kava - the unfolding story: report on a work-in-progressclok.uclan.ac.uk/9455/1/kava...

28
Article Kava - the unfolding story: Report on a work-in- progress. Denham, Alison, McIntyre, Michael and Whitehouse, Jule Available at http://clok.uclan.ac.uk/9455/ Denham, Alison, McIntyre, Michael and Whitehouse, Jule Kava - the unfolding story: Report on a work-in-progress. Journal of Alternative and Complementary Medicine, 8 (3). pp. 237-263.  It is advisable to refer to the publisher’s version if you intend to cite from the work. For more information about UCLan’s research in this area go to http://www.uclan.ac.uk/researchgroups/ and search for <name of research Group>. For information about Research generally at UCLan please go to http://www.uclan.ac.uk/research/ All outputs in CLoK are protected by Intellectual Property Rights law, including Copyright law. Copyright, IPR and Moral Rights for the works on this site are retained by the individual authors and/or other copyright owners. Terms and conditions for use of this material are defined in the http://clok.uclan.ac.uk/policies/ CLoK Central Lancashire online Knowledge www.clok.uclan.ac.uk

Upload: buihanh

Post on 17-Apr-2018

217 views

Category:

Documents


4 download

TRANSCRIPT

Article

Kava shy the unfolding story Report on a workshyinshyprogress

Denham Alison McIntyre Michael and Whitehouse Jule

Available at httpclokuclanacuk9455

Denham Alison McIntyre Michael and Whitehouse Jule Kava shy the unfolding story Report on a workshyinshyprogress Journal of Alternative and Complementary Medicine 8 (3) pp 237shy263

It is advisable to refer to the publisherrsquos version if you intend to cite from the work

For more information about UCLanrsquos research in this area go to httpwwwuclanacukresearchgroups and search for ltname of research Groupgt

For information about Research generally at UCLan please go to httpwwwuclanacukresearch

All outputs in CLoK are protected by Intellectual Property Rights law includingCopyright law Copyright IPR and Moral Rights for the works on this site are retained by the individual authors andor other copyright owners Terms and conditions for use of this material are defined in the httpclokuclanacukpolicies

CLoKCentral Lancashire online Knowledgewwwclokuclanacuk

SPECIAL REPORT

THE JOURNAL OF ALTERNATIVE AND COMPLEMENTARY MEDICINEVolume 8 Number 3 2002 pp 237ndash263copy Mary Ann Liebert Inc

Kavamdashthe Unfolding Story Report on a Work-in-Progress

ALISON DENHAM BA (Soc) MNIMH1

MICHAEL McINTYRE MA FNIMH FRCHM MBAcC2

and JULIE WHITEHOUSE PhD MNIMH3

ABSTRACT

This paper originated as a submission (now updated) to the UK Medicines Control Agencyand Committee of Safety of Medicines (CSM) on January 11 2002 in response to a report circu-lated by the German Federal Institute for Drugs and Medical Products (German initials are BfArM)a compilation of which is summarized in Appendix 2 This agency issued notification in late No-vember 2001 of some thirty adverse events associated with the use of concentrated standardizedpreparations of kava (Piper methysticum Forst f) reported from Germany and Switzerland Ananalysis of the summary of the BfArM case reports (see Appendix 2) shows that these contain du-plications among the cases cited The original submission that was sent to the CSM January 2002has been updated to the version published here This new version was completed in April 2002

As a result of the alert from BfArM the evaluation of kavarsquos safety is now occurring on aworldwide basis and being that this a matter of considerable importance to the public the healthcare community and regulatory authorities as well as to kava farmers throughout Polynesia itis it important to depict this progress report As such this updated report does not provide fi-nal answers The material released by the BfArM is lacking in detail however it is hoped thatthis report will shed light on the kava controversy It is anticipated that there will be further up-dates shortly

This report prepared on behalf of the Traditional Medicines Evaluation Committee a subcom-mittee of the European Herbal Practitioners Association argues that many of the adverse eventscited by the BfArM should not be attributed to kava In addition the report states that the prop-erties of concentrated standardized kava extractsmdashas opposed to preparations that closely ap-proximate those created for traditional usemdashcontribute to causing adverse events This report pro-poses a number of simple measures that will ensure that safe kava preparations may continue tobe available in the United Kingdom

237

BENEFITS OF KAVA

Kava (Piper methysticum Forst f) is an im-portant herbal medicine with unique

properties The ability of herbal practitioners to

care for their patients would be significantly af-fected if this herbrsquos use were restricted or cur-tailed so that practitioners were unable to pre-scribe it Since the early 1900s kava has beenused in Britain by herbal practitioners mainly

1University of Central Lancashire Preston United Kingdom2Midsummer Cottage Clinic Oxon United Kingdom3University of Westminster London United Kingdom

for urinary problems (Ellingwood 1919) Theindications given in the British Herbal Pharma-copoeia (Anonymous 1983) are ldquoCystitis Ure-thritis Rheumatism and Infection of the gen-ito-urinary tractrdquo More recent texts emphasizethe herbrsquos usage as a nervine For example in-dications given by Mills and Bone (2000) areldquoanxiety of nervous origin nervous tensionrestlessness or mild depression of non-psy-chotic origin menopausal symptoms and in-flammation and infections of the genitourinarytracts of both men and women muscular andnervous pain and insomniardquo Kava use as anervine (which is its traditional therapeuticuse) has increased markedly in recent yearspartly because of the evidence for clinical effec-tiveness found in clinical trials (Pittler and Ernst2000) However it is interesting to note thatsuch usage is not completely new Felter (1905)notes kavarsquos application inter alia for treatingneuralgia dizziness and despondency Theherb has a particular value for treating agita-tion and anxiety (Spinella 2001) when othernervines have proved to be ineffective

Clinical trials have been reviewed recentlyby Pittler and Ernst (2000) who stated that for treating anxiety kavarsquos superiority overplacebo was suggested by all seven trials un-der review and that for the three trials in-cluded in the meta-analysis there was a sig-nificant reduction in score on the HamiltonAnxiety Scale The clinical trials were all car-ried out in Germany where kava is prescribedby physicians and sold over the counter inpharmacies for treating anxiety and insomniaA recent review (Loew 2002) listed nine dou-ble-blinded randomized controlled trials in-volving 808 patients and stated that kava wassignificantly superior to placebo for treatingsymptoms associated with anxiety These trialsused a range of products standardized on15ndash70 kavalactones providing a dailydosage of 60ndash210 mg per day of kavalactonesThe use of kava is being increasingly advocatedas an alternative to benzodiazepines for treat-ing anxiety and has been tested in at least onetrial (Loew 2002) The trial was 28 days longand involved 61 people And for example DeLeo et al (2001) showed a significant relativedecrease in anxiety in a double-blinded ran-domized trial on 40 menopausal women when

kava was combined with hormone replacementtherapy This trial lasted for 6 months and thedosage of kava extract was 100 mg per day (con-taining 55 mg of kavain) It has been claimed thatthe German Federal Institute for Drugs and Med-ical Products (German initials are BfArM) citeddoubts about kavarsquos efficacy for treating anxietyas one reason for starting an investigation on theherb (Anonymous 2001) Surprise was ex-pressed in Germany for this reason because itwas claimed (Anonymous 2001) that a recentrandomized controlled trial awaiting publicationhad demonstrated kavarsquos efficacy for this use atdoses that conform to the Commission E mono-graph on the herb (Blumenthal 2000)

The modern use of kava as a nervine is in ac-cordance with its traditional usage in the SouthPacific Kava drinkers report a sense of relax-ation and tranquillity and manifest a sociableattitude (Chanwai 2000) The herb is used as asocial and ceremonial drink among men inPolynesia In modern times as the influence ofthe Christian missionaries decreases kavarsquosuse has become more widespread and fre-quent For instance Kava is an important partof social life in Fiji Fijian spiritual leaders arecalled dauvaguna the literal translation ofwhich means ldquoexpert at drinking kavardquo(Greenwood-Robinson 1999) Kava is not ad-dictive and does not seem to produce the vio-lent antisocial effects that alcohol produces(Lebot et al 1997) However there are seriousconcerns in Australia (Clough et al 2000)about the herbrsquos abuse as a recreational drugwhen it is used to excess It has been associatedwith inactivity confusional states and generalill-health possibly as a result of associated mal-nutrition caused by irregular eating habits(Chanwai 2000) In Aboriginal communitiesmalnutrition is not uncommon and this associ-ation is considered to be unproven

WHAT IS KAVA

Root bark and root of kava are used eitherfresh or dried for its preparation

Based on an archaeological study of charac-teristic drinking bowls it has been proposedthat the herb was first domesticated in Polyne-sia more than 2000 years ago (Green 1974 [cited

DENHAM ET AL238

in Lebot et al 1997]) A comprehensive surveyby Lebot and Levesque in 1989 (cited in Lebotet al 1997) suggests that kava was originallydomesticated in the Vanuatu islands

Kava is now obtained from a wide range ofcultivars in the South Pacific The plant is al-ways propagated vegetatively from stem cut-tings as it does not reproduce via fertilizationmethods There are many cultivars and newstrains continue to be developed by Polynesianfarmers because each strain is considered tohave different psychoactive effects

In recent years there has been increasingpressure on the market because of the increasedworldwide demand for kava In 1998 it wasamong the top-selling herbs in the UnitedStates with a turnover of $8 million repre-senting a growth rate of 473 (Pittler and Ernst2000) It is possible that the current hepatotox-ity problems are to some extent a consequenceof poor quality control caused by a rapid andextraordinary increase in the size of the market(Murray 2000)

There are also concerns that intensive culti-vation and harvesting may affect the quality ofthe product (Aarlbersberg 1999) However be-ing that a range of products is implicated thisis unlikely to provide a satisfactory explanationfor all the reported adverse reactions It ishoped that the BfArM will release relevant dataabout the source and quality assurance of thekava supply in due course

KAVA PHARMACOLOGY

Kava is a well-researched herb The crys-talline resin was first isolated in 1857 by aFrench naval pharmacist and a detailed mono-graph was published in 1886 (Lewin 1886[cited by Lebot et al 1997]) The kavalactonesare considered to be the active constituents andhave been shown in animal studies to have asedative action (Hansel 1996) although themechanism is unclear (Spinella 2001) Thekavalactones are found in the resinous portion(5ndash9) of the plant material and are poorlysoluble in water

Kavalactones are 4-methoxy-2-pyrones withphenyl or styryl substituents at the 6th position(Lebot et al 1997) Total kavalactone content

varies from 3 to 20 dry weight Eighteen lac-tones have so far been isolated (He et al 1997)with the following six compounds (includingfour chiral enantiomers and two achiral enan-tiomers) being considered most important(Haberlain 1997 see Fig 1)

Chiral enantiomers(1) (1)-kavain(2) (1)-dihydrokavain(3) (1)-methysticin(4) (1)-dihydromethysticin

Achiral enantiomers(5) yangonin(6) demethoxyangonin

TRADITIONAL PREPARATION TECHNIQUES

Kava is traditionally prepared in the SouthPacific by grinding and mixing the root or rootbark with cold water This makes an emulsionthat is a suspension of the resinous constituentsin water (Lebot et al 1997) The herb is alsoprepared as an emulsion (also traditionallyprepared without heating) in coconut milk(Johnson 1999) The efficiency of extraction ofthe active constituents which is measured bykavalactone extraction into water varies con-siderably (Murray 2000) but is higher fromfresh material than from the dried plant Kavaconsumed in Vanuatu is reputed to be thestrongest anywhere in the South Pacific The is-lands of Tanna and Pentecost are especiallynoted for their potent brews It is thought thatpart of this increase in potency is the result ofpreparing the drink from the raw fresh rootswhereas in Fiji and elsewhere it is made fromdried rootstock (Greenwood-Robinson 1999)

MODERN PREPARATION TECHNIQUES

The bioavailability of kava constituentsvaries substantially depending on the methodof extraction (Hansel et al 1994 [cited in Schulzet al 1997] Loew 2002) Kava is predomi-nantly available in Germany as a so-called con-centrated standardized extract that is designedto maximize extraction of the kavalactones

KAVA WORK-IN-PROGRESS 239

Schulz noted that to create these concentratedstandardized extracts kava is dissolved in ahigh percentage of an ethanolndashwater mixtureto obtain extracts containing approximately30 kavalactones or alternatively using anacetonendashwater mixture to obtain extracts con-taining approximately 70 kavalactones Whit-ton Whitehouse and Evans (Appendix 1)make the same point about enhanced kavalac-tone extraction using a high ethanol or acetonemedium but detail somewhat different extrac-tion values Both types of products have a herb-to-extract ratio of approximately 12ndash201(Schulz 1997) The dosage recommended by theGerman Commission E is expressed as theequivalent of 60ndash120 mg of kavalactones per day(Blumenthal 1998)

The preparation methods used for stan-dardized products are highly technical and extraction rates vary (Kubatova et al 2001)depending on the solvents used and the tem-perature at which the products are preparedAs Whitton et al propose (Appendix 1) bothefficacy and safety may depend on thekavalactones remaining in their natural formsand on the extraction of the other natural con-stituents of the plant

Varying extraction techniques and preparationmethods may result in an unnatural variation inthe relative concentration of each lactone or inproduction artifacts that may be pathologic to theliver It should be noted that some commercialkava products may also contain syntheticracemic kavain that may have other characteris-tics than the naturally occurring product has Itis clear that these technical matters related to ex-traction techniques require further elucidation

Proponents of concentrated standardizedproducts assert that they provide an effectivedose within a consistent range The traditionalwater-based kava preparations of the Polyne-sian peoples and low-alcohol kava tincturesused by herbal practitioners have been consid-ered to be unreliable because the concentrationof active constituents is relatively low andvaries from kava batch to batch Howeverthere are four relevant counterarguments

(1) The whole range of the constituents mayproduce a more effective and safe medicine(Williamson 2001)

(2) Some constituents not necessarily consid-ered active may enhance the safety of themedicine (Appendix 1)

DENHAM ET AL240

OCH3

H3COyangonin

(+)-methysticin

O O

OCH3

demethoxyangonin

O O

OCH3

OO

OH

O

(+)-dihydromethysticin

OCH3

OO

OH

O

(+)-kavain

OCH3

OH

O

(+)-dihydrokavain

OCH3

OH

O

FIG 1 Kavapyrones of kava (Piper methysticum) Per Haberlein et al 1997

(3) Definitive isolation of the active constituentis elusive in other medicinal plants such asHypericum perforatum (McIntyre 2000Barnes et al 2001)

(4) Herbal practitioners rely on the synergy be-tween a whole range of constituents in theherb or herb within a herbal prescriptionwhich is individually prescribed for a pa-tient This positive interaction may alsohave the benefit of keeping levels of anyone constituent below the safety threshold

LOW-ALCOHOL TINCTURES

The Traditional Medicines Evaluation Com-mittee (TMEC) strongly advocates the use ofextraction techniques that closely approximatethose traditionally used in Polynesia Thiswould require the use of low-alcohol tincturesmade by the traditional cold macerationprocesses common to UK tincture makingThe reasons for this opinion are set out belowthey have also been explored by Whitton et al(Appendix 1)

Tinctures used by herbal practitioners areprepared by macerating dried kava in a mix-ture of water and ethanol It has been shownthat such extracts using 25 ethanol75 wa-ter contain up to 30 times fewer kavalactonesthan the concentrated standardized prepara-tions (Appendix 1) The traditional preparationmethod using a mixture of 25 ethanol75water extracts a wider range of the naturalkava constituents (Appendix 1)

DOSAGE AND OVERDOSAGE

Assuming a 15 25 tincture and an upperlimit of 20 kavalactones in the dried herb(concentrations stated as 3ndash20 see sectionon Pharmacology below) then 500 mL of theherb would contain (100 3 02 3 015) 5 3 g(3000 mg) of kavalactones In addition assum-ing a daily dosage of 5ndash10 mL of the 15 25tincture the daily dose of kavalactonesamounts to a maximum of 30ndash60 mg If the con-centration of kavalactones were lower for ex-ample at approximately 10 as appears to bethe case with regard to Australian kava dis-

cussed by Clough et al (2000) then this dosagefalls to 15ndash30 mg per day It is noteworthy thatthe maximum daily dosage here is equivalentto the minimum daily dosages of the 60ndash210mg kavalactones given in clinical trials of kavaconducted in Germany

Although standardized extracts provide ahigher dosage of kavalactones than low-alco-hol tinctures overdosage in itself is unlikelyto be the cause of hepatotoxicity Strong evi-dence for this is the fact that kava is takendaily at high doses as a normal part of dailylife in large areas of the South Pacific Indeedsome of the accounts of high kava intake areremarkable For example Chanwai (2000) de-scribed the case of a man who was admittedto a hospital after an overdose but ldquoslept offrdquohis symptoms and admitted to consuming upto 40 bowls of a kava preparation per day forthe last 14 years In Australia missionaries in-troduced kava to the aborigines in the 1980sas a substitute for alcohol and it is claimed thatthis has led to abuse of kava Clough et al(2000) discussed this concern and reviewedtwelve studies on the amount of kava usedThe researchers found that social setting ap-pears to determine the amount used with lonedrinkers consuming much more than peoplewho enjoy kava in a family group The re-searchers described normal use of kava in theNorthern Territory as being 37 g of kava pow-der (containing approximately 3800 mg ofkavalactones) per hour with heavy consumersusing approximately 610 g per week preparedas a drink The incidence of serious illness re-sulting from hepatotoxicity associated withregular kava usage would surely have beenobserved by the medical services in Polynesiaand Australia if overdosage of kavalactoneswere the main cause of hepatotoxicity

UNTOWARD EFFECTS

There is a justified concern in Europe thatidiosyncratic hepatotoxicity associated with us-ing some herbal medicines may not be identi-fied because the population that takes herbalmedicines is not large enough to produce suf-ficient cases for the association to be noted Butthe fact that kava remains in traditional usageto such a wide extent is a powerful argument

KAVA WORK-IN-PROGRESS 241

that idiosyncratic hepatotoxicity would havebeen noted

Two postmarketing observation studies inGermany each on more than 3000 people werecited by Pittler and Ernst (2000) in addition tothe abovementioned clinical trials In these ob-servational studies the rate of adverse eventswas 23 (with a daily dose of 120ndash240 mg ofkavalactones) and 15 (with a daily dose of105 mg of kavalactones) The most frequent ad-verse reports were gastrointestinal complaintsallergic skin reactions headaches and photo-sensitivity

There is evidence in the South Pacific of a char-acteristic kava-induced skin disease a scaly rashthat is suggestive of icthyosismdasha condition calledldquokava dermopathyrdquo (Ruze 1990) Although theskin becomes yellow the description does notsuggest an underlying hepatic condition in thatthe patient remains well the rash is not itchyand the condition is ameliorated without treat-ment if heavy use of kava is reduced

The German and Swiss reports cited by theBfArM are of concern because there have beenprevious reports of hepatotoxicity associatedwith the use of some medicinal plants (Larrey1997) The kava case reports from the BfArM(see Appendix 2) include all three of the mainforms of acute damage that can result from ad-verse drug reactions (1) necrosis (2) drug-in-duced hepatitis and (3) cholestatic hepatitis(Hodgson and Levi 1997) This suggests thatthere is a range of causes rather than just onecause in these cases The BfArM case reportshave been circulated worldwide and are cur-rently being evaluated by government agenciesin Europe Australia Canada the UnitedStates and elsewhere We have received anumber of informal case assessments fromthese sources that cannot be specifically citedbecause of their confidential status To achievetransparency and encourage a full debate aboutkava however the BfArM cases are evaluatedin the section entitled Discussion of Cases Re-ported by the BfArM

CRITERIA FOR ASSESSING THE CASE REPORTS

A recent review of the information availableon the case reports (Schmidt and Nahrstedt

2002) is supported by details of the case re-ports on the Web site of the University ofMuenster (wwwuni-muensterdechemiepbkavaanalysehtml)

The criteria for causality assessment of ad-verse reactions used are as follows (Edwardsand Aronson 2000)Probable is defined as

A clinical event including a laboratory testabnormality that occurs in a plausible timerelation to drug administration and that can-not be explained by coincidental or concur-rent disease or other drugs or chemicals

The response to withdrawal of the drug(dechallange) should be clinically plausible

The event must be definitive pharmacolog-ically or phenomenologically using a satis-factory rechallenge procedure if necessary

Possible is defined as

A clinical event including a laboratory testabnormality with a reasonable time relationto administration of the drug but that couldbe explained by concurrent disease or otherdrugs or chemicals

Information on drug withdrawal may belacking or unclear

Unlikely is defined as

A clinical event including a laboratory testabnormality with a temporal relation to theadministration of the drug which makes acausal relation improbable and in whichother drugs chemicals or underlying dis-ease provide plausible explanations

Unassessable is defined as

A report suggesting an adverse reaction thatcannot be judged because information is in-sufficient or contradictory and cannot besupplemented or verified

DISCUSSION OF CASES REPORTED BY THE BfArM

The cases discussed below are analyzed andcategorized by common factors of note withour own assessments

DENHAM ET AL242

(1) Cases of most concern

This group includes five cases 10 13 16 18and 28 According to the assessments made by various government agencies these casescause the most concern and are often cited asbeing probable For this reason these cases aredealt with first As discussed below mostmdashifnot allmdashof these cases have associated factorsthat put this probable categorization into ques-tion

Case 10 This case described necrotizing hep-atitis in a 39-year old female patient with pos-itive reexposure (Strahl et al 1998) During thefirst period that the kava product was takenan oral contraceptive and paroxetine were con-comitant medications It appears that paroxe-tine was not the only antidepressant taken bythis patient who also occasionally took StJohnrsquos wort (Hypericum perforatum) The Exec-utive Summary issued by the Bundesverbandder Arzneimittel Hersteller eV (BAH) andBundesverband der Pharmazeutischen Indus-trie eV (BP) (see Appendix 3) stated ldquoA causalrelationship with kava cannot be excluded butthe patientrsquos history and a potential preexistingliver damage must be taken into account In ad-dition the kava preparation used was not iden-tified by the physicianrdquo

Moreover in this case taking paroxetine incombination with an oral contraceptive maywell have led to overburdening the liver a sit-uation that could have been exacerbated by tak-ing a kava preparation Schmidt and Nahrstedt(2002) suggested that this case may be associ-ated with an immunologic reaction After re-viewing all of the cases in detail Schmidt andNahrstedt (2002) concluded that this is the onlycase for which there was sufficient informationto make an association with kava appear prob-able and for which the dose of kava also con-formed to that recommended by the Com-mission E monograph (Blumenthal 2000)However as discussed below Russmann et al(2001) tested this patient for CYP2D6 and as inCase 16 found this patient to be CYP2D6 defi-cient which appears to have made her partic-ularly vulnerable to the cocktail of drugs shewas taking Given the complicating features ofthis case we submit that this case should beclassed as possible rather than probable

Case 13 This was a case of a 62-year-oldwoman with jaundice The BfArM table (seeAppendix 2) noted regarding concomitantmedication that there was ldquonone denotedrdquo butit was claimed that concomitant medication didexist but was ldquounknownrdquo The insufficiency ofdata provided for this case was highlighted byBfArMrsquos warning note ldquoNo medical messagerdquoIn addition it should be noted that no detailsof the dosage of kava or period of its adminis-tration were apparently recorded for this caseThis is clearly insufficient information onwhich to base a probable assessment

Case 16 This case concerned a 33-year-oldwoman with jaundice The woman wasrecorded as having taken an overdose of alco-hol measured at 60 g (Russman et al 2001) andthen analgesics including paracetamol fol-lowing this alcohol binge Despite the massiveintake of alcohol a liver biopsy indicated thata drug rather than an alcohol induced toxicgenesis However this case like that of Case 10above was discussed by Russman et al (2001)who demonstrated afterward that this patientwas shown to be CYP2D6 deficient which (asdiscussed below) seems to be a risk factor forthe hepatotoxicity that was ascribed to kavaWe submit that given these circumstances thiscase should be considered possible rather thanprobable

Case 18 This case concerned a 50-year-oldman who had necrosis leading to a liver trans-plant This patient took a product manufac-tured by acetone extraction at a dose deliver-ing 210ndash280 mg of kavalactones per day for 15months ldquomoderate alcoholrdquo (ldquomoderaterdquo is notdefined by BfArM) evening primrose(Oenothera biennis) and a yeast preparationThe dosage of kava was well above the Ger-man Commission E recommended dose ofkavalactones (Blumenthal 1998) It was alsorecorded that 500ndash1000 mg of paracetamol wastaken by this patient shortly before transplan-tation The combination of paracetamol and al-cohol plus the very high dose of kava extractedin acetone taken by this man casts seriousdoubts on the assessment of probable in thiscase

Case 28 (BAH) This case concerned a

KAVA WORK-IN-PROGRESS 243

woman age unknown with hepatitis This caseis hard to assess because neither the patientrsquosage nor diagnosis was given and the womanwas taking eleven medications including estra-diol valerate acetylcysteine losartan (which israrely be associated with hepatitis) and mepra-zole (which can be associated with liver diseasealthough this is rare) Omeprazole is metabo-lized by the polymorphic CYP2C19 which is absent in 3 of Caucasians (Flockhart et al2000) The woman was also taking echinacea(Echinacea purpurea) and five products that ap-peared to be for upper respiratory problems Itshould be noted that this patient was taking syn-thetic kavain not kava A comment from BfArMconcerning this case noted ldquorecurrence of the he-patic side-effectsrdquo which has evidently been in-terpreted by some authorities as being equiva-lent to a ldquopositive rechallengerdquo Whether or notthis was actually so was not clear from the datasupplied It appears (Schmidt and Nahrstedt2002) that Case 28 has been published as twocases with slightly different details This is con-fusing and considering that the woman was tak-ing 11 other medications together with a syn-thetic kava (which we submit is not equivalentto natural kava) and that no diagnosis of hercondition was supplied this calls the assessmentof probable in this case into question

(2) Cases associated with taking synthetic kavain

In this category there were 4 cases 1 2 19and 28

In each of these cases the patients concernedwere taking a product made from synthetickavain Although the outcome was hepatitis inall four cases kavain cannot be equated withthe naturally occurring form of kava whichcontains many other constituents that may playan important role in ensuring the safety of thisherb Therefore we submit that no inferenceshould be drawn from these cases Traditionalusage should not be taken as evidence for safeusage of synthetic products

(3) Patients who were taking oral contraceptivepills or hormone replacement therapy (HRT)together with drugs that can also be associatedwith liver damage

The cases in this category were 4 10 12 2021 and 28

Cholestatic jaundice associated with use ofestrogen-containing medications is extremelyrare (Lindberg 1992) but does occur In these6 cases the women were also taking drugs thatcan also be associated with jaundice

Case 4 This case involved a 39-year-oldwoman with jaundice She was on diazepam10 mg PRN for 6 months Some authoritiescalled this case possible Our assessment is thatthe case is unassessable

Case 10 This case involved a 39-year-oldwoman with necrotizing hepatitis For a de-tailed assessment see above

Case 12 This case involved a 37-year-oldwoman with hepatitis She was on 150 mg ofdiclofenac via intramuscular injection Hepato-toxic reactions associated with nonsteroidalanti-inflammatory drug use are extremely rareand concomitant exposure to other hepatotoxicdrugs is considered to be an important factor(Bareille et al 2001) This case of hepatitis isdifficult to interpret because it occurred inBrazil and because ldquoreexposure was said to benegative for all three drugsrdquo We regard thiscase as unassessable

Case 20 This case involved 50-year-oldwoman with necrosis who had a liver trans-plant She had a 20-year history of combinedoral contraceptive use but had changed monthsearlier to estradiol valerate (which was appar-ently taken alone) as HRT She had also startedglimepiride 8 months earlier This is used fortreating type II diabetes and is rarely associatedwith cholestatic jaundice and liver failure Weregard this case as unlikely

Case 21 This case involved a 22-year-oldwoman with necrosis who had a liver trans-plant This woman had changed from Valette(Jenapharm GmbH Jena Germany) (2 mg ofdienogest and 003 mg of ethinlestradiol) toPramino (180215250 mcg of norgestimateand mcg 25 of ethinylestradiol) She also tookrizatriptan if required for migraine reliefRizatriptan should be used with caution in he-patic impairment and avoided if a patient hassevere liver disease Some authorities considerthis case as being possible but our assessment

DENHAM ET AL244

is in view of the other medications taken isthat this case is unassessable

Case 28 This case involved a woman age un-known with hepatitis This case is discussed atlength above As noted above this patient wastaking synthetic kavain not kava

(4) Patients who were taking drugs that can beassociated with liver damage

There were ten cases in this category 1 6 914 15 17 19 23 2627 and 29

Case 1 This case involved a woman age 69with cholestatic hepatitis She was taking pen-toxifylline (which can be associated with intra-hepatic cholestasis) and a diuretic including thepotassium-sparing triamterene (which can beassociated with jaundice) As noted above thispatient was taking synthetic kavain not kavaWe consider this case unassessable

Case 6 This case involved a woman age 50with hepatitis She was taking frusemide(which can be associated with cholestatic jaun-dice) triamterene atenolol and a large dose ofterfenadine (300 mg) The recommended doseof terfenadine in the British National Formu-lary (March 2001) is 60ndash120 mg The Formularyrecommends avoiding this drug in patientswho have hepatic impairment and also says toldquoavoid concomitant administration of drugs li-able to produce electrolyte imbalance such asdiureticsrdquo (British National Formulary 2001)Despite this warning this woman was also tak-ing the diuretic frusemide The InterkantonalenKontrollstelle der Schweiz of Switzerland con-sidered this case of hepatitis to be caused byterfenadine And although some authoritiesregard this case as possible our assessment isthat this case is unlikely

Case 9 This case involved an 81-year-oldwoman who had liver failure and subsequentdeath She was taking hydrochlorothiazide(which can occasionally be associated with in-trahepatic cholestasis) However according toSchmidt and Nahrstedt (2002) there was evi-dence of chronic alcohol abuse and they re-ported that the autopsy showed chronic pan-creatitis that was characteristic of alcoholabuse The autopsy report (Schmidt and

Nahrstedt 2002) apparently said that thesymptoms must have occurred over a periodof at least 18 months The report conceded thatldquohepatic impairment by alcohol [was] not ex-cludedrdquo In these circumstances it seems en-tirely reasonable to claim that this case is un-related to kava use We regard this case asunlikely

Case 14 This case involved a 33-year-oldwoman with hepatitis Cisapride may havebeen taken (which can cause reversible changesthat show in liver-function tests) Cirrhosis ina woman of 33 is an unexplained finding andthe detail in this case is inadequate to elucidateit We consider this case to be unassessable

Case 15 This case involved a 46-year-oldwoman with jaundice She had been taking hy-drochlorothiazide (which can be associatedwith intrahepatic cholestasis) for 55 monthsplus 80 mg of valsartan and 80 mg ofpropanolol per day Some authorities regardthis case as possible but we consider it to beunassessable

Case 17 This case involved a 59-year-oldwoman with jaundice She had taken 100ndash200mg of celecoxib a cyclo-oxygenase-2 inhibitorper day According to the criteria for causalityassessment of adverse reactions some author-ities consider this case to be possible but our as-sessment is that it is unassessable

Case 19 This case involved a 21-year-oldwoman with hepatitis She was taking panto-prazole (which as with omeprazole can be as-sociated with liver disease) She was also takingparacetamol and metoclopramide and had over-dosed on kavain More detail is needed on othermedical conditions suffered by this patient in or-der to interpret this case It is suggested bySchmidt that this woman was using up to 10tablets per day of the product (the recom-mended dose is up to 6 tablets per day) and thatthere was apparently a discussion in her med-ical record file that she may also have used Ec-stasy (substance that has been associated with

KAVA WORK-IN-PROGRESS 245

Personal communication from M McGuffin to M McIn-tyre available as an online document at ehpaglobalnetcouk

fulminant hepatic failure) This case appears tobe unassessable

Case 23 This case involved a 35-year-oldwoman with jaundice According to the BfArM(see Appendix 2) this patient also took parac-etamol but no dosage or details were providedThis case and case 25 in the BfArM listing ap-pear to be the same case Both cases have beenlabeled as possible by some authorities butgiven the lack of information about the dosageof paracetamol and the apparent confusion re-garding cases 23 and 25 we submit that theonly logical assessment is unassessable

Case 2627 This case involved a woman whowas either 38 or 39 yearsrsquo old with hepatitis Itappears that the two cases have been duplicated(Schmidt and Nahrstedt 2002) The confusionwith this case is another example of inaccuratedata provided by the BfArM Information re-garding these cases (or case) depending onwhether the two reports concern the samewoman is unclear Penicillin can be associatedwith hypersensitivity and cholestatic jaundicebut the information given is inadequate to makeany meaningful assessment For this reason weclass this case as unassessable

Case 29 This case involved a 60-year-oldwoman who had a liver transplant This womanwas taking piretamide (which is a loop diuretic)Frusemide another loop diuretic can be associ-ated with cholestatic jaundice According to theBfArM chart (see Appendix 2) she was also tak-ing a sympathomimetic drug etilefrin Thedosage of kava varied but was up to 480ndash1200mg per day (Schmidt and Nahrstedt 2002)which is up to ten times the German Commis-sion E maximum recommended dose (Blumen-thal 1998) Although some authorities have re-garded this case as possible in view of themarked overdosing of kava and the concomitantmedication this case can hardly be said to be areflection on the proper therapeutic use of kava

(5) Cases in which drugs not associated withliver damage herbal medicines or dietarysupplements or kavain alone were taken

This category had eight cases 2 78 11 1322 24 and 25

For these cases detail was limited and theBfArM did not implicate any other drugs ormedications (although this may not be thecase)

All patients in this group apart from the pa-tient in Case 78 for whom no information wasgiven were reported to have made full recov-eries In some of these cases it is not clearwhether the patients were ill or whether thesecases merely recorded raised liver-function en-zymes

Case 2 This case involved a 35-year-old manwith cholestatic hepatitis Concomitant med-ication was ldquounknownrdquo Apart from Cases 18and 30 this is the only case for which it is pos-sible that no other concomitant medication wastaken but there is a marked lack of informationfor this case As noted above this patient wastaking synthetic kavain not kava We regardthis case as unassessable

Case 5 This case involved a woman who waseither 68 or 69 yearsrsquo old with cholestatic he-patitis She was also taking a St Johnrsquos wort(Hypericum perforatum) product which hasbeen associated with CYP3A4 A biopsyshowed ldquoimmunologic hypersensitivityrdquo Thiscase may be regarded as possible but in viewof the immunologic hypersensitivity it maywell have been an idiosyncratic event that wasnot necessarily associated with kava usage

Case 78 This case involved a woman or twowomen ages 72 andor 75 with cholestatic he-patitis These two cases appear to be actuallyone case The woman was taking twoherbalvitamin products one of which in-cluded 06 mg of kavalactones Given the con-fusion involved these ldquocasesrdquo must be re-garded as unassessable

Case 11 This case involved a 59-year-oldwoman who was taking hyoscine butylbro-mide as a suppository Schmidt and Nahrstedt(2002) commented that according to additionalinformation obtained from the BfArM it is un-certain as to whether this patient was taking akava product at all We regard this case asunassessable

DENHAM ET AL246

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

SPECIAL REPORT

THE JOURNAL OF ALTERNATIVE AND COMPLEMENTARY MEDICINEVolume 8 Number 3 2002 pp 237ndash263copy Mary Ann Liebert Inc

Kavamdashthe Unfolding Story Report on a Work-in-Progress

ALISON DENHAM BA (Soc) MNIMH1

MICHAEL McINTYRE MA FNIMH FRCHM MBAcC2

and JULIE WHITEHOUSE PhD MNIMH3

ABSTRACT

This paper originated as a submission (now updated) to the UK Medicines Control Agencyand Committee of Safety of Medicines (CSM) on January 11 2002 in response to a report circu-lated by the German Federal Institute for Drugs and Medical Products (German initials are BfArM)a compilation of which is summarized in Appendix 2 This agency issued notification in late No-vember 2001 of some thirty adverse events associated with the use of concentrated standardizedpreparations of kava (Piper methysticum Forst f) reported from Germany and Switzerland Ananalysis of the summary of the BfArM case reports (see Appendix 2) shows that these contain du-plications among the cases cited The original submission that was sent to the CSM January 2002has been updated to the version published here This new version was completed in April 2002

As a result of the alert from BfArM the evaluation of kavarsquos safety is now occurring on aworldwide basis and being that this a matter of considerable importance to the public the healthcare community and regulatory authorities as well as to kava farmers throughout Polynesia itis it important to depict this progress report As such this updated report does not provide fi-nal answers The material released by the BfArM is lacking in detail however it is hoped thatthis report will shed light on the kava controversy It is anticipated that there will be further up-dates shortly

This report prepared on behalf of the Traditional Medicines Evaluation Committee a subcom-mittee of the European Herbal Practitioners Association argues that many of the adverse eventscited by the BfArM should not be attributed to kava In addition the report states that the prop-erties of concentrated standardized kava extractsmdashas opposed to preparations that closely ap-proximate those created for traditional usemdashcontribute to causing adverse events This report pro-poses a number of simple measures that will ensure that safe kava preparations may continue tobe available in the United Kingdom

237

BENEFITS OF KAVA

Kava (Piper methysticum Forst f) is an im-portant herbal medicine with unique

properties The ability of herbal practitioners to

care for their patients would be significantly af-fected if this herbrsquos use were restricted or cur-tailed so that practitioners were unable to pre-scribe it Since the early 1900s kava has beenused in Britain by herbal practitioners mainly

1University of Central Lancashire Preston United Kingdom2Midsummer Cottage Clinic Oxon United Kingdom3University of Westminster London United Kingdom

for urinary problems (Ellingwood 1919) Theindications given in the British Herbal Pharma-copoeia (Anonymous 1983) are ldquoCystitis Ure-thritis Rheumatism and Infection of the gen-ito-urinary tractrdquo More recent texts emphasizethe herbrsquos usage as a nervine For example in-dications given by Mills and Bone (2000) areldquoanxiety of nervous origin nervous tensionrestlessness or mild depression of non-psy-chotic origin menopausal symptoms and in-flammation and infections of the genitourinarytracts of both men and women muscular andnervous pain and insomniardquo Kava use as anervine (which is its traditional therapeuticuse) has increased markedly in recent yearspartly because of the evidence for clinical effec-tiveness found in clinical trials (Pittler and Ernst2000) However it is interesting to note thatsuch usage is not completely new Felter (1905)notes kavarsquos application inter alia for treatingneuralgia dizziness and despondency Theherb has a particular value for treating agita-tion and anxiety (Spinella 2001) when othernervines have proved to be ineffective

Clinical trials have been reviewed recentlyby Pittler and Ernst (2000) who stated that for treating anxiety kavarsquos superiority overplacebo was suggested by all seven trials un-der review and that for the three trials in-cluded in the meta-analysis there was a sig-nificant reduction in score on the HamiltonAnxiety Scale The clinical trials were all car-ried out in Germany where kava is prescribedby physicians and sold over the counter inpharmacies for treating anxiety and insomniaA recent review (Loew 2002) listed nine dou-ble-blinded randomized controlled trials in-volving 808 patients and stated that kava wassignificantly superior to placebo for treatingsymptoms associated with anxiety These trialsused a range of products standardized on15ndash70 kavalactones providing a dailydosage of 60ndash210 mg per day of kavalactonesThe use of kava is being increasingly advocatedas an alternative to benzodiazepines for treat-ing anxiety and has been tested in at least onetrial (Loew 2002) The trial was 28 days longand involved 61 people And for example DeLeo et al (2001) showed a significant relativedecrease in anxiety in a double-blinded ran-domized trial on 40 menopausal women when

kava was combined with hormone replacementtherapy This trial lasted for 6 months and thedosage of kava extract was 100 mg per day (con-taining 55 mg of kavain) It has been claimed thatthe German Federal Institute for Drugs and Med-ical Products (German initials are BfArM) citeddoubts about kavarsquos efficacy for treating anxietyas one reason for starting an investigation on theherb (Anonymous 2001) Surprise was ex-pressed in Germany for this reason because itwas claimed (Anonymous 2001) that a recentrandomized controlled trial awaiting publicationhad demonstrated kavarsquos efficacy for this use atdoses that conform to the Commission E mono-graph on the herb (Blumenthal 2000)

The modern use of kava as a nervine is in ac-cordance with its traditional usage in the SouthPacific Kava drinkers report a sense of relax-ation and tranquillity and manifest a sociableattitude (Chanwai 2000) The herb is used as asocial and ceremonial drink among men inPolynesia In modern times as the influence ofthe Christian missionaries decreases kavarsquosuse has become more widespread and fre-quent For instance Kava is an important partof social life in Fiji Fijian spiritual leaders arecalled dauvaguna the literal translation ofwhich means ldquoexpert at drinking kavardquo(Greenwood-Robinson 1999) Kava is not ad-dictive and does not seem to produce the vio-lent antisocial effects that alcohol produces(Lebot et al 1997) However there are seriousconcerns in Australia (Clough et al 2000)about the herbrsquos abuse as a recreational drugwhen it is used to excess It has been associatedwith inactivity confusional states and generalill-health possibly as a result of associated mal-nutrition caused by irregular eating habits(Chanwai 2000) In Aboriginal communitiesmalnutrition is not uncommon and this associ-ation is considered to be unproven

WHAT IS KAVA

Root bark and root of kava are used eitherfresh or dried for its preparation

Based on an archaeological study of charac-teristic drinking bowls it has been proposedthat the herb was first domesticated in Polyne-sia more than 2000 years ago (Green 1974 [cited

DENHAM ET AL238

in Lebot et al 1997]) A comprehensive surveyby Lebot and Levesque in 1989 (cited in Lebotet al 1997) suggests that kava was originallydomesticated in the Vanuatu islands

Kava is now obtained from a wide range ofcultivars in the South Pacific The plant is al-ways propagated vegetatively from stem cut-tings as it does not reproduce via fertilizationmethods There are many cultivars and newstrains continue to be developed by Polynesianfarmers because each strain is considered tohave different psychoactive effects

In recent years there has been increasingpressure on the market because of the increasedworldwide demand for kava In 1998 it wasamong the top-selling herbs in the UnitedStates with a turnover of $8 million repre-senting a growth rate of 473 (Pittler and Ernst2000) It is possible that the current hepatotox-ity problems are to some extent a consequenceof poor quality control caused by a rapid andextraordinary increase in the size of the market(Murray 2000)

There are also concerns that intensive culti-vation and harvesting may affect the quality ofthe product (Aarlbersberg 1999) However be-ing that a range of products is implicated thisis unlikely to provide a satisfactory explanationfor all the reported adverse reactions It ishoped that the BfArM will release relevant dataabout the source and quality assurance of thekava supply in due course

KAVA PHARMACOLOGY

Kava is a well-researched herb The crys-talline resin was first isolated in 1857 by aFrench naval pharmacist and a detailed mono-graph was published in 1886 (Lewin 1886[cited by Lebot et al 1997]) The kavalactonesare considered to be the active constituents andhave been shown in animal studies to have asedative action (Hansel 1996) although themechanism is unclear (Spinella 2001) Thekavalactones are found in the resinous portion(5ndash9) of the plant material and are poorlysoluble in water

Kavalactones are 4-methoxy-2-pyrones withphenyl or styryl substituents at the 6th position(Lebot et al 1997) Total kavalactone content

varies from 3 to 20 dry weight Eighteen lac-tones have so far been isolated (He et al 1997)with the following six compounds (includingfour chiral enantiomers and two achiral enan-tiomers) being considered most important(Haberlain 1997 see Fig 1)

Chiral enantiomers(1) (1)-kavain(2) (1)-dihydrokavain(3) (1)-methysticin(4) (1)-dihydromethysticin

Achiral enantiomers(5) yangonin(6) demethoxyangonin

TRADITIONAL PREPARATION TECHNIQUES

Kava is traditionally prepared in the SouthPacific by grinding and mixing the root or rootbark with cold water This makes an emulsionthat is a suspension of the resinous constituentsin water (Lebot et al 1997) The herb is alsoprepared as an emulsion (also traditionallyprepared without heating) in coconut milk(Johnson 1999) The efficiency of extraction ofthe active constituents which is measured bykavalactone extraction into water varies con-siderably (Murray 2000) but is higher fromfresh material than from the dried plant Kavaconsumed in Vanuatu is reputed to be thestrongest anywhere in the South Pacific The is-lands of Tanna and Pentecost are especiallynoted for their potent brews It is thought thatpart of this increase in potency is the result ofpreparing the drink from the raw fresh rootswhereas in Fiji and elsewhere it is made fromdried rootstock (Greenwood-Robinson 1999)

MODERN PREPARATION TECHNIQUES

The bioavailability of kava constituentsvaries substantially depending on the methodof extraction (Hansel et al 1994 [cited in Schulzet al 1997] Loew 2002) Kava is predomi-nantly available in Germany as a so-called con-centrated standardized extract that is designedto maximize extraction of the kavalactones

KAVA WORK-IN-PROGRESS 239

Schulz noted that to create these concentratedstandardized extracts kava is dissolved in ahigh percentage of an ethanolndashwater mixtureto obtain extracts containing approximately30 kavalactones or alternatively using anacetonendashwater mixture to obtain extracts con-taining approximately 70 kavalactones Whit-ton Whitehouse and Evans (Appendix 1)make the same point about enhanced kavalac-tone extraction using a high ethanol or acetonemedium but detail somewhat different extrac-tion values Both types of products have a herb-to-extract ratio of approximately 12ndash201(Schulz 1997) The dosage recommended by theGerman Commission E is expressed as theequivalent of 60ndash120 mg of kavalactones per day(Blumenthal 1998)

The preparation methods used for stan-dardized products are highly technical and extraction rates vary (Kubatova et al 2001)depending on the solvents used and the tem-perature at which the products are preparedAs Whitton et al propose (Appendix 1) bothefficacy and safety may depend on thekavalactones remaining in their natural formsand on the extraction of the other natural con-stituents of the plant

Varying extraction techniques and preparationmethods may result in an unnatural variation inthe relative concentration of each lactone or inproduction artifacts that may be pathologic to theliver It should be noted that some commercialkava products may also contain syntheticracemic kavain that may have other characteris-tics than the naturally occurring product has Itis clear that these technical matters related to ex-traction techniques require further elucidation

Proponents of concentrated standardizedproducts assert that they provide an effectivedose within a consistent range The traditionalwater-based kava preparations of the Polyne-sian peoples and low-alcohol kava tincturesused by herbal practitioners have been consid-ered to be unreliable because the concentrationof active constituents is relatively low andvaries from kava batch to batch Howeverthere are four relevant counterarguments

(1) The whole range of the constituents mayproduce a more effective and safe medicine(Williamson 2001)

(2) Some constituents not necessarily consid-ered active may enhance the safety of themedicine (Appendix 1)

DENHAM ET AL240

OCH3

H3COyangonin

(+)-methysticin

O O

OCH3

demethoxyangonin

O O

OCH3

OO

OH

O

(+)-dihydromethysticin

OCH3

OO

OH

O

(+)-kavain

OCH3

OH

O

(+)-dihydrokavain

OCH3

OH

O

FIG 1 Kavapyrones of kava (Piper methysticum) Per Haberlein et al 1997

(3) Definitive isolation of the active constituentis elusive in other medicinal plants such asHypericum perforatum (McIntyre 2000Barnes et al 2001)

(4) Herbal practitioners rely on the synergy be-tween a whole range of constituents in theherb or herb within a herbal prescriptionwhich is individually prescribed for a pa-tient This positive interaction may alsohave the benefit of keeping levels of anyone constituent below the safety threshold

LOW-ALCOHOL TINCTURES

The Traditional Medicines Evaluation Com-mittee (TMEC) strongly advocates the use ofextraction techniques that closely approximatethose traditionally used in Polynesia Thiswould require the use of low-alcohol tincturesmade by the traditional cold macerationprocesses common to UK tincture makingThe reasons for this opinion are set out belowthey have also been explored by Whitton et al(Appendix 1)

Tinctures used by herbal practitioners areprepared by macerating dried kava in a mix-ture of water and ethanol It has been shownthat such extracts using 25 ethanol75 wa-ter contain up to 30 times fewer kavalactonesthan the concentrated standardized prepara-tions (Appendix 1) The traditional preparationmethod using a mixture of 25 ethanol75water extracts a wider range of the naturalkava constituents (Appendix 1)

DOSAGE AND OVERDOSAGE

Assuming a 15 25 tincture and an upperlimit of 20 kavalactones in the dried herb(concentrations stated as 3ndash20 see sectionon Pharmacology below) then 500 mL of theherb would contain (100 3 02 3 015) 5 3 g(3000 mg) of kavalactones In addition assum-ing a daily dosage of 5ndash10 mL of the 15 25tincture the daily dose of kavalactonesamounts to a maximum of 30ndash60 mg If the con-centration of kavalactones were lower for ex-ample at approximately 10 as appears to bethe case with regard to Australian kava dis-

cussed by Clough et al (2000) then this dosagefalls to 15ndash30 mg per day It is noteworthy thatthe maximum daily dosage here is equivalentto the minimum daily dosages of the 60ndash210mg kavalactones given in clinical trials of kavaconducted in Germany

Although standardized extracts provide ahigher dosage of kavalactones than low-alco-hol tinctures overdosage in itself is unlikelyto be the cause of hepatotoxicity Strong evi-dence for this is the fact that kava is takendaily at high doses as a normal part of dailylife in large areas of the South Pacific Indeedsome of the accounts of high kava intake areremarkable For example Chanwai (2000) de-scribed the case of a man who was admittedto a hospital after an overdose but ldquoslept offrdquohis symptoms and admitted to consuming upto 40 bowls of a kava preparation per day forthe last 14 years In Australia missionaries in-troduced kava to the aborigines in the 1980sas a substitute for alcohol and it is claimed thatthis has led to abuse of kava Clough et al(2000) discussed this concern and reviewedtwelve studies on the amount of kava usedThe researchers found that social setting ap-pears to determine the amount used with lonedrinkers consuming much more than peoplewho enjoy kava in a family group The re-searchers described normal use of kava in theNorthern Territory as being 37 g of kava pow-der (containing approximately 3800 mg ofkavalactones) per hour with heavy consumersusing approximately 610 g per week preparedas a drink The incidence of serious illness re-sulting from hepatotoxicity associated withregular kava usage would surely have beenobserved by the medical services in Polynesiaand Australia if overdosage of kavalactoneswere the main cause of hepatotoxicity

UNTOWARD EFFECTS

There is a justified concern in Europe thatidiosyncratic hepatotoxicity associated with us-ing some herbal medicines may not be identi-fied because the population that takes herbalmedicines is not large enough to produce suf-ficient cases for the association to be noted Butthe fact that kava remains in traditional usageto such a wide extent is a powerful argument

KAVA WORK-IN-PROGRESS 241

that idiosyncratic hepatotoxicity would havebeen noted

Two postmarketing observation studies inGermany each on more than 3000 people werecited by Pittler and Ernst (2000) in addition tothe abovementioned clinical trials In these ob-servational studies the rate of adverse eventswas 23 (with a daily dose of 120ndash240 mg ofkavalactones) and 15 (with a daily dose of105 mg of kavalactones) The most frequent ad-verse reports were gastrointestinal complaintsallergic skin reactions headaches and photo-sensitivity

There is evidence in the South Pacific of a char-acteristic kava-induced skin disease a scaly rashthat is suggestive of icthyosismdasha condition calledldquokava dermopathyrdquo (Ruze 1990) Although theskin becomes yellow the description does notsuggest an underlying hepatic condition in thatthe patient remains well the rash is not itchyand the condition is ameliorated without treat-ment if heavy use of kava is reduced

The German and Swiss reports cited by theBfArM are of concern because there have beenprevious reports of hepatotoxicity associatedwith the use of some medicinal plants (Larrey1997) The kava case reports from the BfArM(see Appendix 2) include all three of the mainforms of acute damage that can result from ad-verse drug reactions (1) necrosis (2) drug-in-duced hepatitis and (3) cholestatic hepatitis(Hodgson and Levi 1997) This suggests thatthere is a range of causes rather than just onecause in these cases The BfArM case reportshave been circulated worldwide and are cur-rently being evaluated by government agenciesin Europe Australia Canada the UnitedStates and elsewhere We have received anumber of informal case assessments fromthese sources that cannot be specifically citedbecause of their confidential status To achievetransparency and encourage a full debate aboutkava however the BfArM cases are evaluatedin the section entitled Discussion of Cases Re-ported by the BfArM

CRITERIA FOR ASSESSING THE CASE REPORTS

A recent review of the information availableon the case reports (Schmidt and Nahrstedt

2002) is supported by details of the case re-ports on the Web site of the University ofMuenster (wwwuni-muensterdechemiepbkavaanalysehtml)

The criteria for causality assessment of ad-verse reactions used are as follows (Edwardsand Aronson 2000)Probable is defined as

A clinical event including a laboratory testabnormality that occurs in a plausible timerelation to drug administration and that can-not be explained by coincidental or concur-rent disease or other drugs or chemicals

The response to withdrawal of the drug(dechallange) should be clinically plausible

The event must be definitive pharmacolog-ically or phenomenologically using a satis-factory rechallenge procedure if necessary

Possible is defined as

A clinical event including a laboratory testabnormality with a reasonable time relationto administration of the drug but that couldbe explained by concurrent disease or otherdrugs or chemicals

Information on drug withdrawal may belacking or unclear

Unlikely is defined as

A clinical event including a laboratory testabnormality with a temporal relation to theadministration of the drug which makes acausal relation improbable and in whichother drugs chemicals or underlying dis-ease provide plausible explanations

Unassessable is defined as

A report suggesting an adverse reaction thatcannot be judged because information is in-sufficient or contradictory and cannot besupplemented or verified

DISCUSSION OF CASES REPORTED BY THE BfArM

The cases discussed below are analyzed andcategorized by common factors of note withour own assessments

DENHAM ET AL242

(1) Cases of most concern

This group includes five cases 10 13 16 18and 28 According to the assessments made by various government agencies these casescause the most concern and are often cited asbeing probable For this reason these cases aredealt with first As discussed below mostmdashifnot allmdashof these cases have associated factorsthat put this probable categorization into ques-tion

Case 10 This case described necrotizing hep-atitis in a 39-year old female patient with pos-itive reexposure (Strahl et al 1998) During thefirst period that the kava product was takenan oral contraceptive and paroxetine were con-comitant medications It appears that paroxe-tine was not the only antidepressant taken bythis patient who also occasionally took StJohnrsquos wort (Hypericum perforatum) The Exec-utive Summary issued by the Bundesverbandder Arzneimittel Hersteller eV (BAH) andBundesverband der Pharmazeutischen Indus-trie eV (BP) (see Appendix 3) stated ldquoA causalrelationship with kava cannot be excluded butthe patientrsquos history and a potential preexistingliver damage must be taken into account In ad-dition the kava preparation used was not iden-tified by the physicianrdquo

Moreover in this case taking paroxetine incombination with an oral contraceptive maywell have led to overburdening the liver a sit-uation that could have been exacerbated by tak-ing a kava preparation Schmidt and Nahrstedt(2002) suggested that this case may be associ-ated with an immunologic reaction After re-viewing all of the cases in detail Schmidt andNahrstedt (2002) concluded that this is the onlycase for which there was sufficient informationto make an association with kava appear prob-able and for which the dose of kava also con-formed to that recommended by the Com-mission E monograph (Blumenthal 2000)However as discussed below Russmann et al(2001) tested this patient for CYP2D6 and as inCase 16 found this patient to be CYP2D6 defi-cient which appears to have made her partic-ularly vulnerable to the cocktail of drugs shewas taking Given the complicating features ofthis case we submit that this case should beclassed as possible rather than probable

Case 13 This was a case of a 62-year-oldwoman with jaundice The BfArM table (seeAppendix 2) noted regarding concomitantmedication that there was ldquonone denotedrdquo butit was claimed that concomitant medication didexist but was ldquounknownrdquo The insufficiency ofdata provided for this case was highlighted byBfArMrsquos warning note ldquoNo medical messagerdquoIn addition it should be noted that no detailsof the dosage of kava or period of its adminis-tration were apparently recorded for this caseThis is clearly insufficient information onwhich to base a probable assessment

Case 16 This case concerned a 33-year-oldwoman with jaundice The woman wasrecorded as having taken an overdose of alco-hol measured at 60 g (Russman et al 2001) andthen analgesics including paracetamol fol-lowing this alcohol binge Despite the massiveintake of alcohol a liver biopsy indicated thata drug rather than an alcohol induced toxicgenesis However this case like that of Case 10above was discussed by Russman et al (2001)who demonstrated afterward that this patientwas shown to be CYP2D6 deficient which (asdiscussed below) seems to be a risk factor forthe hepatotoxicity that was ascribed to kavaWe submit that given these circumstances thiscase should be considered possible rather thanprobable

Case 18 This case concerned a 50-year-oldman who had necrosis leading to a liver trans-plant This patient took a product manufac-tured by acetone extraction at a dose deliver-ing 210ndash280 mg of kavalactones per day for 15months ldquomoderate alcoholrdquo (ldquomoderaterdquo is notdefined by BfArM) evening primrose(Oenothera biennis) and a yeast preparationThe dosage of kava was well above the Ger-man Commission E recommended dose ofkavalactones (Blumenthal 1998) It was alsorecorded that 500ndash1000 mg of paracetamol wastaken by this patient shortly before transplan-tation The combination of paracetamol and al-cohol plus the very high dose of kava extractedin acetone taken by this man casts seriousdoubts on the assessment of probable in thiscase

Case 28 (BAH) This case concerned a

KAVA WORK-IN-PROGRESS 243

woman age unknown with hepatitis This caseis hard to assess because neither the patientrsquosage nor diagnosis was given and the womanwas taking eleven medications including estra-diol valerate acetylcysteine losartan (which israrely be associated with hepatitis) and mepra-zole (which can be associated with liver diseasealthough this is rare) Omeprazole is metabo-lized by the polymorphic CYP2C19 which is absent in 3 of Caucasians (Flockhart et al2000) The woman was also taking echinacea(Echinacea purpurea) and five products that ap-peared to be for upper respiratory problems Itshould be noted that this patient was taking syn-thetic kavain not kava A comment from BfArMconcerning this case noted ldquorecurrence of the he-patic side-effectsrdquo which has evidently been in-terpreted by some authorities as being equiva-lent to a ldquopositive rechallengerdquo Whether or notthis was actually so was not clear from the datasupplied It appears (Schmidt and Nahrstedt2002) that Case 28 has been published as twocases with slightly different details This is con-fusing and considering that the woman was tak-ing 11 other medications together with a syn-thetic kava (which we submit is not equivalentto natural kava) and that no diagnosis of hercondition was supplied this calls the assessmentof probable in this case into question

(2) Cases associated with taking synthetic kavain

In this category there were 4 cases 1 2 19and 28

In each of these cases the patients concernedwere taking a product made from synthetickavain Although the outcome was hepatitis inall four cases kavain cannot be equated withthe naturally occurring form of kava whichcontains many other constituents that may playan important role in ensuring the safety of thisherb Therefore we submit that no inferenceshould be drawn from these cases Traditionalusage should not be taken as evidence for safeusage of synthetic products

(3) Patients who were taking oral contraceptivepills or hormone replacement therapy (HRT)together with drugs that can also be associatedwith liver damage

The cases in this category were 4 10 12 2021 and 28

Cholestatic jaundice associated with use ofestrogen-containing medications is extremelyrare (Lindberg 1992) but does occur In these6 cases the women were also taking drugs thatcan also be associated with jaundice

Case 4 This case involved a 39-year-oldwoman with jaundice She was on diazepam10 mg PRN for 6 months Some authoritiescalled this case possible Our assessment is thatthe case is unassessable

Case 10 This case involved a 39-year-oldwoman with necrotizing hepatitis For a de-tailed assessment see above

Case 12 This case involved a 37-year-oldwoman with hepatitis She was on 150 mg ofdiclofenac via intramuscular injection Hepato-toxic reactions associated with nonsteroidalanti-inflammatory drug use are extremely rareand concomitant exposure to other hepatotoxicdrugs is considered to be an important factor(Bareille et al 2001) This case of hepatitis isdifficult to interpret because it occurred inBrazil and because ldquoreexposure was said to benegative for all three drugsrdquo We regard thiscase as unassessable

Case 20 This case involved 50-year-oldwoman with necrosis who had a liver trans-plant She had a 20-year history of combinedoral contraceptive use but had changed monthsearlier to estradiol valerate (which was appar-ently taken alone) as HRT She had also startedglimepiride 8 months earlier This is used fortreating type II diabetes and is rarely associatedwith cholestatic jaundice and liver failure Weregard this case as unlikely

Case 21 This case involved a 22-year-oldwoman with necrosis who had a liver trans-plant This woman had changed from Valette(Jenapharm GmbH Jena Germany) (2 mg ofdienogest and 003 mg of ethinlestradiol) toPramino (180215250 mcg of norgestimateand mcg 25 of ethinylestradiol) She also tookrizatriptan if required for migraine reliefRizatriptan should be used with caution in he-patic impairment and avoided if a patient hassevere liver disease Some authorities considerthis case as being possible but our assessment

DENHAM ET AL244

is in view of the other medications taken isthat this case is unassessable

Case 28 This case involved a woman age un-known with hepatitis This case is discussed atlength above As noted above this patient wastaking synthetic kavain not kava

(4) Patients who were taking drugs that can beassociated with liver damage

There were ten cases in this category 1 6 914 15 17 19 23 2627 and 29

Case 1 This case involved a woman age 69with cholestatic hepatitis She was taking pen-toxifylline (which can be associated with intra-hepatic cholestasis) and a diuretic including thepotassium-sparing triamterene (which can beassociated with jaundice) As noted above thispatient was taking synthetic kavain not kavaWe consider this case unassessable

Case 6 This case involved a woman age 50with hepatitis She was taking frusemide(which can be associated with cholestatic jaun-dice) triamterene atenolol and a large dose ofterfenadine (300 mg) The recommended doseof terfenadine in the British National Formu-lary (March 2001) is 60ndash120 mg The Formularyrecommends avoiding this drug in patientswho have hepatic impairment and also says toldquoavoid concomitant administration of drugs li-able to produce electrolyte imbalance such asdiureticsrdquo (British National Formulary 2001)Despite this warning this woman was also tak-ing the diuretic frusemide The InterkantonalenKontrollstelle der Schweiz of Switzerland con-sidered this case of hepatitis to be caused byterfenadine And although some authoritiesregard this case as possible our assessment isthat this case is unlikely

Case 9 This case involved an 81-year-oldwoman who had liver failure and subsequentdeath She was taking hydrochlorothiazide(which can occasionally be associated with in-trahepatic cholestasis) However according toSchmidt and Nahrstedt (2002) there was evi-dence of chronic alcohol abuse and they re-ported that the autopsy showed chronic pan-creatitis that was characteristic of alcoholabuse The autopsy report (Schmidt and

Nahrstedt 2002) apparently said that thesymptoms must have occurred over a periodof at least 18 months The report conceded thatldquohepatic impairment by alcohol [was] not ex-cludedrdquo In these circumstances it seems en-tirely reasonable to claim that this case is un-related to kava use We regard this case asunlikely

Case 14 This case involved a 33-year-oldwoman with hepatitis Cisapride may havebeen taken (which can cause reversible changesthat show in liver-function tests) Cirrhosis ina woman of 33 is an unexplained finding andthe detail in this case is inadequate to elucidateit We consider this case to be unassessable

Case 15 This case involved a 46-year-oldwoman with jaundice She had been taking hy-drochlorothiazide (which can be associatedwith intrahepatic cholestasis) for 55 monthsplus 80 mg of valsartan and 80 mg ofpropanolol per day Some authorities regardthis case as possible but we consider it to beunassessable

Case 17 This case involved a 59-year-oldwoman with jaundice She had taken 100ndash200mg of celecoxib a cyclo-oxygenase-2 inhibitorper day According to the criteria for causalityassessment of adverse reactions some author-ities consider this case to be possible but our as-sessment is that it is unassessable

Case 19 This case involved a 21-year-oldwoman with hepatitis She was taking panto-prazole (which as with omeprazole can be as-sociated with liver disease) She was also takingparacetamol and metoclopramide and had over-dosed on kavain More detail is needed on othermedical conditions suffered by this patient in or-der to interpret this case It is suggested bySchmidt that this woman was using up to 10tablets per day of the product (the recom-mended dose is up to 6 tablets per day) and thatthere was apparently a discussion in her med-ical record file that she may also have used Ec-stasy (substance that has been associated with

KAVA WORK-IN-PROGRESS 245

Personal communication from M McGuffin to M McIn-tyre available as an online document at ehpaglobalnetcouk

fulminant hepatic failure) This case appears tobe unassessable

Case 23 This case involved a 35-year-oldwoman with jaundice According to the BfArM(see Appendix 2) this patient also took parac-etamol but no dosage or details were providedThis case and case 25 in the BfArM listing ap-pear to be the same case Both cases have beenlabeled as possible by some authorities butgiven the lack of information about the dosageof paracetamol and the apparent confusion re-garding cases 23 and 25 we submit that theonly logical assessment is unassessable

Case 2627 This case involved a woman whowas either 38 or 39 yearsrsquo old with hepatitis Itappears that the two cases have been duplicated(Schmidt and Nahrstedt 2002) The confusionwith this case is another example of inaccuratedata provided by the BfArM Information re-garding these cases (or case) depending onwhether the two reports concern the samewoman is unclear Penicillin can be associatedwith hypersensitivity and cholestatic jaundicebut the information given is inadequate to makeany meaningful assessment For this reason weclass this case as unassessable

Case 29 This case involved a 60-year-oldwoman who had a liver transplant This womanwas taking piretamide (which is a loop diuretic)Frusemide another loop diuretic can be associ-ated with cholestatic jaundice According to theBfArM chart (see Appendix 2) she was also tak-ing a sympathomimetic drug etilefrin Thedosage of kava varied but was up to 480ndash1200mg per day (Schmidt and Nahrstedt 2002)which is up to ten times the German Commis-sion E maximum recommended dose (Blumen-thal 1998) Although some authorities have re-garded this case as possible in view of themarked overdosing of kava and the concomitantmedication this case can hardly be said to be areflection on the proper therapeutic use of kava

(5) Cases in which drugs not associated withliver damage herbal medicines or dietarysupplements or kavain alone were taken

This category had eight cases 2 78 11 1322 24 and 25

For these cases detail was limited and theBfArM did not implicate any other drugs ormedications (although this may not be thecase)

All patients in this group apart from the pa-tient in Case 78 for whom no information wasgiven were reported to have made full recov-eries In some of these cases it is not clearwhether the patients were ill or whether thesecases merely recorded raised liver-function en-zymes

Case 2 This case involved a 35-year-old manwith cholestatic hepatitis Concomitant med-ication was ldquounknownrdquo Apart from Cases 18and 30 this is the only case for which it is pos-sible that no other concomitant medication wastaken but there is a marked lack of informationfor this case As noted above this patient wastaking synthetic kavain not kava We regardthis case as unassessable

Case 5 This case involved a woman who waseither 68 or 69 yearsrsquo old with cholestatic he-patitis She was also taking a St Johnrsquos wort(Hypericum perforatum) product which hasbeen associated with CYP3A4 A biopsyshowed ldquoimmunologic hypersensitivityrdquo Thiscase may be regarded as possible but in viewof the immunologic hypersensitivity it maywell have been an idiosyncratic event that wasnot necessarily associated with kava usage

Case 78 This case involved a woman or twowomen ages 72 andor 75 with cholestatic he-patitis These two cases appear to be actuallyone case The woman was taking twoherbalvitamin products one of which in-cluded 06 mg of kavalactones Given the con-fusion involved these ldquocasesrdquo must be re-garded as unassessable

Case 11 This case involved a 59-year-oldwoman who was taking hyoscine butylbro-mide as a suppository Schmidt and Nahrstedt(2002) commented that according to additionalinformation obtained from the BfArM it is un-certain as to whether this patient was taking akava product at all We regard this case asunassessable

DENHAM ET AL246

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

for urinary problems (Ellingwood 1919) Theindications given in the British Herbal Pharma-copoeia (Anonymous 1983) are ldquoCystitis Ure-thritis Rheumatism and Infection of the gen-ito-urinary tractrdquo More recent texts emphasizethe herbrsquos usage as a nervine For example in-dications given by Mills and Bone (2000) areldquoanxiety of nervous origin nervous tensionrestlessness or mild depression of non-psy-chotic origin menopausal symptoms and in-flammation and infections of the genitourinarytracts of both men and women muscular andnervous pain and insomniardquo Kava use as anervine (which is its traditional therapeuticuse) has increased markedly in recent yearspartly because of the evidence for clinical effec-tiveness found in clinical trials (Pittler and Ernst2000) However it is interesting to note thatsuch usage is not completely new Felter (1905)notes kavarsquos application inter alia for treatingneuralgia dizziness and despondency Theherb has a particular value for treating agita-tion and anxiety (Spinella 2001) when othernervines have proved to be ineffective

Clinical trials have been reviewed recentlyby Pittler and Ernst (2000) who stated that for treating anxiety kavarsquos superiority overplacebo was suggested by all seven trials un-der review and that for the three trials in-cluded in the meta-analysis there was a sig-nificant reduction in score on the HamiltonAnxiety Scale The clinical trials were all car-ried out in Germany where kava is prescribedby physicians and sold over the counter inpharmacies for treating anxiety and insomniaA recent review (Loew 2002) listed nine dou-ble-blinded randomized controlled trials in-volving 808 patients and stated that kava wassignificantly superior to placebo for treatingsymptoms associated with anxiety These trialsused a range of products standardized on15ndash70 kavalactones providing a dailydosage of 60ndash210 mg per day of kavalactonesThe use of kava is being increasingly advocatedas an alternative to benzodiazepines for treat-ing anxiety and has been tested in at least onetrial (Loew 2002) The trial was 28 days longand involved 61 people And for example DeLeo et al (2001) showed a significant relativedecrease in anxiety in a double-blinded ran-domized trial on 40 menopausal women when

kava was combined with hormone replacementtherapy This trial lasted for 6 months and thedosage of kava extract was 100 mg per day (con-taining 55 mg of kavain) It has been claimed thatthe German Federal Institute for Drugs and Med-ical Products (German initials are BfArM) citeddoubts about kavarsquos efficacy for treating anxietyas one reason for starting an investigation on theherb (Anonymous 2001) Surprise was ex-pressed in Germany for this reason because itwas claimed (Anonymous 2001) that a recentrandomized controlled trial awaiting publicationhad demonstrated kavarsquos efficacy for this use atdoses that conform to the Commission E mono-graph on the herb (Blumenthal 2000)

The modern use of kava as a nervine is in ac-cordance with its traditional usage in the SouthPacific Kava drinkers report a sense of relax-ation and tranquillity and manifest a sociableattitude (Chanwai 2000) The herb is used as asocial and ceremonial drink among men inPolynesia In modern times as the influence ofthe Christian missionaries decreases kavarsquosuse has become more widespread and fre-quent For instance Kava is an important partof social life in Fiji Fijian spiritual leaders arecalled dauvaguna the literal translation ofwhich means ldquoexpert at drinking kavardquo(Greenwood-Robinson 1999) Kava is not ad-dictive and does not seem to produce the vio-lent antisocial effects that alcohol produces(Lebot et al 1997) However there are seriousconcerns in Australia (Clough et al 2000)about the herbrsquos abuse as a recreational drugwhen it is used to excess It has been associatedwith inactivity confusional states and generalill-health possibly as a result of associated mal-nutrition caused by irregular eating habits(Chanwai 2000) In Aboriginal communitiesmalnutrition is not uncommon and this associ-ation is considered to be unproven

WHAT IS KAVA

Root bark and root of kava are used eitherfresh or dried for its preparation

Based on an archaeological study of charac-teristic drinking bowls it has been proposedthat the herb was first domesticated in Polyne-sia more than 2000 years ago (Green 1974 [cited

DENHAM ET AL238

in Lebot et al 1997]) A comprehensive surveyby Lebot and Levesque in 1989 (cited in Lebotet al 1997) suggests that kava was originallydomesticated in the Vanuatu islands

Kava is now obtained from a wide range ofcultivars in the South Pacific The plant is al-ways propagated vegetatively from stem cut-tings as it does not reproduce via fertilizationmethods There are many cultivars and newstrains continue to be developed by Polynesianfarmers because each strain is considered tohave different psychoactive effects

In recent years there has been increasingpressure on the market because of the increasedworldwide demand for kava In 1998 it wasamong the top-selling herbs in the UnitedStates with a turnover of $8 million repre-senting a growth rate of 473 (Pittler and Ernst2000) It is possible that the current hepatotox-ity problems are to some extent a consequenceof poor quality control caused by a rapid andextraordinary increase in the size of the market(Murray 2000)

There are also concerns that intensive culti-vation and harvesting may affect the quality ofthe product (Aarlbersberg 1999) However be-ing that a range of products is implicated thisis unlikely to provide a satisfactory explanationfor all the reported adverse reactions It ishoped that the BfArM will release relevant dataabout the source and quality assurance of thekava supply in due course

KAVA PHARMACOLOGY

Kava is a well-researched herb The crys-talline resin was first isolated in 1857 by aFrench naval pharmacist and a detailed mono-graph was published in 1886 (Lewin 1886[cited by Lebot et al 1997]) The kavalactonesare considered to be the active constituents andhave been shown in animal studies to have asedative action (Hansel 1996) although themechanism is unclear (Spinella 2001) Thekavalactones are found in the resinous portion(5ndash9) of the plant material and are poorlysoluble in water

Kavalactones are 4-methoxy-2-pyrones withphenyl or styryl substituents at the 6th position(Lebot et al 1997) Total kavalactone content

varies from 3 to 20 dry weight Eighteen lac-tones have so far been isolated (He et al 1997)with the following six compounds (includingfour chiral enantiomers and two achiral enan-tiomers) being considered most important(Haberlain 1997 see Fig 1)

Chiral enantiomers(1) (1)-kavain(2) (1)-dihydrokavain(3) (1)-methysticin(4) (1)-dihydromethysticin

Achiral enantiomers(5) yangonin(6) demethoxyangonin

TRADITIONAL PREPARATION TECHNIQUES

Kava is traditionally prepared in the SouthPacific by grinding and mixing the root or rootbark with cold water This makes an emulsionthat is a suspension of the resinous constituentsin water (Lebot et al 1997) The herb is alsoprepared as an emulsion (also traditionallyprepared without heating) in coconut milk(Johnson 1999) The efficiency of extraction ofthe active constituents which is measured bykavalactone extraction into water varies con-siderably (Murray 2000) but is higher fromfresh material than from the dried plant Kavaconsumed in Vanuatu is reputed to be thestrongest anywhere in the South Pacific The is-lands of Tanna and Pentecost are especiallynoted for their potent brews It is thought thatpart of this increase in potency is the result ofpreparing the drink from the raw fresh rootswhereas in Fiji and elsewhere it is made fromdried rootstock (Greenwood-Robinson 1999)

MODERN PREPARATION TECHNIQUES

The bioavailability of kava constituentsvaries substantially depending on the methodof extraction (Hansel et al 1994 [cited in Schulzet al 1997] Loew 2002) Kava is predomi-nantly available in Germany as a so-called con-centrated standardized extract that is designedto maximize extraction of the kavalactones

KAVA WORK-IN-PROGRESS 239

Schulz noted that to create these concentratedstandardized extracts kava is dissolved in ahigh percentage of an ethanolndashwater mixtureto obtain extracts containing approximately30 kavalactones or alternatively using anacetonendashwater mixture to obtain extracts con-taining approximately 70 kavalactones Whit-ton Whitehouse and Evans (Appendix 1)make the same point about enhanced kavalac-tone extraction using a high ethanol or acetonemedium but detail somewhat different extrac-tion values Both types of products have a herb-to-extract ratio of approximately 12ndash201(Schulz 1997) The dosage recommended by theGerman Commission E is expressed as theequivalent of 60ndash120 mg of kavalactones per day(Blumenthal 1998)

The preparation methods used for stan-dardized products are highly technical and extraction rates vary (Kubatova et al 2001)depending on the solvents used and the tem-perature at which the products are preparedAs Whitton et al propose (Appendix 1) bothefficacy and safety may depend on thekavalactones remaining in their natural formsand on the extraction of the other natural con-stituents of the plant

Varying extraction techniques and preparationmethods may result in an unnatural variation inthe relative concentration of each lactone or inproduction artifacts that may be pathologic to theliver It should be noted that some commercialkava products may also contain syntheticracemic kavain that may have other characteris-tics than the naturally occurring product has Itis clear that these technical matters related to ex-traction techniques require further elucidation

Proponents of concentrated standardizedproducts assert that they provide an effectivedose within a consistent range The traditionalwater-based kava preparations of the Polyne-sian peoples and low-alcohol kava tincturesused by herbal practitioners have been consid-ered to be unreliable because the concentrationof active constituents is relatively low andvaries from kava batch to batch Howeverthere are four relevant counterarguments

(1) The whole range of the constituents mayproduce a more effective and safe medicine(Williamson 2001)

(2) Some constituents not necessarily consid-ered active may enhance the safety of themedicine (Appendix 1)

DENHAM ET AL240

OCH3

H3COyangonin

(+)-methysticin

O O

OCH3

demethoxyangonin

O O

OCH3

OO

OH

O

(+)-dihydromethysticin

OCH3

OO

OH

O

(+)-kavain

OCH3

OH

O

(+)-dihydrokavain

OCH3

OH

O

FIG 1 Kavapyrones of kava (Piper methysticum) Per Haberlein et al 1997

(3) Definitive isolation of the active constituentis elusive in other medicinal plants such asHypericum perforatum (McIntyre 2000Barnes et al 2001)

(4) Herbal practitioners rely on the synergy be-tween a whole range of constituents in theherb or herb within a herbal prescriptionwhich is individually prescribed for a pa-tient This positive interaction may alsohave the benefit of keeping levels of anyone constituent below the safety threshold

LOW-ALCOHOL TINCTURES

The Traditional Medicines Evaluation Com-mittee (TMEC) strongly advocates the use ofextraction techniques that closely approximatethose traditionally used in Polynesia Thiswould require the use of low-alcohol tincturesmade by the traditional cold macerationprocesses common to UK tincture makingThe reasons for this opinion are set out belowthey have also been explored by Whitton et al(Appendix 1)

Tinctures used by herbal practitioners areprepared by macerating dried kava in a mix-ture of water and ethanol It has been shownthat such extracts using 25 ethanol75 wa-ter contain up to 30 times fewer kavalactonesthan the concentrated standardized prepara-tions (Appendix 1) The traditional preparationmethod using a mixture of 25 ethanol75water extracts a wider range of the naturalkava constituents (Appendix 1)

DOSAGE AND OVERDOSAGE

Assuming a 15 25 tincture and an upperlimit of 20 kavalactones in the dried herb(concentrations stated as 3ndash20 see sectionon Pharmacology below) then 500 mL of theherb would contain (100 3 02 3 015) 5 3 g(3000 mg) of kavalactones In addition assum-ing a daily dosage of 5ndash10 mL of the 15 25tincture the daily dose of kavalactonesamounts to a maximum of 30ndash60 mg If the con-centration of kavalactones were lower for ex-ample at approximately 10 as appears to bethe case with regard to Australian kava dis-

cussed by Clough et al (2000) then this dosagefalls to 15ndash30 mg per day It is noteworthy thatthe maximum daily dosage here is equivalentto the minimum daily dosages of the 60ndash210mg kavalactones given in clinical trials of kavaconducted in Germany

Although standardized extracts provide ahigher dosage of kavalactones than low-alco-hol tinctures overdosage in itself is unlikelyto be the cause of hepatotoxicity Strong evi-dence for this is the fact that kava is takendaily at high doses as a normal part of dailylife in large areas of the South Pacific Indeedsome of the accounts of high kava intake areremarkable For example Chanwai (2000) de-scribed the case of a man who was admittedto a hospital after an overdose but ldquoslept offrdquohis symptoms and admitted to consuming upto 40 bowls of a kava preparation per day forthe last 14 years In Australia missionaries in-troduced kava to the aborigines in the 1980sas a substitute for alcohol and it is claimed thatthis has led to abuse of kava Clough et al(2000) discussed this concern and reviewedtwelve studies on the amount of kava usedThe researchers found that social setting ap-pears to determine the amount used with lonedrinkers consuming much more than peoplewho enjoy kava in a family group The re-searchers described normal use of kava in theNorthern Territory as being 37 g of kava pow-der (containing approximately 3800 mg ofkavalactones) per hour with heavy consumersusing approximately 610 g per week preparedas a drink The incidence of serious illness re-sulting from hepatotoxicity associated withregular kava usage would surely have beenobserved by the medical services in Polynesiaand Australia if overdosage of kavalactoneswere the main cause of hepatotoxicity

UNTOWARD EFFECTS

There is a justified concern in Europe thatidiosyncratic hepatotoxicity associated with us-ing some herbal medicines may not be identi-fied because the population that takes herbalmedicines is not large enough to produce suf-ficient cases for the association to be noted Butthe fact that kava remains in traditional usageto such a wide extent is a powerful argument

KAVA WORK-IN-PROGRESS 241

that idiosyncratic hepatotoxicity would havebeen noted

Two postmarketing observation studies inGermany each on more than 3000 people werecited by Pittler and Ernst (2000) in addition tothe abovementioned clinical trials In these ob-servational studies the rate of adverse eventswas 23 (with a daily dose of 120ndash240 mg ofkavalactones) and 15 (with a daily dose of105 mg of kavalactones) The most frequent ad-verse reports were gastrointestinal complaintsallergic skin reactions headaches and photo-sensitivity

There is evidence in the South Pacific of a char-acteristic kava-induced skin disease a scaly rashthat is suggestive of icthyosismdasha condition calledldquokava dermopathyrdquo (Ruze 1990) Although theskin becomes yellow the description does notsuggest an underlying hepatic condition in thatthe patient remains well the rash is not itchyand the condition is ameliorated without treat-ment if heavy use of kava is reduced

The German and Swiss reports cited by theBfArM are of concern because there have beenprevious reports of hepatotoxicity associatedwith the use of some medicinal plants (Larrey1997) The kava case reports from the BfArM(see Appendix 2) include all three of the mainforms of acute damage that can result from ad-verse drug reactions (1) necrosis (2) drug-in-duced hepatitis and (3) cholestatic hepatitis(Hodgson and Levi 1997) This suggests thatthere is a range of causes rather than just onecause in these cases The BfArM case reportshave been circulated worldwide and are cur-rently being evaluated by government agenciesin Europe Australia Canada the UnitedStates and elsewhere We have received anumber of informal case assessments fromthese sources that cannot be specifically citedbecause of their confidential status To achievetransparency and encourage a full debate aboutkava however the BfArM cases are evaluatedin the section entitled Discussion of Cases Re-ported by the BfArM

CRITERIA FOR ASSESSING THE CASE REPORTS

A recent review of the information availableon the case reports (Schmidt and Nahrstedt

2002) is supported by details of the case re-ports on the Web site of the University ofMuenster (wwwuni-muensterdechemiepbkavaanalysehtml)

The criteria for causality assessment of ad-verse reactions used are as follows (Edwardsand Aronson 2000)Probable is defined as

A clinical event including a laboratory testabnormality that occurs in a plausible timerelation to drug administration and that can-not be explained by coincidental or concur-rent disease or other drugs or chemicals

The response to withdrawal of the drug(dechallange) should be clinically plausible

The event must be definitive pharmacolog-ically or phenomenologically using a satis-factory rechallenge procedure if necessary

Possible is defined as

A clinical event including a laboratory testabnormality with a reasonable time relationto administration of the drug but that couldbe explained by concurrent disease or otherdrugs or chemicals

Information on drug withdrawal may belacking or unclear

Unlikely is defined as

A clinical event including a laboratory testabnormality with a temporal relation to theadministration of the drug which makes acausal relation improbable and in whichother drugs chemicals or underlying dis-ease provide plausible explanations

Unassessable is defined as

A report suggesting an adverse reaction thatcannot be judged because information is in-sufficient or contradictory and cannot besupplemented or verified

DISCUSSION OF CASES REPORTED BY THE BfArM

The cases discussed below are analyzed andcategorized by common factors of note withour own assessments

DENHAM ET AL242

(1) Cases of most concern

This group includes five cases 10 13 16 18and 28 According to the assessments made by various government agencies these casescause the most concern and are often cited asbeing probable For this reason these cases aredealt with first As discussed below mostmdashifnot allmdashof these cases have associated factorsthat put this probable categorization into ques-tion

Case 10 This case described necrotizing hep-atitis in a 39-year old female patient with pos-itive reexposure (Strahl et al 1998) During thefirst period that the kava product was takenan oral contraceptive and paroxetine were con-comitant medications It appears that paroxe-tine was not the only antidepressant taken bythis patient who also occasionally took StJohnrsquos wort (Hypericum perforatum) The Exec-utive Summary issued by the Bundesverbandder Arzneimittel Hersteller eV (BAH) andBundesverband der Pharmazeutischen Indus-trie eV (BP) (see Appendix 3) stated ldquoA causalrelationship with kava cannot be excluded butthe patientrsquos history and a potential preexistingliver damage must be taken into account In ad-dition the kava preparation used was not iden-tified by the physicianrdquo

Moreover in this case taking paroxetine incombination with an oral contraceptive maywell have led to overburdening the liver a sit-uation that could have been exacerbated by tak-ing a kava preparation Schmidt and Nahrstedt(2002) suggested that this case may be associ-ated with an immunologic reaction After re-viewing all of the cases in detail Schmidt andNahrstedt (2002) concluded that this is the onlycase for which there was sufficient informationto make an association with kava appear prob-able and for which the dose of kava also con-formed to that recommended by the Com-mission E monograph (Blumenthal 2000)However as discussed below Russmann et al(2001) tested this patient for CYP2D6 and as inCase 16 found this patient to be CYP2D6 defi-cient which appears to have made her partic-ularly vulnerable to the cocktail of drugs shewas taking Given the complicating features ofthis case we submit that this case should beclassed as possible rather than probable

Case 13 This was a case of a 62-year-oldwoman with jaundice The BfArM table (seeAppendix 2) noted regarding concomitantmedication that there was ldquonone denotedrdquo butit was claimed that concomitant medication didexist but was ldquounknownrdquo The insufficiency ofdata provided for this case was highlighted byBfArMrsquos warning note ldquoNo medical messagerdquoIn addition it should be noted that no detailsof the dosage of kava or period of its adminis-tration were apparently recorded for this caseThis is clearly insufficient information onwhich to base a probable assessment

Case 16 This case concerned a 33-year-oldwoman with jaundice The woman wasrecorded as having taken an overdose of alco-hol measured at 60 g (Russman et al 2001) andthen analgesics including paracetamol fol-lowing this alcohol binge Despite the massiveintake of alcohol a liver biopsy indicated thata drug rather than an alcohol induced toxicgenesis However this case like that of Case 10above was discussed by Russman et al (2001)who demonstrated afterward that this patientwas shown to be CYP2D6 deficient which (asdiscussed below) seems to be a risk factor forthe hepatotoxicity that was ascribed to kavaWe submit that given these circumstances thiscase should be considered possible rather thanprobable

Case 18 This case concerned a 50-year-oldman who had necrosis leading to a liver trans-plant This patient took a product manufac-tured by acetone extraction at a dose deliver-ing 210ndash280 mg of kavalactones per day for 15months ldquomoderate alcoholrdquo (ldquomoderaterdquo is notdefined by BfArM) evening primrose(Oenothera biennis) and a yeast preparationThe dosage of kava was well above the Ger-man Commission E recommended dose ofkavalactones (Blumenthal 1998) It was alsorecorded that 500ndash1000 mg of paracetamol wastaken by this patient shortly before transplan-tation The combination of paracetamol and al-cohol plus the very high dose of kava extractedin acetone taken by this man casts seriousdoubts on the assessment of probable in thiscase

Case 28 (BAH) This case concerned a

KAVA WORK-IN-PROGRESS 243

woman age unknown with hepatitis This caseis hard to assess because neither the patientrsquosage nor diagnosis was given and the womanwas taking eleven medications including estra-diol valerate acetylcysteine losartan (which israrely be associated with hepatitis) and mepra-zole (which can be associated with liver diseasealthough this is rare) Omeprazole is metabo-lized by the polymorphic CYP2C19 which is absent in 3 of Caucasians (Flockhart et al2000) The woman was also taking echinacea(Echinacea purpurea) and five products that ap-peared to be for upper respiratory problems Itshould be noted that this patient was taking syn-thetic kavain not kava A comment from BfArMconcerning this case noted ldquorecurrence of the he-patic side-effectsrdquo which has evidently been in-terpreted by some authorities as being equiva-lent to a ldquopositive rechallengerdquo Whether or notthis was actually so was not clear from the datasupplied It appears (Schmidt and Nahrstedt2002) that Case 28 has been published as twocases with slightly different details This is con-fusing and considering that the woman was tak-ing 11 other medications together with a syn-thetic kava (which we submit is not equivalentto natural kava) and that no diagnosis of hercondition was supplied this calls the assessmentof probable in this case into question

(2) Cases associated with taking synthetic kavain

In this category there were 4 cases 1 2 19and 28

In each of these cases the patients concernedwere taking a product made from synthetickavain Although the outcome was hepatitis inall four cases kavain cannot be equated withthe naturally occurring form of kava whichcontains many other constituents that may playan important role in ensuring the safety of thisherb Therefore we submit that no inferenceshould be drawn from these cases Traditionalusage should not be taken as evidence for safeusage of synthetic products

(3) Patients who were taking oral contraceptivepills or hormone replacement therapy (HRT)together with drugs that can also be associatedwith liver damage

The cases in this category were 4 10 12 2021 and 28

Cholestatic jaundice associated with use ofestrogen-containing medications is extremelyrare (Lindberg 1992) but does occur In these6 cases the women were also taking drugs thatcan also be associated with jaundice

Case 4 This case involved a 39-year-oldwoman with jaundice She was on diazepam10 mg PRN for 6 months Some authoritiescalled this case possible Our assessment is thatthe case is unassessable

Case 10 This case involved a 39-year-oldwoman with necrotizing hepatitis For a de-tailed assessment see above

Case 12 This case involved a 37-year-oldwoman with hepatitis She was on 150 mg ofdiclofenac via intramuscular injection Hepato-toxic reactions associated with nonsteroidalanti-inflammatory drug use are extremely rareand concomitant exposure to other hepatotoxicdrugs is considered to be an important factor(Bareille et al 2001) This case of hepatitis isdifficult to interpret because it occurred inBrazil and because ldquoreexposure was said to benegative for all three drugsrdquo We regard thiscase as unassessable

Case 20 This case involved 50-year-oldwoman with necrosis who had a liver trans-plant She had a 20-year history of combinedoral contraceptive use but had changed monthsearlier to estradiol valerate (which was appar-ently taken alone) as HRT She had also startedglimepiride 8 months earlier This is used fortreating type II diabetes and is rarely associatedwith cholestatic jaundice and liver failure Weregard this case as unlikely

Case 21 This case involved a 22-year-oldwoman with necrosis who had a liver trans-plant This woman had changed from Valette(Jenapharm GmbH Jena Germany) (2 mg ofdienogest and 003 mg of ethinlestradiol) toPramino (180215250 mcg of norgestimateand mcg 25 of ethinylestradiol) She also tookrizatriptan if required for migraine reliefRizatriptan should be used with caution in he-patic impairment and avoided if a patient hassevere liver disease Some authorities considerthis case as being possible but our assessment

DENHAM ET AL244

is in view of the other medications taken isthat this case is unassessable

Case 28 This case involved a woman age un-known with hepatitis This case is discussed atlength above As noted above this patient wastaking synthetic kavain not kava

(4) Patients who were taking drugs that can beassociated with liver damage

There were ten cases in this category 1 6 914 15 17 19 23 2627 and 29

Case 1 This case involved a woman age 69with cholestatic hepatitis She was taking pen-toxifylline (which can be associated with intra-hepatic cholestasis) and a diuretic including thepotassium-sparing triamterene (which can beassociated with jaundice) As noted above thispatient was taking synthetic kavain not kavaWe consider this case unassessable

Case 6 This case involved a woman age 50with hepatitis She was taking frusemide(which can be associated with cholestatic jaun-dice) triamterene atenolol and a large dose ofterfenadine (300 mg) The recommended doseof terfenadine in the British National Formu-lary (March 2001) is 60ndash120 mg The Formularyrecommends avoiding this drug in patientswho have hepatic impairment and also says toldquoavoid concomitant administration of drugs li-able to produce electrolyte imbalance such asdiureticsrdquo (British National Formulary 2001)Despite this warning this woman was also tak-ing the diuretic frusemide The InterkantonalenKontrollstelle der Schweiz of Switzerland con-sidered this case of hepatitis to be caused byterfenadine And although some authoritiesregard this case as possible our assessment isthat this case is unlikely

Case 9 This case involved an 81-year-oldwoman who had liver failure and subsequentdeath She was taking hydrochlorothiazide(which can occasionally be associated with in-trahepatic cholestasis) However according toSchmidt and Nahrstedt (2002) there was evi-dence of chronic alcohol abuse and they re-ported that the autopsy showed chronic pan-creatitis that was characteristic of alcoholabuse The autopsy report (Schmidt and

Nahrstedt 2002) apparently said that thesymptoms must have occurred over a periodof at least 18 months The report conceded thatldquohepatic impairment by alcohol [was] not ex-cludedrdquo In these circumstances it seems en-tirely reasonable to claim that this case is un-related to kava use We regard this case asunlikely

Case 14 This case involved a 33-year-oldwoman with hepatitis Cisapride may havebeen taken (which can cause reversible changesthat show in liver-function tests) Cirrhosis ina woman of 33 is an unexplained finding andthe detail in this case is inadequate to elucidateit We consider this case to be unassessable

Case 15 This case involved a 46-year-oldwoman with jaundice She had been taking hy-drochlorothiazide (which can be associatedwith intrahepatic cholestasis) for 55 monthsplus 80 mg of valsartan and 80 mg ofpropanolol per day Some authorities regardthis case as possible but we consider it to beunassessable

Case 17 This case involved a 59-year-oldwoman with jaundice She had taken 100ndash200mg of celecoxib a cyclo-oxygenase-2 inhibitorper day According to the criteria for causalityassessment of adverse reactions some author-ities consider this case to be possible but our as-sessment is that it is unassessable

Case 19 This case involved a 21-year-oldwoman with hepatitis She was taking panto-prazole (which as with omeprazole can be as-sociated with liver disease) She was also takingparacetamol and metoclopramide and had over-dosed on kavain More detail is needed on othermedical conditions suffered by this patient in or-der to interpret this case It is suggested bySchmidt that this woman was using up to 10tablets per day of the product (the recom-mended dose is up to 6 tablets per day) and thatthere was apparently a discussion in her med-ical record file that she may also have used Ec-stasy (substance that has been associated with

KAVA WORK-IN-PROGRESS 245

Personal communication from M McGuffin to M McIn-tyre available as an online document at ehpaglobalnetcouk

fulminant hepatic failure) This case appears tobe unassessable

Case 23 This case involved a 35-year-oldwoman with jaundice According to the BfArM(see Appendix 2) this patient also took parac-etamol but no dosage or details were providedThis case and case 25 in the BfArM listing ap-pear to be the same case Both cases have beenlabeled as possible by some authorities butgiven the lack of information about the dosageof paracetamol and the apparent confusion re-garding cases 23 and 25 we submit that theonly logical assessment is unassessable

Case 2627 This case involved a woman whowas either 38 or 39 yearsrsquo old with hepatitis Itappears that the two cases have been duplicated(Schmidt and Nahrstedt 2002) The confusionwith this case is another example of inaccuratedata provided by the BfArM Information re-garding these cases (or case) depending onwhether the two reports concern the samewoman is unclear Penicillin can be associatedwith hypersensitivity and cholestatic jaundicebut the information given is inadequate to makeany meaningful assessment For this reason weclass this case as unassessable

Case 29 This case involved a 60-year-oldwoman who had a liver transplant This womanwas taking piretamide (which is a loop diuretic)Frusemide another loop diuretic can be associ-ated with cholestatic jaundice According to theBfArM chart (see Appendix 2) she was also tak-ing a sympathomimetic drug etilefrin Thedosage of kava varied but was up to 480ndash1200mg per day (Schmidt and Nahrstedt 2002)which is up to ten times the German Commis-sion E maximum recommended dose (Blumen-thal 1998) Although some authorities have re-garded this case as possible in view of themarked overdosing of kava and the concomitantmedication this case can hardly be said to be areflection on the proper therapeutic use of kava

(5) Cases in which drugs not associated withliver damage herbal medicines or dietarysupplements or kavain alone were taken

This category had eight cases 2 78 11 1322 24 and 25

For these cases detail was limited and theBfArM did not implicate any other drugs ormedications (although this may not be thecase)

All patients in this group apart from the pa-tient in Case 78 for whom no information wasgiven were reported to have made full recov-eries In some of these cases it is not clearwhether the patients were ill or whether thesecases merely recorded raised liver-function en-zymes

Case 2 This case involved a 35-year-old manwith cholestatic hepatitis Concomitant med-ication was ldquounknownrdquo Apart from Cases 18and 30 this is the only case for which it is pos-sible that no other concomitant medication wastaken but there is a marked lack of informationfor this case As noted above this patient wastaking synthetic kavain not kava We regardthis case as unassessable

Case 5 This case involved a woman who waseither 68 or 69 yearsrsquo old with cholestatic he-patitis She was also taking a St Johnrsquos wort(Hypericum perforatum) product which hasbeen associated with CYP3A4 A biopsyshowed ldquoimmunologic hypersensitivityrdquo Thiscase may be regarded as possible but in viewof the immunologic hypersensitivity it maywell have been an idiosyncratic event that wasnot necessarily associated with kava usage

Case 78 This case involved a woman or twowomen ages 72 andor 75 with cholestatic he-patitis These two cases appear to be actuallyone case The woman was taking twoherbalvitamin products one of which in-cluded 06 mg of kavalactones Given the con-fusion involved these ldquocasesrdquo must be re-garded as unassessable

Case 11 This case involved a 59-year-oldwoman who was taking hyoscine butylbro-mide as a suppository Schmidt and Nahrstedt(2002) commented that according to additionalinformation obtained from the BfArM it is un-certain as to whether this patient was taking akava product at all We regard this case asunassessable

DENHAM ET AL246

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

in Lebot et al 1997]) A comprehensive surveyby Lebot and Levesque in 1989 (cited in Lebotet al 1997) suggests that kava was originallydomesticated in the Vanuatu islands

Kava is now obtained from a wide range ofcultivars in the South Pacific The plant is al-ways propagated vegetatively from stem cut-tings as it does not reproduce via fertilizationmethods There are many cultivars and newstrains continue to be developed by Polynesianfarmers because each strain is considered tohave different psychoactive effects

In recent years there has been increasingpressure on the market because of the increasedworldwide demand for kava In 1998 it wasamong the top-selling herbs in the UnitedStates with a turnover of $8 million repre-senting a growth rate of 473 (Pittler and Ernst2000) It is possible that the current hepatotox-ity problems are to some extent a consequenceof poor quality control caused by a rapid andextraordinary increase in the size of the market(Murray 2000)

There are also concerns that intensive culti-vation and harvesting may affect the quality ofthe product (Aarlbersberg 1999) However be-ing that a range of products is implicated thisis unlikely to provide a satisfactory explanationfor all the reported adverse reactions It ishoped that the BfArM will release relevant dataabout the source and quality assurance of thekava supply in due course

KAVA PHARMACOLOGY

Kava is a well-researched herb The crys-talline resin was first isolated in 1857 by aFrench naval pharmacist and a detailed mono-graph was published in 1886 (Lewin 1886[cited by Lebot et al 1997]) The kavalactonesare considered to be the active constituents andhave been shown in animal studies to have asedative action (Hansel 1996) although themechanism is unclear (Spinella 2001) Thekavalactones are found in the resinous portion(5ndash9) of the plant material and are poorlysoluble in water

Kavalactones are 4-methoxy-2-pyrones withphenyl or styryl substituents at the 6th position(Lebot et al 1997) Total kavalactone content

varies from 3 to 20 dry weight Eighteen lac-tones have so far been isolated (He et al 1997)with the following six compounds (includingfour chiral enantiomers and two achiral enan-tiomers) being considered most important(Haberlain 1997 see Fig 1)

Chiral enantiomers(1) (1)-kavain(2) (1)-dihydrokavain(3) (1)-methysticin(4) (1)-dihydromethysticin

Achiral enantiomers(5) yangonin(6) demethoxyangonin

TRADITIONAL PREPARATION TECHNIQUES

Kava is traditionally prepared in the SouthPacific by grinding and mixing the root or rootbark with cold water This makes an emulsionthat is a suspension of the resinous constituentsin water (Lebot et al 1997) The herb is alsoprepared as an emulsion (also traditionallyprepared without heating) in coconut milk(Johnson 1999) The efficiency of extraction ofthe active constituents which is measured bykavalactone extraction into water varies con-siderably (Murray 2000) but is higher fromfresh material than from the dried plant Kavaconsumed in Vanuatu is reputed to be thestrongest anywhere in the South Pacific The is-lands of Tanna and Pentecost are especiallynoted for their potent brews It is thought thatpart of this increase in potency is the result ofpreparing the drink from the raw fresh rootswhereas in Fiji and elsewhere it is made fromdried rootstock (Greenwood-Robinson 1999)

MODERN PREPARATION TECHNIQUES

The bioavailability of kava constituentsvaries substantially depending on the methodof extraction (Hansel et al 1994 [cited in Schulzet al 1997] Loew 2002) Kava is predomi-nantly available in Germany as a so-called con-centrated standardized extract that is designedto maximize extraction of the kavalactones

KAVA WORK-IN-PROGRESS 239

Schulz noted that to create these concentratedstandardized extracts kava is dissolved in ahigh percentage of an ethanolndashwater mixtureto obtain extracts containing approximately30 kavalactones or alternatively using anacetonendashwater mixture to obtain extracts con-taining approximately 70 kavalactones Whit-ton Whitehouse and Evans (Appendix 1)make the same point about enhanced kavalac-tone extraction using a high ethanol or acetonemedium but detail somewhat different extrac-tion values Both types of products have a herb-to-extract ratio of approximately 12ndash201(Schulz 1997) The dosage recommended by theGerman Commission E is expressed as theequivalent of 60ndash120 mg of kavalactones per day(Blumenthal 1998)

The preparation methods used for stan-dardized products are highly technical and extraction rates vary (Kubatova et al 2001)depending on the solvents used and the tem-perature at which the products are preparedAs Whitton et al propose (Appendix 1) bothefficacy and safety may depend on thekavalactones remaining in their natural formsand on the extraction of the other natural con-stituents of the plant

Varying extraction techniques and preparationmethods may result in an unnatural variation inthe relative concentration of each lactone or inproduction artifacts that may be pathologic to theliver It should be noted that some commercialkava products may also contain syntheticracemic kavain that may have other characteris-tics than the naturally occurring product has Itis clear that these technical matters related to ex-traction techniques require further elucidation

Proponents of concentrated standardizedproducts assert that they provide an effectivedose within a consistent range The traditionalwater-based kava preparations of the Polyne-sian peoples and low-alcohol kava tincturesused by herbal practitioners have been consid-ered to be unreliable because the concentrationof active constituents is relatively low andvaries from kava batch to batch Howeverthere are four relevant counterarguments

(1) The whole range of the constituents mayproduce a more effective and safe medicine(Williamson 2001)

(2) Some constituents not necessarily consid-ered active may enhance the safety of themedicine (Appendix 1)

DENHAM ET AL240

OCH3

H3COyangonin

(+)-methysticin

O O

OCH3

demethoxyangonin

O O

OCH3

OO

OH

O

(+)-dihydromethysticin

OCH3

OO

OH

O

(+)-kavain

OCH3

OH

O

(+)-dihydrokavain

OCH3

OH

O

FIG 1 Kavapyrones of kava (Piper methysticum) Per Haberlein et al 1997

(3) Definitive isolation of the active constituentis elusive in other medicinal plants such asHypericum perforatum (McIntyre 2000Barnes et al 2001)

(4) Herbal practitioners rely on the synergy be-tween a whole range of constituents in theherb or herb within a herbal prescriptionwhich is individually prescribed for a pa-tient This positive interaction may alsohave the benefit of keeping levels of anyone constituent below the safety threshold

LOW-ALCOHOL TINCTURES

The Traditional Medicines Evaluation Com-mittee (TMEC) strongly advocates the use ofextraction techniques that closely approximatethose traditionally used in Polynesia Thiswould require the use of low-alcohol tincturesmade by the traditional cold macerationprocesses common to UK tincture makingThe reasons for this opinion are set out belowthey have also been explored by Whitton et al(Appendix 1)

Tinctures used by herbal practitioners areprepared by macerating dried kava in a mix-ture of water and ethanol It has been shownthat such extracts using 25 ethanol75 wa-ter contain up to 30 times fewer kavalactonesthan the concentrated standardized prepara-tions (Appendix 1) The traditional preparationmethod using a mixture of 25 ethanol75water extracts a wider range of the naturalkava constituents (Appendix 1)

DOSAGE AND OVERDOSAGE

Assuming a 15 25 tincture and an upperlimit of 20 kavalactones in the dried herb(concentrations stated as 3ndash20 see sectionon Pharmacology below) then 500 mL of theherb would contain (100 3 02 3 015) 5 3 g(3000 mg) of kavalactones In addition assum-ing a daily dosage of 5ndash10 mL of the 15 25tincture the daily dose of kavalactonesamounts to a maximum of 30ndash60 mg If the con-centration of kavalactones were lower for ex-ample at approximately 10 as appears to bethe case with regard to Australian kava dis-

cussed by Clough et al (2000) then this dosagefalls to 15ndash30 mg per day It is noteworthy thatthe maximum daily dosage here is equivalentto the minimum daily dosages of the 60ndash210mg kavalactones given in clinical trials of kavaconducted in Germany

Although standardized extracts provide ahigher dosage of kavalactones than low-alco-hol tinctures overdosage in itself is unlikelyto be the cause of hepatotoxicity Strong evi-dence for this is the fact that kava is takendaily at high doses as a normal part of dailylife in large areas of the South Pacific Indeedsome of the accounts of high kava intake areremarkable For example Chanwai (2000) de-scribed the case of a man who was admittedto a hospital after an overdose but ldquoslept offrdquohis symptoms and admitted to consuming upto 40 bowls of a kava preparation per day forthe last 14 years In Australia missionaries in-troduced kava to the aborigines in the 1980sas a substitute for alcohol and it is claimed thatthis has led to abuse of kava Clough et al(2000) discussed this concern and reviewedtwelve studies on the amount of kava usedThe researchers found that social setting ap-pears to determine the amount used with lonedrinkers consuming much more than peoplewho enjoy kava in a family group The re-searchers described normal use of kava in theNorthern Territory as being 37 g of kava pow-der (containing approximately 3800 mg ofkavalactones) per hour with heavy consumersusing approximately 610 g per week preparedas a drink The incidence of serious illness re-sulting from hepatotoxicity associated withregular kava usage would surely have beenobserved by the medical services in Polynesiaand Australia if overdosage of kavalactoneswere the main cause of hepatotoxicity

UNTOWARD EFFECTS

There is a justified concern in Europe thatidiosyncratic hepatotoxicity associated with us-ing some herbal medicines may not be identi-fied because the population that takes herbalmedicines is not large enough to produce suf-ficient cases for the association to be noted Butthe fact that kava remains in traditional usageto such a wide extent is a powerful argument

KAVA WORK-IN-PROGRESS 241

that idiosyncratic hepatotoxicity would havebeen noted

Two postmarketing observation studies inGermany each on more than 3000 people werecited by Pittler and Ernst (2000) in addition tothe abovementioned clinical trials In these ob-servational studies the rate of adverse eventswas 23 (with a daily dose of 120ndash240 mg ofkavalactones) and 15 (with a daily dose of105 mg of kavalactones) The most frequent ad-verse reports were gastrointestinal complaintsallergic skin reactions headaches and photo-sensitivity

There is evidence in the South Pacific of a char-acteristic kava-induced skin disease a scaly rashthat is suggestive of icthyosismdasha condition calledldquokava dermopathyrdquo (Ruze 1990) Although theskin becomes yellow the description does notsuggest an underlying hepatic condition in thatthe patient remains well the rash is not itchyand the condition is ameliorated without treat-ment if heavy use of kava is reduced

The German and Swiss reports cited by theBfArM are of concern because there have beenprevious reports of hepatotoxicity associatedwith the use of some medicinal plants (Larrey1997) The kava case reports from the BfArM(see Appendix 2) include all three of the mainforms of acute damage that can result from ad-verse drug reactions (1) necrosis (2) drug-in-duced hepatitis and (3) cholestatic hepatitis(Hodgson and Levi 1997) This suggests thatthere is a range of causes rather than just onecause in these cases The BfArM case reportshave been circulated worldwide and are cur-rently being evaluated by government agenciesin Europe Australia Canada the UnitedStates and elsewhere We have received anumber of informal case assessments fromthese sources that cannot be specifically citedbecause of their confidential status To achievetransparency and encourage a full debate aboutkava however the BfArM cases are evaluatedin the section entitled Discussion of Cases Re-ported by the BfArM

CRITERIA FOR ASSESSING THE CASE REPORTS

A recent review of the information availableon the case reports (Schmidt and Nahrstedt

2002) is supported by details of the case re-ports on the Web site of the University ofMuenster (wwwuni-muensterdechemiepbkavaanalysehtml)

The criteria for causality assessment of ad-verse reactions used are as follows (Edwardsand Aronson 2000)Probable is defined as

A clinical event including a laboratory testabnormality that occurs in a plausible timerelation to drug administration and that can-not be explained by coincidental or concur-rent disease or other drugs or chemicals

The response to withdrawal of the drug(dechallange) should be clinically plausible

The event must be definitive pharmacolog-ically or phenomenologically using a satis-factory rechallenge procedure if necessary

Possible is defined as

A clinical event including a laboratory testabnormality with a reasonable time relationto administration of the drug but that couldbe explained by concurrent disease or otherdrugs or chemicals

Information on drug withdrawal may belacking or unclear

Unlikely is defined as

A clinical event including a laboratory testabnormality with a temporal relation to theadministration of the drug which makes acausal relation improbable and in whichother drugs chemicals or underlying dis-ease provide plausible explanations

Unassessable is defined as

A report suggesting an adverse reaction thatcannot be judged because information is in-sufficient or contradictory and cannot besupplemented or verified

DISCUSSION OF CASES REPORTED BY THE BfArM

The cases discussed below are analyzed andcategorized by common factors of note withour own assessments

DENHAM ET AL242

(1) Cases of most concern

This group includes five cases 10 13 16 18and 28 According to the assessments made by various government agencies these casescause the most concern and are often cited asbeing probable For this reason these cases aredealt with first As discussed below mostmdashifnot allmdashof these cases have associated factorsthat put this probable categorization into ques-tion

Case 10 This case described necrotizing hep-atitis in a 39-year old female patient with pos-itive reexposure (Strahl et al 1998) During thefirst period that the kava product was takenan oral contraceptive and paroxetine were con-comitant medications It appears that paroxe-tine was not the only antidepressant taken bythis patient who also occasionally took StJohnrsquos wort (Hypericum perforatum) The Exec-utive Summary issued by the Bundesverbandder Arzneimittel Hersteller eV (BAH) andBundesverband der Pharmazeutischen Indus-trie eV (BP) (see Appendix 3) stated ldquoA causalrelationship with kava cannot be excluded butthe patientrsquos history and a potential preexistingliver damage must be taken into account In ad-dition the kava preparation used was not iden-tified by the physicianrdquo

Moreover in this case taking paroxetine incombination with an oral contraceptive maywell have led to overburdening the liver a sit-uation that could have been exacerbated by tak-ing a kava preparation Schmidt and Nahrstedt(2002) suggested that this case may be associ-ated with an immunologic reaction After re-viewing all of the cases in detail Schmidt andNahrstedt (2002) concluded that this is the onlycase for which there was sufficient informationto make an association with kava appear prob-able and for which the dose of kava also con-formed to that recommended by the Com-mission E monograph (Blumenthal 2000)However as discussed below Russmann et al(2001) tested this patient for CYP2D6 and as inCase 16 found this patient to be CYP2D6 defi-cient which appears to have made her partic-ularly vulnerable to the cocktail of drugs shewas taking Given the complicating features ofthis case we submit that this case should beclassed as possible rather than probable

Case 13 This was a case of a 62-year-oldwoman with jaundice The BfArM table (seeAppendix 2) noted regarding concomitantmedication that there was ldquonone denotedrdquo butit was claimed that concomitant medication didexist but was ldquounknownrdquo The insufficiency ofdata provided for this case was highlighted byBfArMrsquos warning note ldquoNo medical messagerdquoIn addition it should be noted that no detailsof the dosage of kava or period of its adminis-tration were apparently recorded for this caseThis is clearly insufficient information onwhich to base a probable assessment

Case 16 This case concerned a 33-year-oldwoman with jaundice The woman wasrecorded as having taken an overdose of alco-hol measured at 60 g (Russman et al 2001) andthen analgesics including paracetamol fol-lowing this alcohol binge Despite the massiveintake of alcohol a liver biopsy indicated thata drug rather than an alcohol induced toxicgenesis However this case like that of Case 10above was discussed by Russman et al (2001)who demonstrated afterward that this patientwas shown to be CYP2D6 deficient which (asdiscussed below) seems to be a risk factor forthe hepatotoxicity that was ascribed to kavaWe submit that given these circumstances thiscase should be considered possible rather thanprobable

Case 18 This case concerned a 50-year-oldman who had necrosis leading to a liver trans-plant This patient took a product manufac-tured by acetone extraction at a dose deliver-ing 210ndash280 mg of kavalactones per day for 15months ldquomoderate alcoholrdquo (ldquomoderaterdquo is notdefined by BfArM) evening primrose(Oenothera biennis) and a yeast preparationThe dosage of kava was well above the Ger-man Commission E recommended dose ofkavalactones (Blumenthal 1998) It was alsorecorded that 500ndash1000 mg of paracetamol wastaken by this patient shortly before transplan-tation The combination of paracetamol and al-cohol plus the very high dose of kava extractedin acetone taken by this man casts seriousdoubts on the assessment of probable in thiscase

Case 28 (BAH) This case concerned a

KAVA WORK-IN-PROGRESS 243

woman age unknown with hepatitis This caseis hard to assess because neither the patientrsquosage nor diagnosis was given and the womanwas taking eleven medications including estra-diol valerate acetylcysteine losartan (which israrely be associated with hepatitis) and mepra-zole (which can be associated with liver diseasealthough this is rare) Omeprazole is metabo-lized by the polymorphic CYP2C19 which is absent in 3 of Caucasians (Flockhart et al2000) The woman was also taking echinacea(Echinacea purpurea) and five products that ap-peared to be for upper respiratory problems Itshould be noted that this patient was taking syn-thetic kavain not kava A comment from BfArMconcerning this case noted ldquorecurrence of the he-patic side-effectsrdquo which has evidently been in-terpreted by some authorities as being equiva-lent to a ldquopositive rechallengerdquo Whether or notthis was actually so was not clear from the datasupplied It appears (Schmidt and Nahrstedt2002) that Case 28 has been published as twocases with slightly different details This is con-fusing and considering that the woman was tak-ing 11 other medications together with a syn-thetic kava (which we submit is not equivalentto natural kava) and that no diagnosis of hercondition was supplied this calls the assessmentof probable in this case into question

(2) Cases associated with taking synthetic kavain

In this category there were 4 cases 1 2 19and 28

In each of these cases the patients concernedwere taking a product made from synthetickavain Although the outcome was hepatitis inall four cases kavain cannot be equated withthe naturally occurring form of kava whichcontains many other constituents that may playan important role in ensuring the safety of thisherb Therefore we submit that no inferenceshould be drawn from these cases Traditionalusage should not be taken as evidence for safeusage of synthetic products

(3) Patients who were taking oral contraceptivepills or hormone replacement therapy (HRT)together with drugs that can also be associatedwith liver damage

The cases in this category were 4 10 12 2021 and 28

Cholestatic jaundice associated with use ofestrogen-containing medications is extremelyrare (Lindberg 1992) but does occur In these6 cases the women were also taking drugs thatcan also be associated with jaundice

Case 4 This case involved a 39-year-oldwoman with jaundice She was on diazepam10 mg PRN for 6 months Some authoritiescalled this case possible Our assessment is thatthe case is unassessable

Case 10 This case involved a 39-year-oldwoman with necrotizing hepatitis For a de-tailed assessment see above

Case 12 This case involved a 37-year-oldwoman with hepatitis She was on 150 mg ofdiclofenac via intramuscular injection Hepato-toxic reactions associated with nonsteroidalanti-inflammatory drug use are extremely rareand concomitant exposure to other hepatotoxicdrugs is considered to be an important factor(Bareille et al 2001) This case of hepatitis isdifficult to interpret because it occurred inBrazil and because ldquoreexposure was said to benegative for all three drugsrdquo We regard thiscase as unassessable

Case 20 This case involved 50-year-oldwoman with necrosis who had a liver trans-plant She had a 20-year history of combinedoral contraceptive use but had changed monthsearlier to estradiol valerate (which was appar-ently taken alone) as HRT She had also startedglimepiride 8 months earlier This is used fortreating type II diabetes and is rarely associatedwith cholestatic jaundice and liver failure Weregard this case as unlikely

Case 21 This case involved a 22-year-oldwoman with necrosis who had a liver trans-plant This woman had changed from Valette(Jenapharm GmbH Jena Germany) (2 mg ofdienogest and 003 mg of ethinlestradiol) toPramino (180215250 mcg of norgestimateand mcg 25 of ethinylestradiol) She also tookrizatriptan if required for migraine reliefRizatriptan should be used with caution in he-patic impairment and avoided if a patient hassevere liver disease Some authorities considerthis case as being possible but our assessment

DENHAM ET AL244

is in view of the other medications taken isthat this case is unassessable

Case 28 This case involved a woman age un-known with hepatitis This case is discussed atlength above As noted above this patient wastaking synthetic kavain not kava

(4) Patients who were taking drugs that can beassociated with liver damage

There were ten cases in this category 1 6 914 15 17 19 23 2627 and 29

Case 1 This case involved a woman age 69with cholestatic hepatitis She was taking pen-toxifylline (which can be associated with intra-hepatic cholestasis) and a diuretic including thepotassium-sparing triamterene (which can beassociated with jaundice) As noted above thispatient was taking synthetic kavain not kavaWe consider this case unassessable

Case 6 This case involved a woman age 50with hepatitis She was taking frusemide(which can be associated with cholestatic jaun-dice) triamterene atenolol and a large dose ofterfenadine (300 mg) The recommended doseof terfenadine in the British National Formu-lary (March 2001) is 60ndash120 mg The Formularyrecommends avoiding this drug in patientswho have hepatic impairment and also says toldquoavoid concomitant administration of drugs li-able to produce electrolyte imbalance such asdiureticsrdquo (British National Formulary 2001)Despite this warning this woman was also tak-ing the diuretic frusemide The InterkantonalenKontrollstelle der Schweiz of Switzerland con-sidered this case of hepatitis to be caused byterfenadine And although some authoritiesregard this case as possible our assessment isthat this case is unlikely

Case 9 This case involved an 81-year-oldwoman who had liver failure and subsequentdeath She was taking hydrochlorothiazide(which can occasionally be associated with in-trahepatic cholestasis) However according toSchmidt and Nahrstedt (2002) there was evi-dence of chronic alcohol abuse and they re-ported that the autopsy showed chronic pan-creatitis that was characteristic of alcoholabuse The autopsy report (Schmidt and

Nahrstedt 2002) apparently said that thesymptoms must have occurred over a periodof at least 18 months The report conceded thatldquohepatic impairment by alcohol [was] not ex-cludedrdquo In these circumstances it seems en-tirely reasonable to claim that this case is un-related to kava use We regard this case asunlikely

Case 14 This case involved a 33-year-oldwoman with hepatitis Cisapride may havebeen taken (which can cause reversible changesthat show in liver-function tests) Cirrhosis ina woman of 33 is an unexplained finding andthe detail in this case is inadequate to elucidateit We consider this case to be unassessable

Case 15 This case involved a 46-year-oldwoman with jaundice She had been taking hy-drochlorothiazide (which can be associatedwith intrahepatic cholestasis) for 55 monthsplus 80 mg of valsartan and 80 mg ofpropanolol per day Some authorities regardthis case as possible but we consider it to beunassessable

Case 17 This case involved a 59-year-oldwoman with jaundice She had taken 100ndash200mg of celecoxib a cyclo-oxygenase-2 inhibitorper day According to the criteria for causalityassessment of adverse reactions some author-ities consider this case to be possible but our as-sessment is that it is unassessable

Case 19 This case involved a 21-year-oldwoman with hepatitis She was taking panto-prazole (which as with omeprazole can be as-sociated with liver disease) She was also takingparacetamol and metoclopramide and had over-dosed on kavain More detail is needed on othermedical conditions suffered by this patient in or-der to interpret this case It is suggested bySchmidt that this woman was using up to 10tablets per day of the product (the recom-mended dose is up to 6 tablets per day) and thatthere was apparently a discussion in her med-ical record file that she may also have used Ec-stasy (substance that has been associated with

KAVA WORK-IN-PROGRESS 245

Personal communication from M McGuffin to M McIn-tyre available as an online document at ehpaglobalnetcouk

fulminant hepatic failure) This case appears tobe unassessable

Case 23 This case involved a 35-year-oldwoman with jaundice According to the BfArM(see Appendix 2) this patient also took parac-etamol but no dosage or details were providedThis case and case 25 in the BfArM listing ap-pear to be the same case Both cases have beenlabeled as possible by some authorities butgiven the lack of information about the dosageof paracetamol and the apparent confusion re-garding cases 23 and 25 we submit that theonly logical assessment is unassessable

Case 2627 This case involved a woman whowas either 38 or 39 yearsrsquo old with hepatitis Itappears that the two cases have been duplicated(Schmidt and Nahrstedt 2002) The confusionwith this case is another example of inaccuratedata provided by the BfArM Information re-garding these cases (or case) depending onwhether the two reports concern the samewoman is unclear Penicillin can be associatedwith hypersensitivity and cholestatic jaundicebut the information given is inadequate to makeany meaningful assessment For this reason weclass this case as unassessable

Case 29 This case involved a 60-year-oldwoman who had a liver transplant This womanwas taking piretamide (which is a loop diuretic)Frusemide another loop diuretic can be associ-ated with cholestatic jaundice According to theBfArM chart (see Appendix 2) she was also tak-ing a sympathomimetic drug etilefrin Thedosage of kava varied but was up to 480ndash1200mg per day (Schmidt and Nahrstedt 2002)which is up to ten times the German Commis-sion E maximum recommended dose (Blumen-thal 1998) Although some authorities have re-garded this case as possible in view of themarked overdosing of kava and the concomitantmedication this case can hardly be said to be areflection on the proper therapeutic use of kava

(5) Cases in which drugs not associated withliver damage herbal medicines or dietarysupplements or kavain alone were taken

This category had eight cases 2 78 11 1322 24 and 25

For these cases detail was limited and theBfArM did not implicate any other drugs ormedications (although this may not be thecase)

All patients in this group apart from the pa-tient in Case 78 for whom no information wasgiven were reported to have made full recov-eries In some of these cases it is not clearwhether the patients were ill or whether thesecases merely recorded raised liver-function en-zymes

Case 2 This case involved a 35-year-old manwith cholestatic hepatitis Concomitant med-ication was ldquounknownrdquo Apart from Cases 18and 30 this is the only case for which it is pos-sible that no other concomitant medication wastaken but there is a marked lack of informationfor this case As noted above this patient wastaking synthetic kavain not kava We regardthis case as unassessable

Case 5 This case involved a woman who waseither 68 or 69 yearsrsquo old with cholestatic he-patitis She was also taking a St Johnrsquos wort(Hypericum perforatum) product which hasbeen associated with CYP3A4 A biopsyshowed ldquoimmunologic hypersensitivityrdquo Thiscase may be regarded as possible but in viewof the immunologic hypersensitivity it maywell have been an idiosyncratic event that wasnot necessarily associated with kava usage

Case 78 This case involved a woman or twowomen ages 72 andor 75 with cholestatic he-patitis These two cases appear to be actuallyone case The woman was taking twoherbalvitamin products one of which in-cluded 06 mg of kavalactones Given the con-fusion involved these ldquocasesrdquo must be re-garded as unassessable

Case 11 This case involved a 59-year-oldwoman who was taking hyoscine butylbro-mide as a suppository Schmidt and Nahrstedt(2002) commented that according to additionalinformation obtained from the BfArM it is un-certain as to whether this patient was taking akava product at all We regard this case asunassessable

DENHAM ET AL246

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

Schulz noted that to create these concentratedstandardized extracts kava is dissolved in ahigh percentage of an ethanolndashwater mixtureto obtain extracts containing approximately30 kavalactones or alternatively using anacetonendashwater mixture to obtain extracts con-taining approximately 70 kavalactones Whit-ton Whitehouse and Evans (Appendix 1)make the same point about enhanced kavalac-tone extraction using a high ethanol or acetonemedium but detail somewhat different extrac-tion values Both types of products have a herb-to-extract ratio of approximately 12ndash201(Schulz 1997) The dosage recommended by theGerman Commission E is expressed as theequivalent of 60ndash120 mg of kavalactones per day(Blumenthal 1998)

The preparation methods used for stan-dardized products are highly technical and extraction rates vary (Kubatova et al 2001)depending on the solvents used and the tem-perature at which the products are preparedAs Whitton et al propose (Appendix 1) bothefficacy and safety may depend on thekavalactones remaining in their natural formsand on the extraction of the other natural con-stituents of the plant

Varying extraction techniques and preparationmethods may result in an unnatural variation inthe relative concentration of each lactone or inproduction artifacts that may be pathologic to theliver It should be noted that some commercialkava products may also contain syntheticracemic kavain that may have other characteris-tics than the naturally occurring product has Itis clear that these technical matters related to ex-traction techniques require further elucidation

Proponents of concentrated standardizedproducts assert that they provide an effectivedose within a consistent range The traditionalwater-based kava preparations of the Polyne-sian peoples and low-alcohol kava tincturesused by herbal practitioners have been consid-ered to be unreliable because the concentrationof active constituents is relatively low andvaries from kava batch to batch Howeverthere are four relevant counterarguments

(1) The whole range of the constituents mayproduce a more effective and safe medicine(Williamson 2001)

(2) Some constituents not necessarily consid-ered active may enhance the safety of themedicine (Appendix 1)

DENHAM ET AL240

OCH3

H3COyangonin

(+)-methysticin

O O

OCH3

demethoxyangonin

O O

OCH3

OO

OH

O

(+)-dihydromethysticin

OCH3

OO

OH

O

(+)-kavain

OCH3

OH

O

(+)-dihydrokavain

OCH3

OH

O

FIG 1 Kavapyrones of kava (Piper methysticum) Per Haberlein et al 1997

(3) Definitive isolation of the active constituentis elusive in other medicinal plants such asHypericum perforatum (McIntyre 2000Barnes et al 2001)

(4) Herbal practitioners rely on the synergy be-tween a whole range of constituents in theherb or herb within a herbal prescriptionwhich is individually prescribed for a pa-tient This positive interaction may alsohave the benefit of keeping levels of anyone constituent below the safety threshold

LOW-ALCOHOL TINCTURES

The Traditional Medicines Evaluation Com-mittee (TMEC) strongly advocates the use ofextraction techniques that closely approximatethose traditionally used in Polynesia Thiswould require the use of low-alcohol tincturesmade by the traditional cold macerationprocesses common to UK tincture makingThe reasons for this opinion are set out belowthey have also been explored by Whitton et al(Appendix 1)

Tinctures used by herbal practitioners areprepared by macerating dried kava in a mix-ture of water and ethanol It has been shownthat such extracts using 25 ethanol75 wa-ter contain up to 30 times fewer kavalactonesthan the concentrated standardized prepara-tions (Appendix 1) The traditional preparationmethod using a mixture of 25 ethanol75water extracts a wider range of the naturalkava constituents (Appendix 1)

DOSAGE AND OVERDOSAGE

Assuming a 15 25 tincture and an upperlimit of 20 kavalactones in the dried herb(concentrations stated as 3ndash20 see sectionon Pharmacology below) then 500 mL of theherb would contain (100 3 02 3 015) 5 3 g(3000 mg) of kavalactones In addition assum-ing a daily dosage of 5ndash10 mL of the 15 25tincture the daily dose of kavalactonesamounts to a maximum of 30ndash60 mg If the con-centration of kavalactones were lower for ex-ample at approximately 10 as appears to bethe case with regard to Australian kava dis-

cussed by Clough et al (2000) then this dosagefalls to 15ndash30 mg per day It is noteworthy thatthe maximum daily dosage here is equivalentto the minimum daily dosages of the 60ndash210mg kavalactones given in clinical trials of kavaconducted in Germany

Although standardized extracts provide ahigher dosage of kavalactones than low-alco-hol tinctures overdosage in itself is unlikelyto be the cause of hepatotoxicity Strong evi-dence for this is the fact that kava is takendaily at high doses as a normal part of dailylife in large areas of the South Pacific Indeedsome of the accounts of high kava intake areremarkable For example Chanwai (2000) de-scribed the case of a man who was admittedto a hospital after an overdose but ldquoslept offrdquohis symptoms and admitted to consuming upto 40 bowls of a kava preparation per day forthe last 14 years In Australia missionaries in-troduced kava to the aborigines in the 1980sas a substitute for alcohol and it is claimed thatthis has led to abuse of kava Clough et al(2000) discussed this concern and reviewedtwelve studies on the amount of kava usedThe researchers found that social setting ap-pears to determine the amount used with lonedrinkers consuming much more than peoplewho enjoy kava in a family group The re-searchers described normal use of kava in theNorthern Territory as being 37 g of kava pow-der (containing approximately 3800 mg ofkavalactones) per hour with heavy consumersusing approximately 610 g per week preparedas a drink The incidence of serious illness re-sulting from hepatotoxicity associated withregular kava usage would surely have beenobserved by the medical services in Polynesiaand Australia if overdosage of kavalactoneswere the main cause of hepatotoxicity

UNTOWARD EFFECTS

There is a justified concern in Europe thatidiosyncratic hepatotoxicity associated with us-ing some herbal medicines may not be identi-fied because the population that takes herbalmedicines is not large enough to produce suf-ficient cases for the association to be noted Butthe fact that kava remains in traditional usageto such a wide extent is a powerful argument

KAVA WORK-IN-PROGRESS 241

that idiosyncratic hepatotoxicity would havebeen noted

Two postmarketing observation studies inGermany each on more than 3000 people werecited by Pittler and Ernst (2000) in addition tothe abovementioned clinical trials In these ob-servational studies the rate of adverse eventswas 23 (with a daily dose of 120ndash240 mg ofkavalactones) and 15 (with a daily dose of105 mg of kavalactones) The most frequent ad-verse reports were gastrointestinal complaintsallergic skin reactions headaches and photo-sensitivity

There is evidence in the South Pacific of a char-acteristic kava-induced skin disease a scaly rashthat is suggestive of icthyosismdasha condition calledldquokava dermopathyrdquo (Ruze 1990) Although theskin becomes yellow the description does notsuggest an underlying hepatic condition in thatthe patient remains well the rash is not itchyand the condition is ameliorated without treat-ment if heavy use of kava is reduced

The German and Swiss reports cited by theBfArM are of concern because there have beenprevious reports of hepatotoxicity associatedwith the use of some medicinal plants (Larrey1997) The kava case reports from the BfArM(see Appendix 2) include all three of the mainforms of acute damage that can result from ad-verse drug reactions (1) necrosis (2) drug-in-duced hepatitis and (3) cholestatic hepatitis(Hodgson and Levi 1997) This suggests thatthere is a range of causes rather than just onecause in these cases The BfArM case reportshave been circulated worldwide and are cur-rently being evaluated by government agenciesin Europe Australia Canada the UnitedStates and elsewhere We have received anumber of informal case assessments fromthese sources that cannot be specifically citedbecause of their confidential status To achievetransparency and encourage a full debate aboutkava however the BfArM cases are evaluatedin the section entitled Discussion of Cases Re-ported by the BfArM

CRITERIA FOR ASSESSING THE CASE REPORTS

A recent review of the information availableon the case reports (Schmidt and Nahrstedt

2002) is supported by details of the case re-ports on the Web site of the University ofMuenster (wwwuni-muensterdechemiepbkavaanalysehtml)

The criteria for causality assessment of ad-verse reactions used are as follows (Edwardsand Aronson 2000)Probable is defined as

A clinical event including a laboratory testabnormality that occurs in a plausible timerelation to drug administration and that can-not be explained by coincidental or concur-rent disease or other drugs or chemicals

The response to withdrawal of the drug(dechallange) should be clinically plausible

The event must be definitive pharmacolog-ically or phenomenologically using a satis-factory rechallenge procedure if necessary

Possible is defined as

A clinical event including a laboratory testabnormality with a reasonable time relationto administration of the drug but that couldbe explained by concurrent disease or otherdrugs or chemicals

Information on drug withdrawal may belacking or unclear

Unlikely is defined as

A clinical event including a laboratory testabnormality with a temporal relation to theadministration of the drug which makes acausal relation improbable and in whichother drugs chemicals or underlying dis-ease provide plausible explanations

Unassessable is defined as

A report suggesting an adverse reaction thatcannot be judged because information is in-sufficient or contradictory and cannot besupplemented or verified

DISCUSSION OF CASES REPORTED BY THE BfArM

The cases discussed below are analyzed andcategorized by common factors of note withour own assessments

DENHAM ET AL242

(1) Cases of most concern

This group includes five cases 10 13 16 18and 28 According to the assessments made by various government agencies these casescause the most concern and are often cited asbeing probable For this reason these cases aredealt with first As discussed below mostmdashifnot allmdashof these cases have associated factorsthat put this probable categorization into ques-tion

Case 10 This case described necrotizing hep-atitis in a 39-year old female patient with pos-itive reexposure (Strahl et al 1998) During thefirst period that the kava product was takenan oral contraceptive and paroxetine were con-comitant medications It appears that paroxe-tine was not the only antidepressant taken bythis patient who also occasionally took StJohnrsquos wort (Hypericum perforatum) The Exec-utive Summary issued by the Bundesverbandder Arzneimittel Hersteller eV (BAH) andBundesverband der Pharmazeutischen Indus-trie eV (BP) (see Appendix 3) stated ldquoA causalrelationship with kava cannot be excluded butthe patientrsquos history and a potential preexistingliver damage must be taken into account In ad-dition the kava preparation used was not iden-tified by the physicianrdquo

Moreover in this case taking paroxetine incombination with an oral contraceptive maywell have led to overburdening the liver a sit-uation that could have been exacerbated by tak-ing a kava preparation Schmidt and Nahrstedt(2002) suggested that this case may be associ-ated with an immunologic reaction After re-viewing all of the cases in detail Schmidt andNahrstedt (2002) concluded that this is the onlycase for which there was sufficient informationto make an association with kava appear prob-able and for which the dose of kava also con-formed to that recommended by the Com-mission E monograph (Blumenthal 2000)However as discussed below Russmann et al(2001) tested this patient for CYP2D6 and as inCase 16 found this patient to be CYP2D6 defi-cient which appears to have made her partic-ularly vulnerable to the cocktail of drugs shewas taking Given the complicating features ofthis case we submit that this case should beclassed as possible rather than probable

Case 13 This was a case of a 62-year-oldwoman with jaundice The BfArM table (seeAppendix 2) noted regarding concomitantmedication that there was ldquonone denotedrdquo butit was claimed that concomitant medication didexist but was ldquounknownrdquo The insufficiency ofdata provided for this case was highlighted byBfArMrsquos warning note ldquoNo medical messagerdquoIn addition it should be noted that no detailsof the dosage of kava or period of its adminis-tration were apparently recorded for this caseThis is clearly insufficient information onwhich to base a probable assessment

Case 16 This case concerned a 33-year-oldwoman with jaundice The woman wasrecorded as having taken an overdose of alco-hol measured at 60 g (Russman et al 2001) andthen analgesics including paracetamol fol-lowing this alcohol binge Despite the massiveintake of alcohol a liver biopsy indicated thata drug rather than an alcohol induced toxicgenesis However this case like that of Case 10above was discussed by Russman et al (2001)who demonstrated afterward that this patientwas shown to be CYP2D6 deficient which (asdiscussed below) seems to be a risk factor forthe hepatotoxicity that was ascribed to kavaWe submit that given these circumstances thiscase should be considered possible rather thanprobable

Case 18 This case concerned a 50-year-oldman who had necrosis leading to a liver trans-plant This patient took a product manufac-tured by acetone extraction at a dose deliver-ing 210ndash280 mg of kavalactones per day for 15months ldquomoderate alcoholrdquo (ldquomoderaterdquo is notdefined by BfArM) evening primrose(Oenothera biennis) and a yeast preparationThe dosage of kava was well above the Ger-man Commission E recommended dose ofkavalactones (Blumenthal 1998) It was alsorecorded that 500ndash1000 mg of paracetamol wastaken by this patient shortly before transplan-tation The combination of paracetamol and al-cohol plus the very high dose of kava extractedin acetone taken by this man casts seriousdoubts on the assessment of probable in thiscase

Case 28 (BAH) This case concerned a

KAVA WORK-IN-PROGRESS 243

woman age unknown with hepatitis This caseis hard to assess because neither the patientrsquosage nor diagnosis was given and the womanwas taking eleven medications including estra-diol valerate acetylcysteine losartan (which israrely be associated with hepatitis) and mepra-zole (which can be associated with liver diseasealthough this is rare) Omeprazole is metabo-lized by the polymorphic CYP2C19 which is absent in 3 of Caucasians (Flockhart et al2000) The woman was also taking echinacea(Echinacea purpurea) and five products that ap-peared to be for upper respiratory problems Itshould be noted that this patient was taking syn-thetic kavain not kava A comment from BfArMconcerning this case noted ldquorecurrence of the he-patic side-effectsrdquo which has evidently been in-terpreted by some authorities as being equiva-lent to a ldquopositive rechallengerdquo Whether or notthis was actually so was not clear from the datasupplied It appears (Schmidt and Nahrstedt2002) that Case 28 has been published as twocases with slightly different details This is con-fusing and considering that the woman was tak-ing 11 other medications together with a syn-thetic kava (which we submit is not equivalentto natural kava) and that no diagnosis of hercondition was supplied this calls the assessmentof probable in this case into question

(2) Cases associated with taking synthetic kavain

In this category there were 4 cases 1 2 19and 28

In each of these cases the patients concernedwere taking a product made from synthetickavain Although the outcome was hepatitis inall four cases kavain cannot be equated withthe naturally occurring form of kava whichcontains many other constituents that may playan important role in ensuring the safety of thisherb Therefore we submit that no inferenceshould be drawn from these cases Traditionalusage should not be taken as evidence for safeusage of synthetic products

(3) Patients who were taking oral contraceptivepills or hormone replacement therapy (HRT)together with drugs that can also be associatedwith liver damage

The cases in this category were 4 10 12 2021 and 28

Cholestatic jaundice associated with use ofestrogen-containing medications is extremelyrare (Lindberg 1992) but does occur In these6 cases the women were also taking drugs thatcan also be associated with jaundice

Case 4 This case involved a 39-year-oldwoman with jaundice She was on diazepam10 mg PRN for 6 months Some authoritiescalled this case possible Our assessment is thatthe case is unassessable

Case 10 This case involved a 39-year-oldwoman with necrotizing hepatitis For a de-tailed assessment see above

Case 12 This case involved a 37-year-oldwoman with hepatitis She was on 150 mg ofdiclofenac via intramuscular injection Hepato-toxic reactions associated with nonsteroidalanti-inflammatory drug use are extremely rareand concomitant exposure to other hepatotoxicdrugs is considered to be an important factor(Bareille et al 2001) This case of hepatitis isdifficult to interpret because it occurred inBrazil and because ldquoreexposure was said to benegative for all three drugsrdquo We regard thiscase as unassessable

Case 20 This case involved 50-year-oldwoman with necrosis who had a liver trans-plant She had a 20-year history of combinedoral contraceptive use but had changed monthsearlier to estradiol valerate (which was appar-ently taken alone) as HRT She had also startedglimepiride 8 months earlier This is used fortreating type II diabetes and is rarely associatedwith cholestatic jaundice and liver failure Weregard this case as unlikely

Case 21 This case involved a 22-year-oldwoman with necrosis who had a liver trans-plant This woman had changed from Valette(Jenapharm GmbH Jena Germany) (2 mg ofdienogest and 003 mg of ethinlestradiol) toPramino (180215250 mcg of norgestimateand mcg 25 of ethinylestradiol) She also tookrizatriptan if required for migraine reliefRizatriptan should be used with caution in he-patic impairment and avoided if a patient hassevere liver disease Some authorities considerthis case as being possible but our assessment

DENHAM ET AL244

is in view of the other medications taken isthat this case is unassessable

Case 28 This case involved a woman age un-known with hepatitis This case is discussed atlength above As noted above this patient wastaking synthetic kavain not kava

(4) Patients who were taking drugs that can beassociated with liver damage

There were ten cases in this category 1 6 914 15 17 19 23 2627 and 29

Case 1 This case involved a woman age 69with cholestatic hepatitis She was taking pen-toxifylline (which can be associated with intra-hepatic cholestasis) and a diuretic including thepotassium-sparing triamterene (which can beassociated with jaundice) As noted above thispatient was taking synthetic kavain not kavaWe consider this case unassessable

Case 6 This case involved a woman age 50with hepatitis She was taking frusemide(which can be associated with cholestatic jaun-dice) triamterene atenolol and a large dose ofterfenadine (300 mg) The recommended doseof terfenadine in the British National Formu-lary (March 2001) is 60ndash120 mg The Formularyrecommends avoiding this drug in patientswho have hepatic impairment and also says toldquoavoid concomitant administration of drugs li-able to produce electrolyte imbalance such asdiureticsrdquo (British National Formulary 2001)Despite this warning this woman was also tak-ing the diuretic frusemide The InterkantonalenKontrollstelle der Schweiz of Switzerland con-sidered this case of hepatitis to be caused byterfenadine And although some authoritiesregard this case as possible our assessment isthat this case is unlikely

Case 9 This case involved an 81-year-oldwoman who had liver failure and subsequentdeath She was taking hydrochlorothiazide(which can occasionally be associated with in-trahepatic cholestasis) However according toSchmidt and Nahrstedt (2002) there was evi-dence of chronic alcohol abuse and they re-ported that the autopsy showed chronic pan-creatitis that was characteristic of alcoholabuse The autopsy report (Schmidt and

Nahrstedt 2002) apparently said that thesymptoms must have occurred over a periodof at least 18 months The report conceded thatldquohepatic impairment by alcohol [was] not ex-cludedrdquo In these circumstances it seems en-tirely reasonable to claim that this case is un-related to kava use We regard this case asunlikely

Case 14 This case involved a 33-year-oldwoman with hepatitis Cisapride may havebeen taken (which can cause reversible changesthat show in liver-function tests) Cirrhosis ina woman of 33 is an unexplained finding andthe detail in this case is inadequate to elucidateit We consider this case to be unassessable

Case 15 This case involved a 46-year-oldwoman with jaundice She had been taking hy-drochlorothiazide (which can be associatedwith intrahepatic cholestasis) for 55 monthsplus 80 mg of valsartan and 80 mg ofpropanolol per day Some authorities regardthis case as possible but we consider it to beunassessable

Case 17 This case involved a 59-year-oldwoman with jaundice She had taken 100ndash200mg of celecoxib a cyclo-oxygenase-2 inhibitorper day According to the criteria for causalityassessment of adverse reactions some author-ities consider this case to be possible but our as-sessment is that it is unassessable

Case 19 This case involved a 21-year-oldwoman with hepatitis She was taking panto-prazole (which as with omeprazole can be as-sociated with liver disease) She was also takingparacetamol and metoclopramide and had over-dosed on kavain More detail is needed on othermedical conditions suffered by this patient in or-der to interpret this case It is suggested bySchmidt that this woman was using up to 10tablets per day of the product (the recom-mended dose is up to 6 tablets per day) and thatthere was apparently a discussion in her med-ical record file that she may also have used Ec-stasy (substance that has been associated with

KAVA WORK-IN-PROGRESS 245

Personal communication from M McGuffin to M McIn-tyre available as an online document at ehpaglobalnetcouk

fulminant hepatic failure) This case appears tobe unassessable

Case 23 This case involved a 35-year-oldwoman with jaundice According to the BfArM(see Appendix 2) this patient also took parac-etamol but no dosage or details were providedThis case and case 25 in the BfArM listing ap-pear to be the same case Both cases have beenlabeled as possible by some authorities butgiven the lack of information about the dosageof paracetamol and the apparent confusion re-garding cases 23 and 25 we submit that theonly logical assessment is unassessable

Case 2627 This case involved a woman whowas either 38 or 39 yearsrsquo old with hepatitis Itappears that the two cases have been duplicated(Schmidt and Nahrstedt 2002) The confusionwith this case is another example of inaccuratedata provided by the BfArM Information re-garding these cases (or case) depending onwhether the two reports concern the samewoman is unclear Penicillin can be associatedwith hypersensitivity and cholestatic jaundicebut the information given is inadequate to makeany meaningful assessment For this reason weclass this case as unassessable

Case 29 This case involved a 60-year-oldwoman who had a liver transplant This womanwas taking piretamide (which is a loop diuretic)Frusemide another loop diuretic can be associ-ated with cholestatic jaundice According to theBfArM chart (see Appendix 2) she was also tak-ing a sympathomimetic drug etilefrin Thedosage of kava varied but was up to 480ndash1200mg per day (Schmidt and Nahrstedt 2002)which is up to ten times the German Commis-sion E maximum recommended dose (Blumen-thal 1998) Although some authorities have re-garded this case as possible in view of themarked overdosing of kava and the concomitantmedication this case can hardly be said to be areflection on the proper therapeutic use of kava

(5) Cases in which drugs not associated withliver damage herbal medicines or dietarysupplements or kavain alone were taken

This category had eight cases 2 78 11 1322 24 and 25

For these cases detail was limited and theBfArM did not implicate any other drugs ormedications (although this may not be thecase)

All patients in this group apart from the pa-tient in Case 78 for whom no information wasgiven were reported to have made full recov-eries In some of these cases it is not clearwhether the patients were ill or whether thesecases merely recorded raised liver-function en-zymes

Case 2 This case involved a 35-year-old manwith cholestatic hepatitis Concomitant med-ication was ldquounknownrdquo Apart from Cases 18and 30 this is the only case for which it is pos-sible that no other concomitant medication wastaken but there is a marked lack of informationfor this case As noted above this patient wastaking synthetic kavain not kava We regardthis case as unassessable

Case 5 This case involved a woman who waseither 68 or 69 yearsrsquo old with cholestatic he-patitis She was also taking a St Johnrsquos wort(Hypericum perforatum) product which hasbeen associated with CYP3A4 A biopsyshowed ldquoimmunologic hypersensitivityrdquo Thiscase may be regarded as possible but in viewof the immunologic hypersensitivity it maywell have been an idiosyncratic event that wasnot necessarily associated with kava usage

Case 78 This case involved a woman or twowomen ages 72 andor 75 with cholestatic he-patitis These two cases appear to be actuallyone case The woman was taking twoherbalvitamin products one of which in-cluded 06 mg of kavalactones Given the con-fusion involved these ldquocasesrdquo must be re-garded as unassessable

Case 11 This case involved a 59-year-oldwoman who was taking hyoscine butylbro-mide as a suppository Schmidt and Nahrstedt(2002) commented that according to additionalinformation obtained from the BfArM it is un-certain as to whether this patient was taking akava product at all We regard this case asunassessable

DENHAM ET AL246

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

(3) Definitive isolation of the active constituentis elusive in other medicinal plants such asHypericum perforatum (McIntyre 2000Barnes et al 2001)

(4) Herbal practitioners rely on the synergy be-tween a whole range of constituents in theherb or herb within a herbal prescriptionwhich is individually prescribed for a pa-tient This positive interaction may alsohave the benefit of keeping levels of anyone constituent below the safety threshold

LOW-ALCOHOL TINCTURES

The Traditional Medicines Evaluation Com-mittee (TMEC) strongly advocates the use ofextraction techniques that closely approximatethose traditionally used in Polynesia Thiswould require the use of low-alcohol tincturesmade by the traditional cold macerationprocesses common to UK tincture makingThe reasons for this opinion are set out belowthey have also been explored by Whitton et al(Appendix 1)

Tinctures used by herbal practitioners areprepared by macerating dried kava in a mix-ture of water and ethanol It has been shownthat such extracts using 25 ethanol75 wa-ter contain up to 30 times fewer kavalactonesthan the concentrated standardized prepara-tions (Appendix 1) The traditional preparationmethod using a mixture of 25 ethanol75water extracts a wider range of the naturalkava constituents (Appendix 1)

DOSAGE AND OVERDOSAGE

Assuming a 15 25 tincture and an upperlimit of 20 kavalactones in the dried herb(concentrations stated as 3ndash20 see sectionon Pharmacology below) then 500 mL of theherb would contain (100 3 02 3 015) 5 3 g(3000 mg) of kavalactones In addition assum-ing a daily dosage of 5ndash10 mL of the 15 25tincture the daily dose of kavalactonesamounts to a maximum of 30ndash60 mg If the con-centration of kavalactones were lower for ex-ample at approximately 10 as appears to bethe case with regard to Australian kava dis-

cussed by Clough et al (2000) then this dosagefalls to 15ndash30 mg per day It is noteworthy thatthe maximum daily dosage here is equivalentto the minimum daily dosages of the 60ndash210mg kavalactones given in clinical trials of kavaconducted in Germany

Although standardized extracts provide ahigher dosage of kavalactones than low-alco-hol tinctures overdosage in itself is unlikelyto be the cause of hepatotoxicity Strong evi-dence for this is the fact that kava is takendaily at high doses as a normal part of dailylife in large areas of the South Pacific Indeedsome of the accounts of high kava intake areremarkable For example Chanwai (2000) de-scribed the case of a man who was admittedto a hospital after an overdose but ldquoslept offrdquohis symptoms and admitted to consuming upto 40 bowls of a kava preparation per day forthe last 14 years In Australia missionaries in-troduced kava to the aborigines in the 1980sas a substitute for alcohol and it is claimed thatthis has led to abuse of kava Clough et al(2000) discussed this concern and reviewedtwelve studies on the amount of kava usedThe researchers found that social setting ap-pears to determine the amount used with lonedrinkers consuming much more than peoplewho enjoy kava in a family group The re-searchers described normal use of kava in theNorthern Territory as being 37 g of kava pow-der (containing approximately 3800 mg ofkavalactones) per hour with heavy consumersusing approximately 610 g per week preparedas a drink The incidence of serious illness re-sulting from hepatotoxicity associated withregular kava usage would surely have beenobserved by the medical services in Polynesiaand Australia if overdosage of kavalactoneswere the main cause of hepatotoxicity

UNTOWARD EFFECTS

There is a justified concern in Europe thatidiosyncratic hepatotoxicity associated with us-ing some herbal medicines may not be identi-fied because the population that takes herbalmedicines is not large enough to produce suf-ficient cases for the association to be noted Butthe fact that kava remains in traditional usageto such a wide extent is a powerful argument

KAVA WORK-IN-PROGRESS 241

that idiosyncratic hepatotoxicity would havebeen noted

Two postmarketing observation studies inGermany each on more than 3000 people werecited by Pittler and Ernst (2000) in addition tothe abovementioned clinical trials In these ob-servational studies the rate of adverse eventswas 23 (with a daily dose of 120ndash240 mg ofkavalactones) and 15 (with a daily dose of105 mg of kavalactones) The most frequent ad-verse reports were gastrointestinal complaintsallergic skin reactions headaches and photo-sensitivity

There is evidence in the South Pacific of a char-acteristic kava-induced skin disease a scaly rashthat is suggestive of icthyosismdasha condition calledldquokava dermopathyrdquo (Ruze 1990) Although theskin becomes yellow the description does notsuggest an underlying hepatic condition in thatthe patient remains well the rash is not itchyand the condition is ameliorated without treat-ment if heavy use of kava is reduced

The German and Swiss reports cited by theBfArM are of concern because there have beenprevious reports of hepatotoxicity associatedwith the use of some medicinal plants (Larrey1997) The kava case reports from the BfArM(see Appendix 2) include all three of the mainforms of acute damage that can result from ad-verse drug reactions (1) necrosis (2) drug-in-duced hepatitis and (3) cholestatic hepatitis(Hodgson and Levi 1997) This suggests thatthere is a range of causes rather than just onecause in these cases The BfArM case reportshave been circulated worldwide and are cur-rently being evaluated by government agenciesin Europe Australia Canada the UnitedStates and elsewhere We have received anumber of informal case assessments fromthese sources that cannot be specifically citedbecause of their confidential status To achievetransparency and encourage a full debate aboutkava however the BfArM cases are evaluatedin the section entitled Discussion of Cases Re-ported by the BfArM

CRITERIA FOR ASSESSING THE CASE REPORTS

A recent review of the information availableon the case reports (Schmidt and Nahrstedt

2002) is supported by details of the case re-ports on the Web site of the University ofMuenster (wwwuni-muensterdechemiepbkavaanalysehtml)

The criteria for causality assessment of ad-verse reactions used are as follows (Edwardsand Aronson 2000)Probable is defined as

A clinical event including a laboratory testabnormality that occurs in a plausible timerelation to drug administration and that can-not be explained by coincidental or concur-rent disease or other drugs or chemicals

The response to withdrawal of the drug(dechallange) should be clinically plausible

The event must be definitive pharmacolog-ically or phenomenologically using a satis-factory rechallenge procedure if necessary

Possible is defined as

A clinical event including a laboratory testabnormality with a reasonable time relationto administration of the drug but that couldbe explained by concurrent disease or otherdrugs or chemicals

Information on drug withdrawal may belacking or unclear

Unlikely is defined as

A clinical event including a laboratory testabnormality with a temporal relation to theadministration of the drug which makes acausal relation improbable and in whichother drugs chemicals or underlying dis-ease provide plausible explanations

Unassessable is defined as

A report suggesting an adverse reaction thatcannot be judged because information is in-sufficient or contradictory and cannot besupplemented or verified

DISCUSSION OF CASES REPORTED BY THE BfArM

The cases discussed below are analyzed andcategorized by common factors of note withour own assessments

DENHAM ET AL242

(1) Cases of most concern

This group includes five cases 10 13 16 18and 28 According to the assessments made by various government agencies these casescause the most concern and are often cited asbeing probable For this reason these cases aredealt with first As discussed below mostmdashifnot allmdashof these cases have associated factorsthat put this probable categorization into ques-tion

Case 10 This case described necrotizing hep-atitis in a 39-year old female patient with pos-itive reexposure (Strahl et al 1998) During thefirst period that the kava product was takenan oral contraceptive and paroxetine were con-comitant medications It appears that paroxe-tine was not the only antidepressant taken bythis patient who also occasionally took StJohnrsquos wort (Hypericum perforatum) The Exec-utive Summary issued by the Bundesverbandder Arzneimittel Hersteller eV (BAH) andBundesverband der Pharmazeutischen Indus-trie eV (BP) (see Appendix 3) stated ldquoA causalrelationship with kava cannot be excluded butthe patientrsquos history and a potential preexistingliver damage must be taken into account In ad-dition the kava preparation used was not iden-tified by the physicianrdquo

Moreover in this case taking paroxetine incombination with an oral contraceptive maywell have led to overburdening the liver a sit-uation that could have been exacerbated by tak-ing a kava preparation Schmidt and Nahrstedt(2002) suggested that this case may be associ-ated with an immunologic reaction After re-viewing all of the cases in detail Schmidt andNahrstedt (2002) concluded that this is the onlycase for which there was sufficient informationto make an association with kava appear prob-able and for which the dose of kava also con-formed to that recommended by the Com-mission E monograph (Blumenthal 2000)However as discussed below Russmann et al(2001) tested this patient for CYP2D6 and as inCase 16 found this patient to be CYP2D6 defi-cient which appears to have made her partic-ularly vulnerable to the cocktail of drugs shewas taking Given the complicating features ofthis case we submit that this case should beclassed as possible rather than probable

Case 13 This was a case of a 62-year-oldwoman with jaundice The BfArM table (seeAppendix 2) noted regarding concomitantmedication that there was ldquonone denotedrdquo butit was claimed that concomitant medication didexist but was ldquounknownrdquo The insufficiency ofdata provided for this case was highlighted byBfArMrsquos warning note ldquoNo medical messagerdquoIn addition it should be noted that no detailsof the dosage of kava or period of its adminis-tration were apparently recorded for this caseThis is clearly insufficient information onwhich to base a probable assessment

Case 16 This case concerned a 33-year-oldwoman with jaundice The woman wasrecorded as having taken an overdose of alco-hol measured at 60 g (Russman et al 2001) andthen analgesics including paracetamol fol-lowing this alcohol binge Despite the massiveintake of alcohol a liver biopsy indicated thata drug rather than an alcohol induced toxicgenesis However this case like that of Case 10above was discussed by Russman et al (2001)who demonstrated afterward that this patientwas shown to be CYP2D6 deficient which (asdiscussed below) seems to be a risk factor forthe hepatotoxicity that was ascribed to kavaWe submit that given these circumstances thiscase should be considered possible rather thanprobable

Case 18 This case concerned a 50-year-oldman who had necrosis leading to a liver trans-plant This patient took a product manufac-tured by acetone extraction at a dose deliver-ing 210ndash280 mg of kavalactones per day for 15months ldquomoderate alcoholrdquo (ldquomoderaterdquo is notdefined by BfArM) evening primrose(Oenothera biennis) and a yeast preparationThe dosage of kava was well above the Ger-man Commission E recommended dose ofkavalactones (Blumenthal 1998) It was alsorecorded that 500ndash1000 mg of paracetamol wastaken by this patient shortly before transplan-tation The combination of paracetamol and al-cohol plus the very high dose of kava extractedin acetone taken by this man casts seriousdoubts on the assessment of probable in thiscase

Case 28 (BAH) This case concerned a

KAVA WORK-IN-PROGRESS 243

woman age unknown with hepatitis This caseis hard to assess because neither the patientrsquosage nor diagnosis was given and the womanwas taking eleven medications including estra-diol valerate acetylcysteine losartan (which israrely be associated with hepatitis) and mepra-zole (which can be associated with liver diseasealthough this is rare) Omeprazole is metabo-lized by the polymorphic CYP2C19 which is absent in 3 of Caucasians (Flockhart et al2000) The woman was also taking echinacea(Echinacea purpurea) and five products that ap-peared to be for upper respiratory problems Itshould be noted that this patient was taking syn-thetic kavain not kava A comment from BfArMconcerning this case noted ldquorecurrence of the he-patic side-effectsrdquo which has evidently been in-terpreted by some authorities as being equiva-lent to a ldquopositive rechallengerdquo Whether or notthis was actually so was not clear from the datasupplied It appears (Schmidt and Nahrstedt2002) that Case 28 has been published as twocases with slightly different details This is con-fusing and considering that the woman was tak-ing 11 other medications together with a syn-thetic kava (which we submit is not equivalentto natural kava) and that no diagnosis of hercondition was supplied this calls the assessmentof probable in this case into question

(2) Cases associated with taking synthetic kavain

In this category there were 4 cases 1 2 19and 28

In each of these cases the patients concernedwere taking a product made from synthetickavain Although the outcome was hepatitis inall four cases kavain cannot be equated withthe naturally occurring form of kava whichcontains many other constituents that may playan important role in ensuring the safety of thisherb Therefore we submit that no inferenceshould be drawn from these cases Traditionalusage should not be taken as evidence for safeusage of synthetic products

(3) Patients who were taking oral contraceptivepills or hormone replacement therapy (HRT)together with drugs that can also be associatedwith liver damage

The cases in this category were 4 10 12 2021 and 28

Cholestatic jaundice associated with use ofestrogen-containing medications is extremelyrare (Lindberg 1992) but does occur In these6 cases the women were also taking drugs thatcan also be associated with jaundice

Case 4 This case involved a 39-year-oldwoman with jaundice She was on diazepam10 mg PRN for 6 months Some authoritiescalled this case possible Our assessment is thatthe case is unassessable

Case 10 This case involved a 39-year-oldwoman with necrotizing hepatitis For a de-tailed assessment see above

Case 12 This case involved a 37-year-oldwoman with hepatitis She was on 150 mg ofdiclofenac via intramuscular injection Hepato-toxic reactions associated with nonsteroidalanti-inflammatory drug use are extremely rareand concomitant exposure to other hepatotoxicdrugs is considered to be an important factor(Bareille et al 2001) This case of hepatitis isdifficult to interpret because it occurred inBrazil and because ldquoreexposure was said to benegative for all three drugsrdquo We regard thiscase as unassessable

Case 20 This case involved 50-year-oldwoman with necrosis who had a liver trans-plant She had a 20-year history of combinedoral contraceptive use but had changed monthsearlier to estradiol valerate (which was appar-ently taken alone) as HRT She had also startedglimepiride 8 months earlier This is used fortreating type II diabetes and is rarely associatedwith cholestatic jaundice and liver failure Weregard this case as unlikely

Case 21 This case involved a 22-year-oldwoman with necrosis who had a liver trans-plant This woman had changed from Valette(Jenapharm GmbH Jena Germany) (2 mg ofdienogest and 003 mg of ethinlestradiol) toPramino (180215250 mcg of norgestimateand mcg 25 of ethinylestradiol) She also tookrizatriptan if required for migraine reliefRizatriptan should be used with caution in he-patic impairment and avoided if a patient hassevere liver disease Some authorities considerthis case as being possible but our assessment

DENHAM ET AL244

is in view of the other medications taken isthat this case is unassessable

Case 28 This case involved a woman age un-known with hepatitis This case is discussed atlength above As noted above this patient wastaking synthetic kavain not kava

(4) Patients who were taking drugs that can beassociated with liver damage

There were ten cases in this category 1 6 914 15 17 19 23 2627 and 29

Case 1 This case involved a woman age 69with cholestatic hepatitis She was taking pen-toxifylline (which can be associated with intra-hepatic cholestasis) and a diuretic including thepotassium-sparing triamterene (which can beassociated with jaundice) As noted above thispatient was taking synthetic kavain not kavaWe consider this case unassessable

Case 6 This case involved a woman age 50with hepatitis She was taking frusemide(which can be associated with cholestatic jaun-dice) triamterene atenolol and a large dose ofterfenadine (300 mg) The recommended doseof terfenadine in the British National Formu-lary (March 2001) is 60ndash120 mg The Formularyrecommends avoiding this drug in patientswho have hepatic impairment and also says toldquoavoid concomitant administration of drugs li-able to produce electrolyte imbalance such asdiureticsrdquo (British National Formulary 2001)Despite this warning this woman was also tak-ing the diuretic frusemide The InterkantonalenKontrollstelle der Schweiz of Switzerland con-sidered this case of hepatitis to be caused byterfenadine And although some authoritiesregard this case as possible our assessment isthat this case is unlikely

Case 9 This case involved an 81-year-oldwoman who had liver failure and subsequentdeath She was taking hydrochlorothiazide(which can occasionally be associated with in-trahepatic cholestasis) However according toSchmidt and Nahrstedt (2002) there was evi-dence of chronic alcohol abuse and they re-ported that the autopsy showed chronic pan-creatitis that was characteristic of alcoholabuse The autopsy report (Schmidt and

Nahrstedt 2002) apparently said that thesymptoms must have occurred over a periodof at least 18 months The report conceded thatldquohepatic impairment by alcohol [was] not ex-cludedrdquo In these circumstances it seems en-tirely reasonable to claim that this case is un-related to kava use We regard this case asunlikely

Case 14 This case involved a 33-year-oldwoman with hepatitis Cisapride may havebeen taken (which can cause reversible changesthat show in liver-function tests) Cirrhosis ina woman of 33 is an unexplained finding andthe detail in this case is inadequate to elucidateit We consider this case to be unassessable

Case 15 This case involved a 46-year-oldwoman with jaundice She had been taking hy-drochlorothiazide (which can be associatedwith intrahepatic cholestasis) for 55 monthsplus 80 mg of valsartan and 80 mg ofpropanolol per day Some authorities regardthis case as possible but we consider it to beunassessable

Case 17 This case involved a 59-year-oldwoman with jaundice She had taken 100ndash200mg of celecoxib a cyclo-oxygenase-2 inhibitorper day According to the criteria for causalityassessment of adverse reactions some author-ities consider this case to be possible but our as-sessment is that it is unassessable

Case 19 This case involved a 21-year-oldwoman with hepatitis She was taking panto-prazole (which as with omeprazole can be as-sociated with liver disease) She was also takingparacetamol and metoclopramide and had over-dosed on kavain More detail is needed on othermedical conditions suffered by this patient in or-der to interpret this case It is suggested bySchmidt that this woman was using up to 10tablets per day of the product (the recom-mended dose is up to 6 tablets per day) and thatthere was apparently a discussion in her med-ical record file that she may also have used Ec-stasy (substance that has been associated with

KAVA WORK-IN-PROGRESS 245

Personal communication from M McGuffin to M McIn-tyre available as an online document at ehpaglobalnetcouk

fulminant hepatic failure) This case appears tobe unassessable

Case 23 This case involved a 35-year-oldwoman with jaundice According to the BfArM(see Appendix 2) this patient also took parac-etamol but no dosage or details were providedThis case and case 25 in the BfArM listing ap-pear to be the same case Both cases have beenlabeled as possible by some authorities butgiven the lack of information about the dosageof paracetamol and the apparent confusion re-garding cases 23 and 25 we submit that theonly logical assessment is unassessable

Case 2627 This case involved a woman whowas either 38 or 39 yearsrsquo old with hepatitis Itappears that the two cases have been duplicated(Schmidt and Nahrstedt 2002) The confusionwith this case is another example of inaccuratedata provided by the BfArM Information re-garding these cases (or case) depending onwhether the two reports concern the samewoman is unclear Penicillin can be associatedwith hypersensitivity and cholestatic jaundicebut the information given is inadequate to makeany meaningful assessment For this reason weclass this case as unassessable

Case 29 This case involved a 60-year-oldwoman who had a liver transplant This womanwas taking piretamide (which is a loop diuretic)Frusemide another loop diuretic can be associ-ated with cholestatic jaundice According to theBfArM chart (see Appendix 2) she was also tak-ing a sympathomimetic drug etilefrin Thedosage of kava varied but was up to 480ndash1200mg per day (Schmidt and Nahrstedt 2002)which is up to ten times the German Commis-sion E maximum recommended dose (Blumen-thal 1998) Although some authorities have re-garded this case as possible in view of themarked overdosing of kava and the concomitantmedication this case can hardly be said to be areflection on the proper therapeutic use of kava

(5) Cases in which drugs not associated withliver damage herbal medicines or dietarysupplements or kavain alone were taken

This category had eight cases 2 78 11 1322 24 and 25

For these cases detail was limited and theBfArM did not implicate any other drugs ormedications (although this may not be thecase)

All patients in this group apart from the pa-tient in Case 78 for whom no information wasgiven were reported to have made full recov-eries In some of these cases it is not clearwhether the patients were ill or whether thesecases merely recorded raised liver-function en-zymes

Case 2 This case involved a 35-year-old manwith cholestatic hepatitis Concomitant med-ication was ldquounknownrdquo Apart from Cases 18and 30 this is the only case for which it is pos-sible that no other concomitant medication wastaken but there is a marked lack of informationfor this case As noted above this patient wastaking synthetic kavain not kava We regardthis case as unassessable

Case 5 This case involved a woman who waseither 68 or 69 yearsrsquo old with cholestatic he-patitis She was also taking a St Johnrsquos wort(Hypericum perforatum) product which hasbeen associated with CYP3A4 A biopsyshowed ldquoimmunologic hypersensitivityrdquo Thiscase may be regarded as possible but in viewof the immunologic hypersensitivity it maywell have been an idiosyncratic event that wasnot necessarily associated with kava usage

Case 78 This case involved a woman or twowomen ages 72 andor 75 with cholestatic he-patitis These two cases appear to be actuallyone case The woman was taking twoherbalvitamin products one of which in-cluded 06 mg of kavalactones Given the con-fusion involved these ldquocasesrdquo must be re-garded as unassessable

Case 11 This case involved a 59-year-oldwoman who was taking hyoscine butylbro-mide as a suppository Schmidt and Nahrstedt(2002) commented that according to additionalinformation obtained from the BfArM it is un-certain as to whether this patient was taking akava product at all We regard this case asunassessable

DENHAM ET AL246

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

that idiosyncratic hepatotoxicity would havebeen noted

Two postmarketing observation studies inGermany each on more than 3000 people werecited by Pittler and Ernst (2000) in addition tothe abovementioned clinical trials In these ob-servational studies the rate of adverse eventswas 23 (with a daily dose of 120ndash240 mg ofkavalactones) and 15 (with a daily dose of105 mg of kavalactones) The most frequent ad-verse reports were gastrointestinal complaintsallergic skin reactions headaches and photo-sensitivity

There is evidence in the South Pacific of a char-acteristic kava-induced skin disease a scaly rashthat is suggestive of icthyosismdasha condition calledldquokava dermopathyrdquo (Ruze 1990) Although theskin becomes yellow the description does notsuggest an underlying hepatic condition in thatthe patient remains well the rash is not itchyand the condition is ameliorated without treat-ment if heavy use of kava is reduced

The German and Swiss reports cited by theBfArM are of concern because there have beenprevious reports of hepatotoxicity associatedwith the use of some medicinal plants (Larrey1997) The kava case reports from the BfArM(see Appendix 2) include all three of the mainforms of acute damage that can result from ad-verse drug reactions (1) necrosis (2) drug-in-duced hepatitis and (3) cholestatic hepatitis(Hodgson and Levi 1997) This suggests thatthere is a range of causes rather than just onecause in these cases The BfArM case reportshave been circulated worldwide and are cur-rently being evaluated by government agenciesin Europe Australia Canada the UnitedStates and elsewhere We have received anumber of informal case assessments fromthese sources that cannot be specifically citedbecause of their confidential status To achievetransparency and encourage a full debate aboutkava however the BfArM cases are evaluatedin the section entitled Discussion of Cases Re-ported by the BfArM

CRITERIA FOR ASSESSING THE CASE REPORTS

A recent review of the information availableon the case reports (Schmidt and Nahrstedt

2002) is supported by details of the case re-ports on the Web site of the University ofMuenster (wwwuni-muensterdechemiepbkavaanalysehtml)

The criteria for causality assessment of ad-verse reactions used are as follows (Edwardsand Aronson 2000)Probable is defined as

A clinical event including a laboratory testabnormality that occurs in a plausible timerelation to drug administration and that can-not be explained by coincidental or concur-rent disease or other drugs or chemicals

The response to withdrawal of the drug(dechallange) should be clinically plausible

The event must be definitive pharmacolog-ically or phenomenologically using a satis-factory rechallenge procedure if necessary

Possible is defined as

A clinical event including a laboratory testabnormality with a reasonable time relationto administration of the drug but that couldbe explained by concurrent disease or otherdrugs or chemicals

Information on drug withdrawal may belacking or unclear

Unlikely is defined as

A clinical event including a laboratory testabnormality with a temporal relation to theadministration of the drug which makes acausal relation improbable and in whichother drugs chemicals or underlying dis-ease provide plausible explanations

Unassessable is defined as

A report suggesting an adverse reaction thatcannot be judged because information is in-sufficient or contradictory and cannot besupplemented or verified

DISCUSSION OF CASES REPORTED BY THE BfArM

The cases discussed below are analyzed andcategorized by common factors of note withour own assessments

DENHAM ET AL242

(1) Cases of most concern

This group includes five cases 10 13 16 18and 28 According to the assessments made by various government agencies these casescause the most concern and are often cited asbeing probable For this reason these cases aredealt with first As discussed below mostmdashifnot allmdashof these cases have associated factorsthat put this probable categorization into ques-tion

Case 10 This case described necrotizing hep-atitis in a 39-year old female patient with pos-itive reexposure (Strahl et al 1998) During thefirst period that the kava product was takenan oral contraceptive and paroxetine were con-comitant medications It appears that paroxe-tine was not the only antidepressant taken bythis patient who also occasionally took StJohnrsquos wort (Hypericum perforatum) The Exec-utive Summary issued by the Bundesverbandder Arzneimittel Hersteller eV (BAH) andBundesverband der Pharmazeutischen Indus-trie eV (BP) (see Appendix 3) stated ldquoA causalrelationship with kava cannot be excluded butthe patientrsquos history and a potential preexistingliver damage must be taken into account In ad-dition the kava preparation used was not iden-tified by the physicianrdquo

Moreover in this case taking paroxetine incombination with an oral contraceptive maywell have led to overburdening the liver a sit-uation that could have been exacerbated by tak-ing a kava preparation Schmidt and Nahrstedt(2002) suggested that this case may be associ-ated with an immunologic reaction After re-viewing all of the cases in detail Schmidt andNahrstedt (2002) concluded that this is the onlycase for which there was sufficient informationto make an association with kava appear prob-able and for which the dose of kava also con-formed to that recommended by the Com-mission E monograph (Blumenthal 2000)However as discussed below Russmann et al(2001) tested this patient for CYP2D6 and as inCase 16 found this patient to be CYP2D6 defi-cient which appears to have made her partic-ularly vulnerable to the cocktail of drugs shewas taking Given the complicating features ofthis case we submit that this case should beclassed as possible rather than probable

Case 13 This was a case of a 62-year-oldwoman with jaundice The BfArM table (seeAppendix 2) noted regarding concomitantmedication that there was ldquonone denotedrdquo butit was claimed that concomitant medication didexist but was ldquounknownrdquo The insufficiency ofdata provided for this case was highlighted byBfArMrsquos warning note ldquoNo medical messagerdquoIn addition it should be noted that no detailsof the dosage of kava or period of its adminis-tration were apparently recorded for this caseThis is clearly insufficient information onwhich to base a probable assessment

Case 16 This case concerned a 33-year-oldwoman with jaundice The woman wasrecorded as having taken an overdose of alco-hol measured at 60 g (Russman et al 2001) andthen analgesics including paracetamol fol-lowing this alcohol binge Despite the massiveintake of alcohol a liver biopsy indicated thata drug rather than an alcohol induced toxicgenesis However this case like that of Case 10above was discussed by Russman et al (2001)who demonstrated afterward that this patientwas shown to be CYP2D6 deficient which (asdiscussed below) seems to be a risk factor forthe hepatotoxicity that was ascribed to kavaWe submit that given these circumstances thiscase should be considered possible rather thanprobable

Case 18 This case concerned a 50-year-oldman who had necrosis leading to a liver trans-plant This patient took a product manufac-tured by acetone extraction at a dose deliver-ing 210ndash280 mg of kavalactones per day for 15months ldquomoderate alcoholrdquo (ldquomoderaterdquo is notdefined by BfArM) evening primrose(Oenothera biennis) and a yeast preparationThe dosage of kava was well above the Ger-man Commission E recommended dose ofkavalactones (Blumenthal 1998) It was alsorecorded that 500ndash1000 mg of paracetamol wastaken by this patient shortly before transplan-tation The combination of paracetamol and al-cohol plus the very high dose of kava extractedin acetone taken by this man casts seriousdoubts on the assessment of probable in thiscase

Case 28 (BAH) This case concerned a

KAVA WORK-IN-PROGRESS 243

woman age unknown with hepatitis This caseis hard to assess because neither the patientrsquosage nor diagnosis was given and the womanwas taking eleven medications including estra-diol valerate acetylcysteine losartan (which israrely be associated with hepatitis) and mepra-zole (which can be associated with liver diseasealthough this is rare) Omeprazole is metabo-lized by the polymorphic CYP2C19 which is absent in 3 of Caucasians (Flockhart et al2000) The woman was also taking echinacea(Echinacea purpurea) and five products that ap-peared to be for upper respiratory problems Itshould be noted that this patient was taking syn-thetic kavain not kava A comment from BfArMconcerning this case noted ldquorecurrence of the he-patic side-effectsrdquo which has evidently been in-terpreted by some authorities as being equiva-lent to a ldquopositive rechallengerdquo Whether or notthis was actually so was not clear from the datasupplied It appears (Schmidt and Nahrstedt2002) that Case 28 has been published as twocases with slightly different details This is con-fusing and considering that the woman was tak-ing 11 other medications together with a syn-thetic kava (which we submit is not equivalentto natural kava) and that no diagnosis of hercondition was supplied this calls the assessmentof probable in this case into question

(2) Cases associated with taking synthetic kavain

In this category there were 4 cases 1 2 19and 28

In each of these cases the patients concernedwere taking a product made from synthetickavain Although the outcome was hepatitis inall four cases kavain cannot be equated withthe naturally occurring form of kava whichcontains many other constituents that may playan important role in ensuring the safety of thisherb Therefore we submit that no inferenceshould be drawn from these cases Traditionalusage should not be taken as evidence for safeusage of synthetic products

(3) Patients who were taking oral contraceptivepills or hormone replacement therapy (HRT)together with drugs that can also be associatedwith liver damage

The cases in this category were 4 10 12 2021 and 28

Cholestatic jaundice associated with use ofestrogen-containing medications is extremelyrare (Lindberg 1992) but does occur In these6 cases the women were also taking drugs thatcan also be associated with jaundice

Case 4 This case involved a 39-year-oldwoman with jaundice She was on diazepam10 mg PRN for 6 months Some authoritiescalled this case possible Our assessment is thatthe case is unassessable

Case 10 This case involved a 39-year-oldwoman with necrotizing hepatitis For a de-tailed assessment see above

Case 12 This case involved a 37-year-oldwoman with hepatitis She was on 150 mg ofdiclofenac via intramuscular injection Hepato-toxic reactions associated with nonsteroidalanti-inflammatory drug use are extremely rareand concomitant exposure to other hepatotoxicdrugs is considered to be an important factor(Bareille et al 2001) This case of hepatitis isdifficult to interpret because it occurred inBrazil and because ldquoreexposure was said to benegative for all three drugsrdquo We regard thiscase as unassessable

Case 20 This case involved 50-year-oldwoman with necrosis who had a liver trans-plant She had a 20-year history of combinedoral contraceptive use but had changed monthsearlier to estradiol valerate (which was appar-ently taken alone) as HRT She had also startedglimepiride 8 months earlier This is used fortreating type II diabetes and is rarely associatedwith cholestatic jaundice and liver failure Weregard this case as unlikely

Case 21 This case involved a 22-year-oldwoman with necrosis who had a liver trans-plant This woman had changed from Valette(Jenapharm GmbH Jena Germany) (2 mg ofdienogest and 003 mg of ethinlestradiol) toPramino (180215250 mcg of norgestimateand mcg 25 of ethinylestradiol) She also tookrizatriptan if required for migraine reliefRizatriptan should be used with caution in he-patic impairment and avoided if a patient hassevere liver disease Some authorities considerthis case as being possible but our assessment

DENHAM ET AL244

is in view of the other medications taken isthat this case is unassessable

Case 28 This case involved a woman age un-known with hepatitis This case is discussed atlength above As noted above this patient wastaking synthetic kavain not kava

(4) Patients who were taking drugs that can beassociated with liver damage

There were ten cases in this category 1 6 914 15 17 19 23 2627 and 29

Case 1 This case involved a woman age 69with cholestatic hepatitis She was taking pen-toxifylline (which can be associated with intra-hepatic cholestasis) and a diuretic including thepotassium-sparing triamterene (which can beassociated with jaundice) As noted above thispatient was taking synthetic kavain not kavaWe consider this case unassessable

Case 6 This case involved a woman age 50with hepatitis She was taking frusemide(which can be associated with cholestatic jaun-dice) triamterene atenolol and a large dose ofterfenadine (300 mg) The recommended doseof terfenadine in the British National Formu-lary (March 2001) is 60ndash120 mg The Formularyrecommends avoiding this drug in patientswho have hepatic impairment and also says toldquoavoid concomitant administration of drugs li-able to produce electrolyte imbalance such asdiureticsrdquo (British National Formulary 2001)Despite this warning this woman was also tak-ing the diuretic frusemide The InterkantonalenKontrollstelle der Schweiz of Switzerland con-sidered this case of hepatitis to be caused byterfenadine And although some authoritiesregard this case as possible our assessment isthat this case is unlikely

Case 9 This case involved an 81-year-oldwoman who had liver failure and subsequentdeath She was taking hydrochlorothiazide(which can occasionally be associated with in-trahepatic cholestasis) However according toSchmidt and Nahrstedt (2002) there was evi-dence of chronic alcohol abuse and they re-ported that the autopsy showed chronic pan-creatitis that was characteristic of alcoholabuse The autopsy report (Schmidt and

Nahrstedt 2002) apparently said that thesymptoms must have occurred over a periodof at least 18 months The report conceded thatldquohepatic impairment by alcohol [was] not ex-cludedrdquo In these circumstances it seems en-tirely reasonable to claim that this case is un-related to kava use We regard this case asunlikely

Case 14 This case involved a 33-year-oldwoman with hepatitis Cisapride may havebeen taken (which can cause reversible changesthat show in liver-function tests) Cirrhosis ina woman of 33 is an unexplained finding andthe detail in this case is inadequate to elucidateit We consider this case to be unassessable

Case 15 This case involved a 46-year-oldwoman with jaundice She had been taking hy-drochlorothiazide (which can be associatedwith intrahepatic cholestasis) for 55 monthsplus 80 mg of valsartan and 80 mg ofpropanolol per day Some authorities regardthis case as possible but we consider it to beunassessable

Case 17 This case involved a 59-year-oldwoman with jaundice She had taken 100ndash200mg of celecoxib a cyclo-oxygenase-2 inhibitorper day According to the criteria for causalityassessment of adverse reactions some author-ities consider this case to be possible but our as-sessment is that it is unassessable

Case 19 This case involved a 21-year-oldwoman with hepatitis She was taking panto-prazole (which as with omeprazole can be as-sociated with liver disease) She was also takingparacetamol and metoclopramide and had over-dosed on kavain More detail is needed on othermedical conditions suffered by this patient in or-der to interpret this case It is suggested bySchmidt that this woman was using up to 10tablets per day of the product (the recom-mended dose is up to 6 tablets per day) and thatthere was apparently a discussion in her med-ical record file that she may also have used Ec-stasy (substance that has been associated with

KAVA WORK-IN-PROGRESS 245

Personal communication from M McGuffin to M McIn-tyre available as an online document at ehpaglobalnetcouk

fulminant hepatic failure) This case appears tobe unassessable

Case 23 This case involved a 35-year-oldwoman with jaundice According to the BfArM(see Appendix 2) this patient also took parac-etamol but no dosage or details were providedThis case and case 25 in the BfArM listing ap-pear to be the same case Both cases have beenlabeled as possible by some authorities butgiven the lack of information about the dosageof paracetamol and the apparent confusion re-garding cases 23 and 25 we submit that theonly logical assessment is unassessable

Case 2627 This case involved a woman whowas either 38 or 39 yearsrsquo old with hepatitis Itappears that the two cases have been duplicated(Schmidt and Nahrstedt 2002) The confusionwith this case is another example of inaccuratedata provided by the BfArM Information re-garding these cases (or case) depending onwhether the two reports concern the samewoman is unclear Penicillin can be associatedwith hypersensitivity and cholestatic jaundicebut the information given is inadequate to makeany meaningful assessment For this reason weclass this case as unassessable

Case 29 This case involved a 60-year-oldwoman who had a liver transplant This womanwas taking piretamide (which is a loop diuretic)Frusemide another loop diuretic can be associ-ated with cholestatic jaundice According to theBfArM chart (see Appendix 2) she was also tak-ing a sympathomimetic drug etilefrin Thedosage of kava varied but was up to 480ndash1200mg per day (Schmidt and Nahrstedt 2002)which is up to ten times the German Commis-sion E maximum recommended dose (Blumen-thal 1998) Although some authorities have re-garded this case as possible in view of themarked overdosing of kava and the concomitantmedication this case can hardly be said to be areflection on the proper therapeutic use of kava

(5) Cases in which drugs not associated withliver damage herbal medicines or dietarysupplements or kavain alone were taken

This category had eight cases 2 78 11 1322 24 and 25

For these cases detail was limited and theBfArM did not implicate any other drugs ormedications (although this may not be thecase)

All patients in this group apart from the pa-tient in Case 78 for whom no information wasgiven were reported to have made full recov-eries In some of these cases it is not clearwhether the patients were ill or whether thesecases merely recorded raised liver-function en-zymes

Case 2 This case involved a 35-year-old manwith cholestatic hepatitis Concomitant med-ication was ldquounknownrdquo Apart from Cases 18and 30 this is the only case for which it is pos-sible that no other concomitant medication wastaken but there is a marked lack of informationfor this case As noted above this patient wastaking synthetic kavain not kava We regardthis case as unassessable

Case 5 This case involved a woman who waseither 68 or 69 yearsrsquo old with cholestatic he-patitis She was also taking a St Johnrsquos wort(Hypericum perforatum) product which hasbeen associated with CYP3A4 A biopsyshowed ldquoimmunologic hypersensitivityrdquo Thiscase may be regarded as possible but in viewof the immunologic hypersensitivity it maywell have been an idiosyncratic event that wasnot necessarily associated with kava usage

Case 78 This case involved a woman or twowomen ages 72 andor 75 with cholestatic he-patitis These two cases appear to be actuallyone case The woman was taking twoherbalvitamin products one of which in-cluded 06 mg of kavalactones Given the con-fusion involved these ldquocasesrdquo must be re-garded as unassessable

Case 11 This case involved a 59-year-oldwoman who was taking hyoscine butylbro-mide as a suppository Schmidt and Nahrstedt(2002) commented that according to additionalinformation obtained from the BfArM it is un-certain as to whether this patient was taking akava product at all We regard this case asunassessable

DENHAM ET AL246

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

(1) Cases of most concern

This group includes five cases 10 13 16 18and 28 According to the assessments made by various government agencies these casescause the most concern and are often cited asbeing probable For this reason these cases aredealt with first As discussed below mostmdashifnot allmdashof these cases have associated factorsthat put this probable categorization into ques-tion

Case 10 This case described necrotizing hep-atitis in a 39-year old female patient with pos-itive reexposure (Strahl et al 1998) During thefirst period that the kava product was takenan oral contraceptive and paroxetine were con-comitant medications It appears that paroxe-tine was not the only antidepressant taken bythis patient who also occasionally took StJohnrsquos wort (Hypericum perforatum) The Exec-utive Summary issued by the Bundesverbandder Arzneimittel Hersteller eV (BAH) andBundesverband der Pharmazeutischen Indus-trie eV (BP) (see Appendix 3) stated ldquoA causalrelationship with kava cannot be excluded butthe patientrsquos history and a potential preexistingliver damage must be taken into account In ad-dition the kava preparation used was not iden-tified by the physicianrdquo

Moreover in this case taking paroxetine incombination with an oral contraceptive maywell have led to overburdening the liver a sit-uation that could have been exacerbated by tak-ing a kava preparation Schmidt and Nahrstedt(2002) suggested that this case may be associ-ated with an immunologic reaction After re-viewing all of the cases in detail Schmidt andNahrstedt (2002) concluded that this is the onlycase for which there was sufficient informationto make an association with kava appear prob-able and for which the dose of kava also con-formed to that recommended by the Com-mission E monograph (Blumenthal 2000)However as discussed below Russmann et al(2001) tested this patient for CYP2D6 and as inCase 16 found this patient to be CYP2D6 defi-cient which appears to have made her partic-ularly vulnerable to the cocktail of drugs shewas taking Given the complicating features ofthis case we submit that this case should beclassed as possible rather than probable

Case 13 This was a case of a 62-year-oldwoman with jaundice The BfArM table (seeAppendix 2) noted regarding concomitantmedication that there was ldquonone denotedrdquo butit was claimed that concomitant medication didexist but was ldquounknownrdquo The insufficiency ofdata provided for this case was highlighted byBfArMrsquos warning note ldquoNo medical messagerdquoIn addition it should be noted that no detailsof the dosage of kava or period of its adminis-tration were apparently recorded for this caseThis is clearly insufficient information onwhich to base a probable assessment

Case 16 This case concerned a 33-year-oldwoman with jaundice The woman wasrecorded as having taken an overdose of alco-hol measured at 60 g (Russman et al 2001) andthen analgesics including paracetamol fol-lowing this alcohol binge Despite the massiveintake of alcohol a liver biopsy indicated thata drug rather than an alcohol induced toxicgenesis However this case like that of Case 10above was discussed by Russman et al (2001)who demonstrated afterward that this patientwas shown to be CYP2D6 deficient which (asdiscussed below) seems to be a risk factor forthe hepatotoxicity that was ascribed to kavaWe submit that given these circumstances thiscase should be considered possible rather thanprobable

Case 18 This case concerned a 50-year-oldman who had necrosis leading to a liver trans-plant This patient took a product manufac-tured by acetone extraction at a dose deliver-ing 210ndash280 mg of kavalactones per day for 15months ldquomoderate alcoholrdquo (ldquomoderaterdquo is notdefined by BfArM) evening primrose(Oenothera biennis) and a yeast preparationThe dosage of kava was well above the Ger-man Commission E recommended dose ofkavalactones (Blumenthal 1998) It was alsorecorded that 500ndash1000 mg of paracetamol wastaken by this patient shortly before transplan-tation The combination of paracetamol and al-cohol plus the very high dose of kava extractedin acetone taken by this man casts seriousdoubts on the assessment of probable in thiscase

Case 28 (BAH) This case concerned a

KAVA WORK-IN-PROGRESS 243

woman age unknown with hepatitis This caseis hard to assess because neither the patientrsquosage nor diagnosis was given and the womanwas taking eleven medications including estra-diol valerate acetylcysteine losartan (which israrely be associated with hepatitis) and mepra-zole (which can be associated with liver diseasealthough this is rare) Omeprazole is metabo-lized by the polymorphic CYP2C19 which is absent in 3 of Caucasians (Flockhart et al2000) The woman was also taking echinacea(Echinacea purpurea) and five products that ap-peared to be for upper respiratory problems Itshould be noted that this patient was taking syn-thetic kavain not kava A comment from BfArMconcerning this case noted ldquorecurrence of the he-patic side-effectsrdquo which has evidently been in-terpreted by some authorities as being equiva-lent to a ldquopositive rechallengerdquo Whether or notthis was actually so was not clear from the datasupplied It appears (Schmidt and Nahrstedt2002) that Case 28 has been published as twocases with slightly different details This is con-fusing and considering that the woman was tak-ing 11 other medications together with a syn-thetic kava (which we submit is not equivalentto natural kava) and that no diagnosis of hercondition was supplied this calls the assessmentof probable in this case into question

(2) Cases associated with taking synthetic kavain

In this category there were 4 cases 1 2 19and 28

In each of these cases the patients concernedwere taking a product made from synthetickavain Although the outcome was hepatitis inall four cases kavain cannot be equated withthe naturally occurring form of kava whichcontains many other constituents that may playan important role in ensuring the safety of thisherb Therefore we submit that no inferenceshould be drawn from these cases Traditionalusage should not be taken as evidence for safeusage of synthetic products

(3) Patients who were taking oral contraceptivepills or hormone replacement therapy (HRT)together with drugs that can also be associatedwith liver damage

The cases in this category were 4 10 12 2021 and 28

Cholestatic jaundice associated with use ofestrogen-containing medications is extremelyrare (Lindberg 1992) but does occur In these6 cases the women were also taking drugs thatcan also be associated with jaundice

Case 4 This case involved a 39-year-oldwoman with jaundice She was on diazepam10 mg PRN for 6 months Some authoritiescalled this case possible Our assessment is thatthe case is unassessable

Case 10 This case involved a 39-year-oldwoman with necrotizing hepatitis For a de-tailed assessment see above

Case 12 This case involved a 37-year-oldwoman with hepatitis She was on 150 mg ofdiclofenac via intramuscular injection Hepato-toxic reactions associated with nonsteroidalanti-inflammatory drug use are extremely rareand concomitant exposure to other hepatotoxicdrugs is considered to be an important factor(Bareille et al 2001) This case of hepatitis isdifficult to interpret because it occurred inBrazil and because ldquoreexposure was said to benegative for all three drugsrdquo We regard thiscase as unassessable

Case 20 This case involved 50-year-oldwoman with necrosis who had a liver trans-plant She had a 20-year history of combinedoral contraceptive use but had changed monthsearlier to estradiol valerate (which was appar-ently taken alone) as HRT She had also startedglimepiride 8 months earlier This is used fortreating type II diabetes and is rarely associatedwith cholestatic jaundice and liver failure Weregard this case as unlikely

Case 21 This case involved a 22-year-oldwoman with necrosis who had a liver trans-plant This woman had changed from Valette(Jenapharm GmbH Jena Germany) (2 mg ofdienogest and 003 mg of ethinlestradiol) toPramino (180215250 mcg of norgestimateand mcg 25 of ethinylestradiol) She also tookrizatriptan if required for migraine reliefRizatriptan should be used with caution in he-patic impairment and avoided if a patient hassevere liver disease Some authorities considerthis case as being possible but our assessment

DENHAM ET AL244

is in view of the other medications taken isthat this case is unassessable

Case 28 This case involved a woman age un-known with hepatitis This case is discussed atlength above As noted above this patient wastaking synthetic kavain not kava

(4) Patients who were taking drugs that can beassociated with liver damage

There were ten cases in this category 1 6 914 15 17 19 23 2627 and 29

Case 1 This case involved a woman age 69with cholestatic hepatitis She was taking pen-toxifylline (which can be associated with intra-hepatic cholestasis) and a diuretic including thepotassium-sparing triamterene (which can beassociated with jaundice) As noted above thispatient was taking synthetic kavain not kavaWe consider this case unassessable

Case 6 This case involved a woman age 50with hepatitis She was taking frusemide(which can be associated with cholestatic jaun-dice) triamterene atenolol and a large dose ofterfenadine (300 mg) The recommended doseof terfenadine in the British National Formu-lary (March 2001) is 60ndash120 mg The Formularyrecommends avoiding this drug in patientswho have hepatic impairment and also says toldquoavoid concomitant administration of drugs li-able to produce electrolyte imbalance such asdiureticsrdquo (British National Formulary 2001)Despite this warning this woman was also tak-ing the diuretic frusemide The InterkantonalenKontrollstelle der Schweiz of Switzerland con-sidered this case of hepatitis to be caused byterfenadine And although some authoritiesregard this case as possible our assessment isthat this case is unlikely

Case 9 This case involved an 81-year-oldwoman who had liver failure and subsequentdeath She was taking hydrochlorothiazide(which can occasionally be associated with in-trahepatic cholestasis) However according toSchmidt and Nahrstedt (2002) there was evi-dence of chronic alcohol abuse and they re-ported that the autopsy showed chronic pan-creatitis that was characteristic of alcoholabuse The autopsy report (Schmidt and

Nahrstedt 2002) apparently said that thesymptoms must have occurred over a periodof at least 18 months The report conceded thatldquohepatic impairment by alcohol [was] not ex-cludedrdquo In these circumstances it seems en-tirely reasonable to claim that this case is un-related to kava use We regard this case asunlikely

Case 14 This case involved a 33-year-oldwoman with hepatitis Cisapride may havebeen taken (which can cause reversible changesthat show in liver-function tests) Cirrhosis ina woman of 33 is an unexplained finding andthe detail in this case is inadequate to elucidateit We consider this case to be unassessable

Case 15 This case involved a 46-year-oldwoman with jaundice She had been taking hy-drochlorothiazide (which can be associatedwith intrahepatic cholestasis) for 55 monthsplus 80 mg of valsartan and 80 mg ofpropanolol per day Some authorities regardthis case as possible but we consider it to beunassessable

Case 17 This case involved a 59-year-oldwoman with jaundice She had taken 100ndash200mg of celecoxib a cyclo-oxygenase-2 inhibitorper day According to the criteria for causalityassessment of adverse reactions some author-ities consider this case to be possible but our as-sessment is that it is unassessable

Case 19 This case involved a 21-year-oldwoman with hepatitis She was taking panto-prazole (which as with omeprazole can be as-sociated with liver disease) She was also takingparacetamol and metoclopramide and had over-dosed on kavain More detail is needed on othermedical conditions suffered by this patient in or-der to interpret this case It is suggested bySchmidt that this woman was using up to 10tablets per day of the product (the recom-mended dose is up to 6 tablets per day) and thatthere was apparently a discussion in her med-ical record file that she may also have used Ec-stasy (substance that has been associated with

KAVA WORK-IN-PROGRESS 245

Personal communication from M McGuffin to M McIn-tyre available as an online document at ehpaglobalnetcouk

fulminant hepatic failure) This case appears tobe unassessable

Case 23 This case involved a 35-year-oldwoman with jaundice According to the BfArM(see Appendix 2) this patient also took parac-etamol but no dosage or details were providedThis case and case 25 in the BfArM listing ap-pear to be the same case Both cases have beenlabeled as possible by some authorities butgiven the lack of information about the dosageof paracetamol and the apparent confusion re-garding cases 23 and 25 we submit that theonly logical assessment is unassessable

Case 2627 This case involved a woman whowas either 38 or 39 yearsrsquo old with hepatitis Itappears that the two cases have been duplicated(Schmidt and Nahrstedt 2002) The confusionwith this case is another example of inaccuratedata provided by the BfArM Information re-garding these cases (or case) depending onwhether the two reports concern the samewoman is unclear Penicillin can be associatedwith hypersensitivity and cholestatic jaundicebut the information given is inadequate to makeany meaningful assessment For this reason weclass this case as unassessable

Case 29 This case involved a 60-year-oldwoman who had a liver transplant This womanwas taking piretamide (which is a loop diuretic)Frusemide another loop diuretic can be associ-ated with cholestatic jaundice According to theBfArM chart (see Appendix 2) she was also tak-ing a sympathomimetic drug etilefrin Thedosage of kava varied but was up to 480ndash1200mg per day (Schmidt and Nahrstedt 2002)which is up to ten times the German Commis-sion E maximum recommended dose (Blumen-thal 1998) Although some authorities have re-garded this case as possible in view of themarked overdosing of kava and the concomitantmedication this case can hardly be said to be areflection on the proper therapeutic use of kava

(5) Cases in which drugs not associated withliver damage herbal medicines or dietarysupplements or kavain alone were taken

This category had eight cases 2 78 11 1322 24 and 25

For these cases detail was limited and theBfArM did not implicate any other drugs ormedications (although this may not be thecase)

All patients in this group apart from the pa-tient in Case 78 for whom no information wasgiven were reported to have made full recov-eries In some of these cases it is not clearwhether the patients were ill or whether thesecases merely recorded raised liver-function en-zymes

Case 2 This case involved a 35-year-old manwith cholestatic hepatitis Concomitant med-ication was ldquounknownrdquo Apart from Cases 18and 30 this is the only case for which it is pos-sible that no other concomitant medication wastaken but there is a marked lack of informationfor this case As noted above this patient wastaking synthetic kavain not kava We regardthis case as unassessable

Case 5 This case involved a woman who waseither 68 or 69 yearsrsquo old with cholestatic he-patitis She was also taking a St Johnrsquos wort(Hypericum perforatum) product which hasbeen associated with CYP3A4 A biopsyshowed ldquoimmunologic hypersensitivityrdquo Thiscase may be regarded as possible but in viewof the immunologic hypersensitivity it maywell have been an idiosyncratic event that wasnot necessarily associated with kava usage

Case 78 This case involved a woman or twowomen ages 72 andor 75 with cholestatic he-patitis These two cases appear to be actuallyone case The woman was taking twoherbalvitamin products one of which in-cluded 06 mg of kavalactones Given the con-fusion involved these ldquocasesrdquo must be re-garded as unassessable

Case 11 This case involved a 59-year-oldwoman who was taking hyoscine butylbro-mide as a suppository Schmidt and Nahrstedt(2002) commented that according to additionalinformation obtained from the BfArM it is un-certain as to whether this patient was taking akava product at all We regard this case asunassessable

DENHAM ET AL246

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

woman age unknown with hepatitis This caseis hard to assess because neither the patientrsquosage nor diagnosis was given and the womanwas taking eleven medications including estra-diol valerate acetylcysteine losartan (which israrely be associated with hepatitis) and mepra-zole (which can be associated with liver diseasealthough this is rare) Omeprazole is metabo-lized by the polymorphic CYP2C19 which is absent in 3 of Caucasians (Flockhart et al2000) The woman was also taking echinacea(Echinacea purpurea) and five products that ap-peared to be for upper respiratory problems Itshould be noted that this patient was taking syn-thetic kavain not kava A comment from BfArMconcerning this case noted ldquorecurrence of the he-patic side-effectsrdquo which has evidently been in-terpreted by some authorities as being equiva-lent to a ldquopositive rechallengerdquo Whether or notthis was actually so was not clear from the datasupplied It appears (Schmidt and Nahrstedt2002) that Case 28 has been published as twocases with slightly different details This is con-fusing and considering that the woman was tak-ing 11 other medications together with a syn-thetic kava (which we submit is not equivalentto natural kava) and that no diagnosis of hercondition was supplied this calls the assessmentof probable in this case into question

(2) Cases associated with taking synthetic kavain

In this category there were 4 cases 1 2 19and 28

In each of these cases the patients concernedwere taking a product made from synthetickavain Although the outcome was hepatitis inall four cases kavain cannot be equated withthe naturally occurring form of kava whichcontains many other constituents that may playan important role in ensuring the safety of thisherb Therefore we submit that no inferenceshould be drawn from these cases Traditionalusage should not be taken as evidence for safeusage of synthetic products

(3) Patients who were taking oral contraceptivepills or hormone replacement therapy (HRT)together with drugs that can also be associatedwith liver damage

The cases in this category were 4 10 12 2021 and 28

Cholestatic jaundice associated with use ofestrogen-containing medications is extremelyrare (Lindberg 1992) but does occur In these6 cases the women were also taking drugs thatcan also be associated with jaundice

Case 4 This case involved a 39-year-oldwoman with jaundice She was on diazepam10 mg PRN for 6 months Some authoritiescalled this case possible Our assessment is thatthe case is unassessable

Case 10 This case involved a 39-year-oldwoman with necrotizing hepatitis For a de-tailed assessment see above

Case 12 This case involved a 37-year-oldwoman with hepatitis She was on 150 mg ofdiclofenac via intramuscular injection Hepato-toxic reactions associated with nonsteroidalanti-inflammatory drug use are extremely rareand concomitant exposure to other hepatotoxicdrugs is considered to be an important factor(Bareille et al 2001) This case of hepatitis isdifficult to interpret because it occurred inBrazil and because ldquoreexposure was said to benegative for all three drugsrdquo We regard thiscase as unassessable

Case 20 This case involved 50-year-oldwoman with necrosis who had a liver trans-plant She had a 20-year history of combinedoral contraceptive use but had changed monthsearlier to estradiol valerate (which was appar-ently taken alone) as HRT She had also startedglimepiride 8 months earlier This is used fortreating type II diabetes and is rarely associatedwith cholestatic jaundice and liver failure Weregard this case as unlikely

Case 21 This case involved a 22-year-oldwoman with necrosis who had a liver trans-plant This woman had changed from Valette(Jenapharm GmbH Jena Germany) (2 mg ofdienogest and 003 mg of ethinlestradiol) toPramino (180215250 mcg of norgestimateand mcg 25 of ethinylestradiol) She also tookrizatriptan if required for migraine reliefRizatriptan should be used with caution in he-patic impairment and avoided if a patient hassevere liver disease Some authorities considerthis case as being possible but our assessment

DENHAM ET AL244

is in view of the other medications taken isthat this case is unassessable

Case 28 This case involved a woman age un-known with hepatitis This case is discussed atlength above As noted above this patient wastaking synthetic kavain not kava

(4) Patients who were taking drugs that can beassociated with liver damage

There were ten cases in this category 1 6 914 15 17 19 23 2627 and 29

Case 1 This case involved a woman age 69with cholestatic hepatitis She was taking pen-toxifylline (which can be associated with intra-hepatic cholestasis) and a diuretic including thepotassium-sparing triamterene (which can beassociated with jaundice) As noted above thispatient was taking synthetic kavain not kavaWe consider this case unassessable

Case 6 This case involved a woman age 50with hepatitis She was taking frusemide(which can be associated with cholestatic jaun-dice) triamterene atenolol and a large dose ofterfenadine (300 mg) The recommended doseof terfenadine in the British National Formu-lary (March 2001) is 60ndash120 mg The Formularyrecommends avoiding this drug in patientswho have hepatic impairment and also says toldquoavoid concomitant administration of drugs li-able to produce electrolyte imbalance such asdiureticsrdquo (British National Formulary 2001)Despite this warning this woman was also tak-ing the diuretic frusemide The InterkantonalenKontrollstelle der Schweiz of Switzerland con-sidered this case of hepatitis to be caused byterfenadine And although some authoritiesregard this case as possible our assessment isthat this case is unlikely

Case 9 This case involved an 81-year-oldwoman who had liver failure and subsequentdeath She was taking hydrochlorothiazide(which can occasionally be associated with in-trahepatic cholestasis) However according toSchmidt and Nahrstedt (2002) there was evi-dence of chronic alcohol abuse and they re-ported that the autopsy showed chronic pan-creatitis that was characteristic of alcoholabuse The autopsy report (Schmidt and

Nahrstedt 2002) apparently said that thesymptoms must have occurred over a periodof at least 18 months The report conceded thatldquohepatic impairment by alcohol [was] not ex-cludedrdquo In these circumstances it seems en-tirely reasonable to claim that this case is un-related to kava use We regard this case asunlikely

Case 14 This case involved a 33-year-oldwoman with hepatitis Cisapride may havebeen taken (which can cause reversible changesthat show in liver-function tests) Cirrhosis ina woman of 33 is an unexplained finding andthe detail in this case is inadequate to elucidateit We consider this case to be unassessable

Case 15 This case involved a 46-year-oldwoman with jaundice She had been taking hy-drochlorothiazide (which can be associatedwith intrahepatic cholestasis) for 55 monthsplus 80 mg of valsartan and 80 mg ofpropanolol per day Some authorities regardthis case as possible but we consider it to beunassessable

Case 17 This case involved a 59-year-oldwoman with jaundice She had taken 100ndash200mg of celecoxib a cyclo-oxygenase-2 inhibitorper day According to the criteria for causalityassessment of adverse reactions some author-ities consider this case to be possible but our as-sessment is that it is unassessable

Case 19 This case involved a 21-year-oldwoman with hepatitis She was taking panto-prazole (which as with omeprazole can be as-sociated with liver disease) She was also takingparacetamol and metoclopramide and had over-dosed on kavain More detail is needed on othermedical conditions suffered by this patient in or-der to interpret this case It is suggested bySchmidt that this woman was using up to 10tablets per day of the product (the recom-mended dose is up to 6 tablets per day) and thatthere was apparently a discussion in her med-ical record file that she may also have used Ec-stasy (substance that has been associated with

KAVA WORK-IN-PROGRESS 245

Personal communication from M McGuffin to M McIn-tyre available as an online document at ehpaglobalnetcouk

fulminant hepatic failure) This case appears tobe unassessable

Case 23 This case involved a 35-year-oldwoman with jaundice According to the BfArM(see Appendix 2) this patient also took parac-etamol but no dosage or details were providedThis case and case 25 in the BfArM listing ap-pear to be the same case Both cases have beenlabeled as possible by some authorities butgiven the lack of information about the dosageof paracetamol and the apparent confusion re-garding cases 23 and 25 we submit that theonly logical assessment is unassessable

Case 2627 This case involved a woman whowas either 38 or 39 yearsrsquo old with hepatitis Itappears that the two cases have been duplicated(Schmidt and Nahrstedt 2002) The confusionwith this case is another example of inaccuratedata provided by the BfArM Information re-garding these cases (or case) depending onwhether the two reports concern the samewoman is unclear Penicillin can be associatedwith hypersensitivity and cholestatic jaundicebut the information given is inadequate to makeany meaningful assessment For this reason weclass this case as unassessable

Case 29 This case involved a 60-year-oldwoman who had a liver transplant This womanwas taking piretamide (which is a loop diuretic)Frusemide another loop diuretic can be associ-ated with cholestatic jaundice According to theBfArM chart (see Appendix 2) she was also tak-ing a sympathomimetic drug etilefrin Thedosage of kava varied but was up to 480ndash1200mg per day (Schmidt and Nahrstedt 2002)which is up to ten times the German Commis-sion E maximum recommended dose (Blumen-thal 1998) Although some authorities have re-garded this case as possible in view of themarked overdosing of kava and the concomitantmedication this case can hardly be said to be areflection on the proper therapeutic use of kava

(5) Cases in which drugs not associated withliver damage herbal medicines or dietarysupplements or kavain alone were taken

This category had eight cases 2 78 11 1322 24 and 25

For these cases detail was limited and theBfArM did not implicate any other drugs ormedications (although this may not be thecase)

All patients in this group apart from the pa-tient in Case 78 for whom no information wasgiven were reported to have made full recov-eries In some of these cases it is not clearwhether the patients were ill or whether thesecases merely recorded raised liver-function en-zymes

Case 2 This case involved a 35-year-old manwith cholestatic hepatitis Concomitant med-ication was ldquounknownrdquo Apart from Cases 18and 30 this is the only case for which it is pos-sible that no other concomitant medication wastaken but there is a marked lack of informationfor this case As noted above this patient wastaking synthetic kavain not kava We regardthis case as unassessable

Case 5 This case involved a woman who waseither 68 or 69 yearsrsquo old with cholestatic he-patitis She was also taking a St Johnrsquos wort(Hypericum perforatum) product which hasbeen associated with CYP3A4 A biopsyshowed ldquoimmunologic hypersensitivityrdquo Thiscase may be regarded as possible but in viewof the immunologic hypersensitivity it maywell have been an idiosyncratic event that wasnot necessarily associated with kava usage

Case 78 This case involved a woman or twowomen ages 72 andor 75 with cholestatic he-patitis These two cases appear to be actuallyone case The woman was taking twoherbalvitamin products one of which in-cluded 06 mg of kavalactones Given the con-fusion involved these ldquocasesrdquo must be re-garded as unassessable

Case 11 This case involved a 59-year-oldwoman who was taking hyoscine butylbro-mide as a suppository Schmidt and Nahrstedt(2002) commented that according to additionalinformation obtained from the BfArM it is un-certain as to whether this patient was taking akava product at all We regard this case asunassessable

DENHAM ET AL246

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

is in view of the other medications taken isthat this case is unassessable

Case 28 This case involved a woman age un-known with hepatitis This case is discussed atlength above As noted above this patient wastaking synthetic kavain not kava

(4) Patients who were taking drugs that can beassociated with liver damage

There were ten cases in this category 1 6 914 15 17 19 23 2627 and 29

Case 1 This case involved a woman age 69with cholestatic hepatitis She was taking pen-toxifylline (which can be associated with intra-hepatic cholestasis) and a diuretic including thepotassium-sparing triamterene (which can beassociated with jaundice) As noted above thispatient was taking synthetic kavain not kavaWe consider this case unassessable

Case 6 This case involved a woman age 50with hepatitis She was taking frusemide(which can be associated with cholestatic jaun-dice) triamterene atenolol and a large dose ofterfenadine (300 mg) The recommended doseof terfenadine in the British National Formu-lary (March 2001) is 60ndash120 mg The Formularyrecommends avoiding this drug in patientswho have hepatic impairment and also says toldquoavoid concomitant administration of drugs li-able to produce electrolyte imbalance such asdiureticsrdquo (British National Formulary 2001)Despite this warning this woman was also tak-ing the diuretic frusemide The InterkantonalenKontrollstelle der Schweiz of Switzerland con-sidered this case of hepatitis to be caused byterfenadine And although some authoritiesregard this case as possible our assessment isthat this case is unlikely

Case 9 This case involved an 81-year-oldwoman who had liver failure and subsequentdeath She was taking hydrochlorothiazide(which can occasionally be associated with in-trahepatic cholestasis) However according toSchmidt and Nahrstedt (2002) there was evi-dence of chronic alcohol abuse and they re-ported that the autopsy showed chronic pan-creatitis that was characteristic of alcoholabuse The autopsy report (Schmidt and

Nahrstedt 2002) apparently said that thesymptoms must have occurred over a periodof at least 18 months The report conceded thatldquohepatic impairment by alcohol [was] not ex-cludedrdquo In these circumstances it seems en-tirely reasonable to claim that this case is un-related to kava use We regard this case asunlikely

Case 14 This case involved a 33-year-oldwoman with hepatitis Cisapride may havebeen taken (which can cause reversible changesthat show in liver-function tests) Cirrhosis ina woman of 33 is an unexplained finding andthe detail in this case is inadequate to elucidateit We consider this case to be unassessable

Case 15 This case involved a 46-year-oldwoman with jaundice She had been taking hy-drochlorothiazide (which can be associatedwith intrahepatic cholestasis) for 55 monthsplus 80 mg of valsartan and 80 mg ofpropanolol per day Some authorities regardthis case as possible but we consider it to beunassessable

Case 17 This case involved a 59-year-oldwoman with jaundice She had taken 100ndash200mg of celecoxib a cyclo-oxygenase-2 inhibitorper day According to the criteria for causalityassessment of adverse reactions some author-ities consider this case to be possible but our as-sessment is that it is unassessable

Case 19 This case involved a 21-year-oldwoman with hepatitis She was taking panto-prazole (which as with omeprazole can be as-sociated with liver disease) She was also takingparacetamol and metoclopramide and had over-dosed on kavain More detail is needed on othermedical conditions suffered by this patient in or-der to interpret this case It is suggested bySchmidt that this woman was using up to 10tablets per day of the product (the recom-mended dose is up to 6 tablets per day) and thatthere was apparently a discussion in her med-ical record file that she may also have used Ec-stasy (substance that has been associated with

KAVA WORK-IN-PROGRESS 245

Personal communication from M McGuffin to M McIn-tyre available as an online document at ehpaglobalnetcouk

fulminant hepatic failure) This case appears tobe unassessable

Case 23 This case involved a 35-year-oldwoman with jaundice According to the BfArM(see Appendix 2) this patient also took parac-etamol but no dosage or details were providedThis case and case 25 in the BfArM listing ap-pear to be the same case Both cases have beenlabeled as possible by some authorities butgiven the lack of information about the dosageof paracetamol and the apparent confusion re-garding cases 23 and 25 we submit that theonly logical assessment is unassessable

Case 2627 This case involved a woman whowas either 38 or 39 yearsrsquo old with hepatitis Itappears that the two cases have been duplicated(Schmidt and Nahrstedt 2002) The confusionwith this case is another example of inaccuratedata provided by the BfArM Information re-garding these cases (or case) depending onwhether the two reports concern the samewoman is unclear Penicillin can be associatedwith hypersensitivity and cholestatic jaundicebut the information given is inadequate to makeany meaningful assessment For this reason weclass this case as unassessable

Case 29 This case involved a 60-year-oldwoman who had a liver transplant This womanwas taking piretamide (which is a loop diuretic)Frusemide another loop diuretic can be associ-ated with cholestatic jaundice According to theBfArM chart (see Appendix 2) she was also tak-ing a sympathomimetic drug etilefrin Thedosage of kava varied but was up to 480ndash1200mg per day (Schmidt and Nahrstedt 2002)which is up to ten times the German Commis-sion E maximum recommended dose (Blumen-thal 1998) Although some authorities have re-garded this case as possible in view of themarked overdosing of kava and the concomitantmedication this case can hardly be said to be areflection on the proper therapeutic use of kava

(5) Cases in which drugs not associated withliver damage herbal medicines or dietarysupplements or kavain alone were taken

This category had eight cases 2 78 11 1322 24 and 25

For these cases detail was limited and theBfArM did not implicate any other drugs ormedications (although this may not be thecase)

All patients in this group apart from the pa-tient in Case 78 for whom no information wasgiven were reported to have made full recov-eries In some of these cases it is not clearwhether the patients were ill or whether thesecases merely recorded raised liver-function en-zymes

Case 2 This case involved a 35-year-old manwith cholestatic hepatitis Concomitant med-ication was ldquounknownrdquo Apart from Cases 18and 30 this is the only case for which it is pos-sible that no other concomitant medication wastaken but there is a marked lack of informationfor this case As noted above this patient wastaking synthetic kavain not kava We regardthis case as unassessable

Case 5 This case involved a woman who waseither 68 or 69 yearsrsquo old with cholestatic he-patitis She was also taking a St Johnrsquos wort(Hypericum perforatum) product which hasbeen associated with CYP3A4 A biopsyshowed ldquoimmunologic hypersensitivityrdquo Thiscase may be regarded as possible but in viewof the immunologic hypersensitivity it maywell have been an idiosyncratic event that wasnot necessarily associated with kava usage

Case 78 This case involved a woman or twowomen ages 72 andor 75 with cholestatic he-patitis These two cases appear to be actuallyone case The woman was taking twoherbalvitamin products one of which in-cluded 06 mg of kavalactones Given the con-fusion involved these ldquocasesrdquo must be re-garded as unassessable

Case 11 This case involved a 59-year-oldwoman who was taking hyoscine butylbro-mide as a suppository Schmidt and Nahrstedt(2002) commented that according to additionalinformation obtained from the BfArM it is un-certain as to whether this patient was taking akava product at all We regard this case asunassessable

DENHAM ET AL246

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

fulminant hepatic failure) This case appears tobe unassessable

Case 23 This case involved a 35-year-oldwoman with jaundice According to the BfArM(see Appendix 2) this patient also took parac-etamol but no dosage or details were providedThis case and case 25 in the BfArM listing ap-pear to be the same case Both cases have beenlabeled as possible by some authorities butgiven the lack of information about the dosageof paracetamol and the apparent confusion re-garding cases 23 and 25 we submit that theonly logical assessment is unassessable

Case 2627 This case involved a woman whowas either 38 or 39 yearsrsquo old with hepatitis Itappears that the two cases have been duplicated(Schmidt and Nahrstedt 2002) The confusionwith this case is another example of inaccuratedata provided by the BfArM Information re-garding these cases (or case) depending onwhether the two reports concern the samewoman is unclear Penicillin can be associatedwith hypersensitivity and cholestatic jaundicebut the information given is inadequate to makeany meaningful assessment For this reason weclass this case as unassessable

Case 29 This case involved a 60-year-oldwoman who had a liver transplant This womanwas taking piretamide (which is a loop diuretic)Frusemide another loop diuretic can be associ-ated with cholestatic jaundice According to theBfArM chart (see Appendix 2) she was also tak-ing a sympathomimetic drug etilefrin Thedosage of kava varied but was up to 480ndash1200mg per day (Schmidt and Nahrstedt 2002)which is up to ten times the German Commis-sion E maximum recommended dose (Blumen-thal 1998) Although some authorities have re-garded this case as possible in view of themarked overdosing of kava and the concomitantmedication this case can hardly be said to be areflection on the proper therapeutic use of kava

(5) Cases in which drugs not associated withliver damage herbal medicines or dietarysupplements or kavain alone were taken

This category had eight cases 2 78 11 1322 24 and 25

For these cases detail was limited and theBfArM did not implicate any other drugs ormedications (although this may not be thecase)

All patients in this group apart from the pa-tient in Case 78 for whom no information wasgiven were reported to have made full recov-eries In some of these cases it is not clearwhether the patients were ill or whether thesecases merely recorded raised liver-function en-zymes

Case 2 This case involved a 35-year-old manwith cholestatic hepatitis Concomitant med-ication was ldquounknownrdquo Apart from Cases 18and 30 this is the only case for which it is pos-sible that no other concomitant medication wastaken but there is a marked lack of informationfor this case As noted above this patient wastaking synthetic kavain not kava We regardthis case as unassessable

Case 5 This case involved a woman who waseither 68 or 69 yearsrsquo old with cholestatic he-patitis She was also taking a St Johnrsquos wort(Hypericum perforatum) product which hasbeen associated with CYP3A4 A biopsyshowed ldquoimmunologic hypersensitivityrdquo Thiscase may be regarded as possible but in viewof the immunologic hypersensitivity it maywell have been an idiosyncratic event that wasnot necessarily associated with kava usage

Case 78 This case involved a woman or twowomen ages 72 andor 75 with cholestatic he-patitis These two cases appear to be actuallyone case The woman was taking twoherbalvitamin products one of which in-cluded 06 mg of kavalactones Given the con-fusion involved these ldquocasesrdquo must be re-garded as unassessable

Case 11 This case involved a 59-year-oldwoman who was taking hyoscine butylbro-mide as a suppository Schmidt and Nahrstedt(2002) commented that according to additionalinformation obtained from the BfArM it is un-certain as to whether this patient was taking akava product at all We regard this case asunassessable

DENHAM ET AL246

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

Case 13 This case involved a 62-year-oldwoman with jaundice See above for the dis-cussion of this case It does appear that therewas concomitant medication but no details ofthis or of the kava dosage are available Thismakes interpretation impossible consequentlywe regard this case as unassessable

Case 22 This case involved a 34-year-oldwoman with hepatitis She was taking L-thy-roxine No information is available on her vi-ral serology differential diagnosis or alcoholintake We regard this case as unassessable

Case 24 This case involved a 47-year-oldwoman who had raised liver-function asshown on a test She had a high intake of fish-oil The report stated that this patientrsquos liver en-zymes returned to normal when she stoppedtaking fish oils but again the detail is insuffi-cient However this case appears to supportthe safe use of kava because report stated thatthe patient was ldquorestored to health after dis-continuation of the concomitant medicationand continuation of the (kava) medicationrdquo Weconsider this case to be unlikely

Case 25 This case involved a 34-year-oldwoman with hepatitis According to the infor-mation provided by the BfArM this womanwas just taking Hypericum perforatum concomi-tantly There is confusion about whether this isthe same case as Case 23 and that as recordedby BfArM (see Appendix 2) paracetamol wasindeed a concomitant medicine This case mustbe classed as unlikely

(6) Cases associated with an overdose of alcohol

This group included two cases 16 and 9

Case 16 This case involved a 33-year-oldwoman with jaundice This case is discussed atlength above because some authorities regardthis case as being probable The woman took anoverdose of alcohol (recorded as 60 g) Thiscase was described in detail by Russman et al(2001) because the woman was deficient in CYP2D6 which as previously noted may havemade her vulnerable to the mixture of kava al-cohol and paracetamol (which were taken for

hangover symptoms) In these circumstancesas stated above this case is unlikely to be prob-able We believe it to be possible

Case 9 This case is discussed in subsection 4above

(7) Cases not associated with other drug usage

This group included two cases 18 and 30These final two cases involved men both of

whom required liver transplants and both ofwhom appeared not to have been taking othermedications For these two cases more detailson the medical histories is required for properassessment

Case 18 This case involved a 50-year-old manwith liver necrosis and who had a liver trans-plant This case is discussed in some detailabove The man took an 210ndash280 mg of an ace-tone preparation per day for 15 months Healso had a ldquomoderate alcoholrdquo intake and tooka yeast preparation This is above the recom-mended dose of kavalactones He may alsohave taken paracetamol (see above) This caseis unassessable

Case 30 This case involved a 32-year-old manwith necrosis of the liver and who had a livertransplant He took a product containing 240mg of kavalactones per day for 3 months andoccasionally a valerian (Valeriana officinalis)product at night This too was above the rec-ommended dose of kavalactones This case can-not be evaluated fully because of lack of de-tailed documentation regarding the manrsquosmedical history or the presenting disease andso must be categorized as unassessable

CYTOCHROME p450 METABOLISM OF XENOBIOTICS AND CYP2D6 DEFICIENCY

In most of these cases the patients were alsotaking drugs concomitantly Assuming that themedications were responsible for the adverseevents and not some other factors such as otherdisease or excessive use of alcohol it is possi-ble that the hepatotoxicity was caused by the

KAVA WORK-IN-PROGRESS 247

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

conventional drugs by the kava by both thedrugs and the kava or mainly by the drugs withthe kava as a cofactor However in assessingthese cases one should take into account theapparent increased risk of adverse effects on theliver where kavalactone concentration is en-hanced in a product In all cases cited by theBfArM the affected patients appear to havebeen taking concentrated standardized prod-ucts which in no way relates to the tradi-tional water-based or low-alcohol extracts thathave not been associated with comparable ad-verse events In any case upon analysis of allrelevant factors the number of cases cited bythe BfArM that can actually be attributed tokava is so low that the only logical conclusionthat can be drawn is that kava has a low levelof incidence of adverse events InterestinglySchmidt and Nahrstedt (2002) came to muchthe same conclusion stating that the relativeincidence of adverse events is a fraction of thatof others connected with anxiolytics such asbenzodiazepines

Interindividual variability in cytochrome-p450metabolism of xenobiotics

Kava may be regarded as a possible cofactorin some of these cases but variable individualresponses (interindividual variability) to drugsor herbs should also be taken into account inthese cases Interindividual variability in drugresponse is now increasingly recognized as amajor cause of adverse drug reactions Muchof this variability is now ascribed to genetic dif-ferences in drug absorption disposition me-tabolism or excretion The variability that hasbeen most investigated and that is consideredto be of most significance is genetic polymor-phism in drug metabolizing enzymes in thehepatocyte This is considered to be an adap-tive response to environmental challenge (Wolfand Smith 1999) so it is not in itself surprisingthat individuals vary and failure to metabolizexenobiotics (ldquoforeignrdquo compounds whetherthese be natural or synthetic) is associated withusing medicines from natural or syntheticsources

Cytochrome p450 (CYP) enzymes are mixedfunction microsomal mono-oxygenases that arelocated on the smooth endoplasmic reticulum

throughout the body primarily in hepatocytesand in the wall of the small intestine There are12 families and a single hepatocyte can containa range of CYP enzymes that metabolize arange of drugs These CYP enzymes are re-sponsible for phase I (oxidation reduction andhydrolysis) metabolism of a wide number ofcompounds and for transforming lipophilicdrugs into more polar compounds that can beexcreted by the kidneys

Phase II of detoxification occurs if a productconjugates in the hepatocyte cytoplasm withthe tripeptide glutathione The resulting solu-ble compound is excreted via the bile or theurine This conjugation is catalyzed by cyto-plasmic glutathione S-transferases Interindi-vidual variations exist in the concentration of hepatocyte glutathione and in the relative con-centration of individual glutathione S-trans-ferases (Mannervik and Widdersten 1995) andin levels of other compounds that are associ-ated with drug metabolism

CYP2D6 deficiency

Many CYP enzymes are genetically polymor-phic and thus there is marked interindividualvariation in drug metabolism (Wolf and Smith1999) CYP2D6 is one of the most extensivelystudied genetic polymorphisms It is thought tocause much of the individual variations seen indrug responses side-effects and drug interac-tions (Poolsup et al 2000) Individuals may bepoor (slow) metabolizers intermediate exten-sive (fast) or ultrafast metabolizers In a Cau-casian population 7ndash9 of individuals are ho-mozygous deficient in CYP2D6 and are thuspoor metabolizers (Poolsup et al 2000) The in-cidence of CYP2D6 deficiency in Asian popula-tions is 1 and it is thought that much ethnicvariation in drug response is associated withCYP polymorphism (Poolsup et al 2000) Drugsubstrates for CYP2D6 include antidepressantsantipsychotics beta-blockers (eg propanololand antiarrythmics) and several antidepres-sants (Fromm et al 1997) A poor metabolizeris at risk of having adverse reactions if his or herrate of biotransformation is inadequate

If xenobiotics are inadequately metabolizedthey may make covalent bonds with DNA RNAnuclear proteins or cytoplasmic proteins and

DENHAM ET AL248

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

breakdown of function occurs within these cellsWhen this breakdown is above a certain rate theresult of this is damage to the hepatocyte lead-ing to centrilobular necrosis (Kaplowitz 1997)

As noted above Russmann et al (2001) dis-cussed Case 16 in detail It is noteworthy thatthe woman had restarted kava for 3 weeks af-ter an initial course of treatment 2 months ear-lier and then became ill 3 weeks later after anoverdose of alcohol The woman was shown tobe CYP2D6-deficient using phenotyping withdebrisoquine The researchers then tested thepatient who was delineated as Case 10 whichwas described by Strahl et al (1998) and foundthat she was also CYP2D6-deficient Strahl et al(1998) argued that CYP2D6 deficiency is a riskfactor for hepatotoxicity that is ascribed to kava

This finding may help to explain the lack ofhepatotoxicity as a result of kava beingrecorded in the South Pacific Wanwirolmuk etal (1998) tested the phenotypes of 100 personsof pure Polynesian descent using a debriso-quine probe and found a 0 incidence ofCYP2D6 deficiency The researchers proposedthat with regard to this factor Polynesiansstrongly resemble Asian populations

As stated many antidepressants are metab-olized by CYP2D6 and it is likely that using an-tidepressants with kava is not uncommon Yetonly one of the above cases involved antide-pressants which suggests that CYP2D6 defi-ciency is more likely to be relevant than com-petition between CYP2D6 substrates

This finding is significant but difficult to pre-dict because most people are unaware of theirCYP2D6 phenotype It should be noted thatwhen CYP2D6 deficiency occurs use of kavaproducts with enhanced kavalactones mighthave implications for the affecting the liver par-ticularly when a concomitant orthodox medi-cine or substantial amounts of alcohol are takenregularly It is proposed that such risks are likelyto be small if low-alcohol tinctures are usedwithin the normal therapeutic dosage range

RECOMMENDATIONS FROM TMEC

TMEC recommends that

(1) Products made from synthetic kavain are

synthetic drugs not herbal-medicinal prod-ucts and should be excluded from theanalysis

(2) None of the cases cited by the BfArM in-volved traditionally prepared tinctures Inthe light of evidence presented above and byWhitton et al (Appendix 1) the safety ofconcentrated standardized products madefrom acetone extracts and high-alcohol con-centrations needs reevaluation Low-alcoholtinctures appear to provide a safe alterna-tive TMEC recommends adopting extrac-tion methods that use 25 alcohol to ensurethat the full spectrum of constituents is ex-tracted resulting in a substantially lowerconcentration of kavalactones thus ensur-ing kavarsquos safe use as a medicine

(3) Consumers need to be informed that kavaproducts should not be taken together withconventional medicines without the adviceof a health professional Even more impor-tantly consumers need to know that kavashould not be taken without consulting ahealth professional if users have estab-lished histories of liver disease

(4) Maximum doses for kava should be set af-ter consultation with interested parties

(5) Doctors nurses pharmacists and otherhealth professionals should be adequatelyinformed about herbal medicines and pos-sible herbndashdrug interactions (Jobst et al2000)

SUMMARY

The Executive Summary issued by two Ger-man pharmaceutical associationsmdashBundesver-band der ArzneimittelndashHersteller e V (BAH)and Bundesverband der Pharmazeutischen In-dustrie eV (BPI) (see Appendix 3)mdashof theirsubmission to the BfArM concerning kavastated that the causality in most of the reportsis unclear because details such as additionalmedication patient history and consumptionof alcohol are not given ldquothus not permitting asound evaluation of these casesrdquo Schmidt andNahrstedt (2002) noted that a number of thecases have been reported in the literature morethan once with different data including asnoted above case 28 and in particular that

KAVA WORK-IN-PROGRESS 249

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

cases 7 and 8 are the same as are cases 26 and27 The details of the case reports given are alsoat variance with regard to case 4 in which a dif-ferent time before onset is given and inconsis-tencies also occur in cases 23 and 25 in whichthe time before the onset of the two cases is re-versed These discrepancies are unsatisfactoryand undermine confidence in the accuracy andveracity of the information provided by theBfArM It has also been claimed (Schmidt andNahrstedt 2002) that the case reports are notpresented in accordance with European Unionguidelines for adverse-event reports

The BfArM document is deficient in other re-spects too It is unclear whether the term ldquoliverdamagerdquo refers to the results of a liver biopsyor to the finding of raised alanine aminotrans-ferase (ALT) blood levels that are interpretedas indicating damage to hepatocytes in hepa-tocellular disease (Pagana and Pagana 1999)However in light of the need for the UK Com-mittee on Safety of Medicines to make an in-formed decision on these cases this paper en-deavors to interpret the evidence as presentedUnless noted otherwise references to possibledrug-induced hepatoxocity are taken from theBritish National Formulary (BNF) (March 2001)

ACKNOWLEDGMENTS

Thanks to Angela Grunwald for translatinga number of German texts that provided valu-able background for this paper

REFERENCES

Aalbersberg B Kava boom hits the Pacific Med PlantConserv 1999520

Anonymous British Herbal Pharmacopoeia KeighleyUnited Kingdom British Herbal Medicine Association1983

Anonymous Kava only on prescription That would be aloss [editorial in German] Artzte Zeitung 20012012

Bareille M Montastruc J Lapeyre-Mestre M Liver dam-age and NSAID Casendashnon-case study in the FrenchPharmacovigilance Database Therapie 200156(1)51ndash55

Barnes J Anderson L Phillipson J St Johnrsquos wort (Hy-pericum perforatum L) A review of its chemistry phar-macology and clinical properties J Pharm Pharmacol200153(5)583ndash600

Blumenthal M ed The Complete German CommissionE Monographs Austin American Botanical Council1998

Blumenthal M ed Herbal Medicine Revised and Ex-panded Commission E Monographs Austin AmericanBotanical Council 2000

British Medical Association and Royal PharmaceuticalAssociation British National Formulary London Phar-maceutical Press March 2001

Chanwai L Kava toxicity Emerg Med 20002142ndash145Clough A Burns C Mununggurr N Kava in Arnhem

Land A review of consumption and its social corre-lates Drugs Alcohol Rev 200019(3)319ndash323

De Leo V la Marca A Morgante G Lanzetta D Florio PPetraglia F Evaluation of combining kava extract withhormone replacement therapy in the treatment of post-menopausal anxiety Maturitas (Eur Menopause J)200139185ndash188

Edwards I Aronson J Adverse drug reactions Defini-tions diagnosis and management Lancet 20003561255ndash1259

Ellingwood F American Materia Medica Therapeuticsand Pharmacognosy [reprint] Portland OR EclecticMedical Publications 1919

Felter H Lloyd J Kingrsquos American Dispensatory[reprinted 1983] Portland OR Eclectic Medical Publi-cations 1905

Flockhart D Desta Z Mahal S Selection of drugs to treatgastro-oesophageal reflux disease The role of drug in-teractions Clin Pharmacokinet 200039(4)295ndash309

Fromm M Kroemer H Eichelbaum M Impact of p450genetic polymorphism on the first-pass extraction ofcardiovascular and neuroactive drugs Adv Drg DelRev 199727171ndash199

Green R Sites with Lapita pottery Importing and voy-aging Mankind 19749253ndash259

Greenwood-Robinson M Kava New York Dell Publish-ing 1999

Haberlein H Boonen G Beck M-A Piper methysticumEnantiomeric separation of kavapyrones by HPLCPlanta Medica 19976363ndash65

Hansel R Kava-Kava in modern medical practice [in Ger-man] Zeitschrift fuumlr Phytotherapie 199617180ndash195

He X Lin L Lian L Electrospray HPLC-MS in phyto-chemical analysis of kava (Piper methysticum) extractPlanta Medica 19976370ndash74

Hodgson E Levi PE Textbook of Modern ToxicologyStamford CT Appleton amp Lange 1997

Jobst K McIntyre M St George D Whitelegg M Safetyof St Johnrsquos wort (Hypericum perforatum) Lancet 20009203 575

Johnson T CRC Ethnobotany Desk Reference Boca Ra-ton FL CRC Press 1999

Kaplowitz N Hepatotoxicity of herbal remedies Insightsinto the intricacies of plantndashanimal warfare and celldeath Gastroenterology 1997113(4)1408ndash1412

Kubatova A Miller D Hawthorne S Comparison of sub-critical water and organic solvents for extractingkavalactones from kava root J Chromatog A 2001923187ndash194

DENHAM ET AL250

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

Larrey D Hepatotoxicity of herbal remedies J Hepatol199726(suppl1)47ndash51

Lebot V Merlin M Lindstrom L Kavamdashthe Pacific ElixirVT Healing Arts Press 1997

Lebot V Levesque J The origin and distribution of kava(Piper methysticum Forstf and Piper wichmanii C DCPiperaceae) A phytochemical approach Allertonia19895 223ndash280

Lewin L On Piper methysticum kava Archives de Med-icine et de Pharmacie Navale 188646210ndash220

Lindberg M Hepatobiliary complications of oral contra-ceptives J Gen Int Med 19927(2)199ndash209

Loew von D Kava-kava-extract Deutsche ApothekerZeitung 2002142(9)1012ndash1020

Mannervik B Widersten M Human glutathione trans-ferases Classification tissue distribution structure andfunctional properties In Pacifici G Fracchia G edsAdvances in Drug Metabolism in Man Brussels Euro-pean Commission 1995

McIntyre M A review of the benefits adverse eventsdrug interactions and safety of St Johnrsquos wort (Hyper-icum perforatum) J Altern Complement Med 20006(2)115ndash124

Mills S Bone K Principles and Practice of PhytotherapyEdinburgh Churchill Livingstone 2000

Murray W Neoliberal globalisation ldquoExoticrdquo agro-exportsand local change in the islands A study of the Fijian kavasector Singapore J Trop Geog 200021(3)355ndash373

Pagana K Pagana T Mosbyrsquos Diagnostic and LaboratoryTest Reference St Louis CV Mosby 1999

Pittler M Ernst E Efficacy of kava extract for treating anx-iety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Poolsup N Po L Knight T Pharmacogenetics and psy-chopharmacotherapy J Clin Pharm Ther 200025197ndash220

Russmann S Lauterburg B Helbling A Kava hepatotox-icity Ann Int Med 2001135(1)68ndash69

Ruse P Kava-induced dermopathy A niacin deficiencyLancet 19903351442ndash1445

Schmidt M Nahrstedt A Is kava hepatotoxic [in Ger-

man] Deutsche Apotheker Zeitung 2002142(9)1006ndash1011

Schulz V Rational Phytotherapy Heidelberg Springer-Verlag 1997

Spinella M The Psychopharmacology of Herbal Medi-cine Massachusetts MIT Press 2001

Strahl S Ehret V Dahm H Maier K Necrotizing he-patitis after taking a plant medicine Deutscher Medi-zinwochenschrift 19981231410ndash1414

Wanwirolmuk S Bhawan S Coville P Chalcroft S Ge-netic polymorphism of debrisoquine (CYP2D6) andproguanil (CYP2C19) in South Pacific Polynesian pop-ulations Eur J Clin Pharmacol 199854431ndash435

Williamson E Synergy and other interactions in phy-tomedicines Phytomedicine 20018(5)401ndash409

Wolf C Smith S Pharmacogenetics Br Med Bull 199955(2)366ndash386

BIBLIOGRAPHY

Andersson T Pharmacokinetics metabolism and interac-tions of acid pump Inhibitors Focus on omeprazolelansoprazole and pantoprazole Clin Pharmacokinet199631(1) 9ndash28

Kircheiner J Brosen K Dahl M Gram L Kasper S RootsI Sjoqvist F Spina E Brockmuller J CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants Acta Psychiatrica Scand 2001104173ndash192

Address reprint requests toAlison Denham BA (Soc) MNIMH

University of Central LancashirePreston PR1 2HEUnited Kingdom

E-mail adenhamuclanacuk

KAVA WORK-IN-PROGRESS 251

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

APPENDIX 1

Response to Reported Hepatotoxicity of High Lactone Extractions of Piper methysticum Forst (Kava)

PETER WHITTON BSc1 JULIE WHITEHOUSE PhD2and CHRISTINE EVANS BSc PhD3

Introduction

This paper (the result of work currently in progress) was produced in response to the reportsby the German BfArM of possible hepatotoxic effects of kava (Piper methysticum Forst) extractsthat has led to concerns regarding the safety of kava products on sale in the United KingdomThere have been thirty cases of hepatotoxicity reported to German and Swiss regulators includingthree transplants and one death allegedly associated with the use of concentrated standardizedkava extracts

In the Oceanic Islands of the South Pacific kava is drunk as an alternative to alcohol or forceremonial purposes and studies have shown in islanders who regularly drink up to ten timesthe recommended therapeutic dose that the only recorded abnormality is a slightly raisedgamma-glutamyltransferase (Barguil 2001)

The analysis presented in this paper based on as-yet-unpublished research by the authors demon-strates the presence of glutathione in the traditional extract which it is postulated may have a he-patoprotective effect Concentrated standardized extracts do not contain glutathione (see below)

Extraction Techniques

In the Oceanic Islands kava is traditionally prepared by macerating the root or root bark ina cold water andor coconut milk solution However in the manufacture of concentrated ex-tracts either ethanol (60 and above) or acetone (60 or above) is used as a solvent to obtainthe maximum yield of kavalactones that have been identified as the ldquoactive constituentrdquo

Research Data (The Result of Work in Progress)

Analysis of kava extraction in different solvents

Kava root was extracted in different solvents and analyzed by high performance liquid chro-matography (HPLC) with diode-array detection Different solvents were used to extract thekavalactones and their extraction is shown in Table 1

The extraction was carried out by reflux percolation for 1 hour for each sample of 5 wvPiper methysticum root The resulting liquids then had their specific gravity and percentage ofdry extract determined by the techniques described in the British Pharmacopoeia (1999) Thetotal kavalactones were measured by HPLC using an acetonitrilewater solvent gradient (Whit-ton 2001)

DENHAM ET AL252

1(Student) School of Biosciences University of Westminster London United Kingdom2University of Westminster London United Kingdom3School of Biosciences University of Westminster London United KingdomPersonal communication from Lamberts Ltd Nottingham United Kingdom

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

Kavalactone standardized extracts are produced using a high acetone or ethanol concentra-tion and are likely to contain only kavalactones and no proteins amino acids or sugars

Further analysis identified one of the other compounds in the aqueous extract and in the 25ethanol extract as glutathione by comparison to a reference sample obtained from Sigma-Aldrich(Poole Dorset United Kingdom)

Samples of commercially available kava extracts were examined and the ratio of kavalactonesto glutathione was calculated the results are shown below in Table 2 and Figure 1

Importance of Glutathione in Kava Extracts

Kavalactones may cause hepatic stress if not mediated by glutathione and are usually me-tabolized in the liver by enzymes called lactone hydrolases (Schmidt et al 1999) It can also bedemonstrated in vitro that kavalactones combine with glutathione in a pH dependant reactionby placing a mixture of kavalactones and glutathione in a test tube and adding 01M of sodiumhydroxide to adjust the pH to between 7 and 10 In the control solution of kavalactones alonein this solution no change occurs The same process is observed when cysteine is used insteadof glutathione This reaction is possibly nonreversible and causes the decolourization of the so-lution This point is important as it shows that the lactone ring structures have been opened andthus changed into other as-yet-unknown compounds The opening of the lactone ring rendersthe complex water-soluble and so it can be absorbed into the gut This may bypass the phase Ienzymatic detoxification pathway in the liver thus causing no depletion of intracellular glu-tathione in the hepatocyte The pH range of this reaction renders it liable to occur in the duo-denum and so most probably occurs in vivo between the kavalactone and the cysteine moiety ofthe glutathione molecule after the glutathione has been broken down to its constituent aminoacids by the gastric enzymes

It could be that the high concentration of kavalactones introduced by concentrated standard-ized extracts has the potential to saturate the enzymatic detoxification pathways resulting in un-due stress on the liver Glutathione has an essential role in the phase II conversion of kavalac-tones into excretable waste products and thus gluathione is relevant in excess dosage of

TABLE 1 EXTRACTION OF KAVALACTONES IN DIFFERENT SOLVENTS SUMMARY OF

RESULTS FOR TEN SAMPLES IN EACH SOLVENT

Extract Kavalactones in dried extract

Acetone extract 10096 Ethanol extract 10025 Ethanol 15Water 297

TABLE 2 KAVALACTONEGLUTATHIONE RATIOS

(RESULTS SUMMARIZED FROM TEN SAMPLES OF EACH TYPE)

Kavalactone content Glutathione content Kavalactoneglutathione Sample (expressed as absorbance) (expressed as absorbance) ratio

Kava standardised extract powder 4031712e6 1346769e3 100003(30 kavalactones) dissolved in 25 ethanol

82 Ethanol extraction 3168906e5 53813e3 1001725 Ethanol extraction (1 part plant 163689e4 188736e4 1115

to 3 parts solvent)25 ethanol extraction (1 part plant 249798e4 51525e4 122

to 1 part solvent)

e napierian logarithm

KAVA WORK-IN-PROGRESS 253

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

kavalactones Glutathione is not soluble in ethanol concentrations above 50 (Merck Index 1996)There has been relevant work on a related group of compounds sesquiterpene lactones It hasbeen demonstrated that sesquiterpene lactones react with the sulfide group on the glutathione mol-ecule in a reversible pH dependent reaction (Schmidt et al 1999) The binding of the sesquiter-pene lactones to the glutathione molecule allows for faster clearance by the lactone hydrolases pre-sent in the hepatocytes (Schmidt et al 2001) It has been demonstrated that oral glutathioneprevents toxicity from sesquiterpene lactones if administered at the same time (Lautermann etal1995) It has also been documented that oral glutathione has to be taken at the time of inges-tion of the kavalactones in order to potentiate the metabolism of the kavalactones

We have shown using Acanthamoeba that when kavalactones are added to the growth mediumthe organisms die rapidly However when the same experiment is carried out using a 5050 mix-ture of kavalactones and glutathione no cell death occurs It was found that no cell death oc-curred in concentrations of the glutathionekavalactone mixture of 50 mgmL whereas cell deathoccurred using kavalactones alone at concentrations of less than 1 mgmL Using photomi-croscopy cell death was assessed via visual criteria of cell movement and cell morphology It isplanned to repeat this procedure using hepatocyte cultures but as the phase I and phase II path-ways are common to most life forms it is considered that these results could show that glu-tathione is indeed required for the metabolism of kavalactones

Glutathione is present in adequate amounts in most cells in the body but some individualscan have a genetic deficiency (Lomaestro and Malone 1995) In these cases high doses of kavalac-tones will lead to rapid depletion in glutathione levels and result in free lactone exposure in thehepatocytes and consequent tissue damage (Zheng et al 2000) Glutathione supplementation(taken orally) has been shown to correct the deficiency (Kidd 1997) It is suggested that the glu-tathione molecule may not be absorbed intact but may be broken down into its constituent aminoacids and regenerated within the hepatocyte

It can be postulated that as the reaction occurs in vitro without the presence of enzymes thebinding of the sulfhydral group of common to the glutathione and the cysteine moiety to thekavalactone may take place in the duodenum before absorption This would allow the same pHeffect as has been seen in vitro and allow the Michael type reaction (Merck Index 1996) to oc-cur It has been demonstrated in vitro (in original research undertaken by the authors) that theaddition of either glutathione or cysteine to kavalactones at a pH between 68 and 100 in anaqueous environment leads to the solubility of the kavalactones with decolourization of the so-lution when compared to the same process without the cysteine or glutathione

DENHAM ET AL254

100

80

60

40

20

096 82 45 25

Kavalactones

Glutathione

FIG 1 Kavalactone and glutathione extraction (expressed as a percentage of dry extract) against ethanol percentagein solvent

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

KAVA WORK-IN-PROGRESS 255

The decolourization strongly suggests that the lactone ring has been opened in this reactionthus making the resultant compound water-soluble This means that this reaction which takesplace without the presence of enzymes can occur in the duodenum This would lead to the ab-sorption of the complex of cysteineglutathione and kavalactone into the hepatocyte withoutputting stress on the phase I pathway and so no depletion of intracellular glutathione wouldtake place This suggests that the liver was able to cope with oxidative stress from other sourceswith no interference from the kava

Previous experiments have been conducted with oral glutathione and acetaminophen (parac-etamol) where no non-enzymatic pathway has been observed These findings are readily ex-plained by the different structural formulae of these compounds (Fig 2)

It can be demonstrated that that the cysteine moiety of the glutathione molecule binds via theMichael reaction to the oxygen atom in the lactone ring with the kavalactones This opens thering and leaves the double-bonded oxygen unit intact As mentioned previously the resultingconjugate is water soluble because of the presence of the thiol group from the cysteine In thecase of acetaminophen (paracetamol) this reaction cannot take place without the presence of thephase I enzymes as the molecule is cleaved at the amine (nitrogen) group in alkaline conditions(as the authors have demonstrated) This is quite possibly the reason why large doses (ten tofifty times the therapeutic dose of 60ndash120 mg per day) of the traditional kava extract have beenshown not to cause hepatic damage whereas the standardized concentrated extract has been as-sociated with these cases

Another strong piece of evidence that the kavalactones may be moderated by other compo-nents in traditional use comes from the epidemiologic studies carried out in Arnhem Land in

FIG 2 Chemical stuctures of (A) glutathione (B) kavalactones (C) acetaminophen and (D) cysteine

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

DENHAM ET AL256

the Northern Territories in Australia These preliminary studies have found no evidence of liverdamage in individuals who habitually ingest kava extracts equivalent to between ten and fiftytimes the daily therapeutic dose (60ndash120 mg) of kavalactones per day

Summary

Kavalactones are normally metabolized by lactone hydrolases which are enhanced by thepresence of glutathione

Glutathione naturally occurs in kava in a 11 ratio with kavalactones and is likely to reducethe likelihood of potential lactone toxicity In contrast to the traditional crude extract stan-dardized extracts contain no glutathione while containing up to 30 times the kavalactone con-centration

It appears that the high kavalactone in concentrated standardized extracts may deplete the re-serves of glutathione in the hepatocytes which could result in liver damage It would thereforeseem prudent to limit the organic solvent level in the extraction of kava to 25 ethanol in orderto ensure the preservation of the hepatoprotective effect of the glutathione Tinctures made with25 ethanol would appear to be safe as a result of this synergistic effect of the glutathione andkavalactones

Conclusions

Traditional preparations have had many years of safe usage (Norton and Ruse 1994) and tox-icity has only been reported in Europe with concentrated standardized extracts (Escher et al2001)

This paper argues that there are significant differences between concentrated extracts and thoseproduced by traditional methods that maintain a satisfactory ratio between glutathione andkavalactones In traditional extracts ratios of at least 11 kavalactoneglutathione should providea safe product with hepatoprotective action It would appear that glutathione has an importantsynergistic action in protecting the liver from potential lactone toxicity

This study suggests that standardized herbal extracts which do not contain all components ofthe traditional plant extract may have a potential to induce hepatotoxicity in susceptible people(eg those taking concomitant orthodox medicines) It is proposed that tinctures manufacturedusing a traditional cold-maceration process (in 25 ethanol and 75 water) that is more nearlyapproximate to traditional water or coconut milk extracts or raw plant material are safe in nor-mal subjects

REFERENCES

Barguil Y Rapid responses Kava and gamma-glutamyltransferase increase Hepatic enyzmatic induction or liverfunction alteration [letter in response to Escher et al 2001] eBMJ March 21 2001 Online document at httpbmjcom

British Pharmacopaeia Commission London The Stationery Office 1999Budavari S Merck Index Monograph 4483 Twelfth Edition Rahway NJ Merck and Co 1996Escher M Desmeules J Giostra E Drug points Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Kidd MD Altern Med Rev 19972(6)155ndash176

Epidemiological studies on kava use and side effects in Arnhem Land Aborigines Menzies University PersonalCommunication Personal communication with Alan Clough of Menzies University Darwin Northern Territory Aus-tralia 2002

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

KAVA WORK-IN-PROGRESS 257

Lautermann J McLaren J Schacht J Glutathione protection against gentamicin otoside effects depends on nutritionalstatus Hearing Res 199586(1ndash2)15ndash24

Lomaestro BM Malone M Glutathione in health and disease Pharmacotherapeutic issues Ann Pharmacother1995291263ndash1273

Norton SA Ruze P Kava dermopathy J Am Acad Dermatol 19943189ndash97Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene Bioorganic Med Chem 200192189ndash2194Schmidt TJ Lyszlig G Pahl HL Merfort I Helenanolide type sesquiterpene lactones Part 5 The role of glutathione ad-dition under physiological conditions Bioorganic Med Chem 19997(9)2849ndash2855

Whitton PA Rapid Responses to Letters page eBMJ March 2001 Online document at httpbmjcomZheng XG Kang JS Kim HM Jin GZ Ahn BZ Naphthazarin derivatives (V) Formation of glutathione conjugate andcytoside effects activity of 2-or 6-substituted 58-dimethoxy-14-napthoquinones in the presence of glutathione-S-transferase in rat liver S-9 fraction and mouse liver perfusate Arch Pharmacol Res 20002322ndash25

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

APPENDIX

2

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatic findings

Concomitant drugs

Notes

169

f

23

200 mg of

Data missing

Data missing

Cho

lestatic hep

atitis

ASS

deh

ydrosano

lRecov

ered

hep

atic side-effects

synthe

tic

Ren

tylin

adescribed

for all co

ncom

itan

t ka

vain

med

ications

235

m

23

200 mg of

Anx

iety states

Anx

iety states

Cho

lestatic hep

atitis

Data missing

Recov

ery after disco

ntinua

tion

synthe

tic

kava

in3

68f

33

70 m

gd

Data missing

Data missing

Increa

sed liver

Data missing

Data missing

of acetone

en

zymes (present

extract)

before beg

inning

kava

med

ication)

439

f

33

70 m

gd

Dep

ressive

4 ye

ars

Upp

er abd

ominal

Diazepam

aRecov

ery after disco

ntinua

tion

of all

of acetone

neur

osis

pressure na

usea

Gravistata

med

ications

hep

atotox

icity also

extract

vomiting icterus

L-Thy

roxin

know

n for the co

ncom

itan

tmed

ications

568

f

33

70 m

gd

Dep

ressive

2 ye

ars

Cho

lestatic hep

atitis

Neu

roplan

t forte

aRecov

ery after 97

day

s spo

radic

of acetone

emotiona

licteru

sMaa

loxa

naif

notification

s of inc

reased

liver

extract

deterioration

requ

ired

param

eters und

er M

aaloxa

na6

50f

33

70 m

gd

Data missing

2 mon

ths

Increa

sed liver

Teldan

eaaten

olol

Hep

atic side-effects also described

for

of acetone

enzy

mes liv

erHyd

rotrix

aconc

omitan

t med

ications

extract

cell-im

pairmen

tacute hep

atitis

with icteru

s 7

72f

Phy

to-

Data missing

6 mon

ths

Jaun

dice cho

lestatic

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

875

f

Phy

to-

Data missing

2 ye

ars

Cho

lestatic hep

atitis

Eun

ovaa

No he

patic side-effects kn

own for

Geriatrikum

ahe

patitis liver-cell

conc

omitan

t med

ication

(with 25

mg

impairm

ent

dry extract

with etha

nol)

981

f

23

60 m

g of

Anx

iety

9 mon

ths

Tox

ic hep

atitis w

ith

HCT-isis 12

5

Exitus seldom

ly icterus

und

er hyd

ro-

etha

nol

restlessne

ssliv

er failure acute

Cralonin Tra

chlorothiazide he

patic im

pairmen

t by

ex

tract

yello

w liver

Bay

oten

sina

alco

hol no

t ex

clude

ddys

trop

hy( bis

198

)

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

1039f

60 m

gd

Data missing

6 mon

ths an

dSe

vere hep

atitis w

ith

Paroxe

tin St John

rsquosRecov

ery after 8 3 weeks

hep

atic

14 day

s after

confluen

t ne

cros

iswort if req

uired

side-effects described

for hormon

alreex

posu

reho

rmon

al ovu

lation

ovulation

inh

ibitors

inhibitors for 6 yea

rs11

59f

23

120 mg

dAnx

iety states

4 mon

ths

Live

r-cell im

pairm

ent

Bus

copan

aSp

orad

ic notifications

of he

patic side-

effects und

er Buscop

ana

1237f

23

70 m

gd

Data missing

Data missing

Hep

atitis

Microdiola

sinc

e Recov

ery after 3 mon

ths hep

atic side-

of acetone

5 ye

ars 2

3effects also kno

wn for co

ncom

itan

tex

tract

diclofena

c IM

med

ications

1362f

Ethan

olData missing

Data missing

Live

r-cell im

pairm

ent

Non

e den

oted

No med

ical m

essage

extract

1433f

Ethan

olData missing

4 mon

ths

Bilir

ubina

emia

Cisap

ride

Hep

atic side-effects also described

for

extract

hepa

titis inc

reased

conc

omitan

t med

ication

liver enz

ymes

cirrho

sis of the

liver

1546f

Data missing

Data missing

Data missing

Seve

re liver dam

age

Prop

anolol HCT

Hep

atic side-effects also described

for

with icteru

sValsartan

aco

ncom

itan

t med

ications

1633f

33

70 m

gd

Data missing

Data missing

Cho

lestatic hep

atitis

13

60

g alcoho

lRecov

ery after 6 weeks

of acetone

with icteru

sex

tract

1760f

70 m

gd of

Dep

ression

Data missing

Increa

sed biliru

bin

Celecox

ibRecov

ery after 2 weeks

he

patic side-

aceton

e-an

d tran

saminases

effects also kno

wn for co

ncom

itan

tex

tract

indolen

t icteru

smed

ication

1850m

3ndash4

370

mg

Nervo

us2 mon

ths

Acu

te necrotizing

Alcoh

ol m

oderately

Trans

plantation notifications

of he

patic

of acetone

-tens

ion

hepa

titis irrev

ersible

1ndash2

3 paracetam

ol

side-effects und

er paracetam

ol exist

extract

liver dam

age

Nachtke

rzen

samen

ola

1921f

8ndash10

350

mg

Data missing

2 mon

ths

Increa

sed liver

Pasp

ertina

Side-effects also

kno

wn for co

ncom

itan

ten

zymes jaund

ice

Pan

toprazo

le

med

ications

hepa

titis

paracetam

ol

Basiliku

m-Tropfen

a

2050f

60 m

gd of

Stress states

7 mon

ths

Fulm

inan

t liv

erAmaryl

a G

luco

pha

geTrans

plantation hep

atic side-effects

etha

nol

failu

reSa G

ravistat

aalso kno

wn for Amaryl

a(cho

lestasis

extract

follo

wed

by

hepatitis) an

d K

limon

orm

aas w

ell as

Klim

onorm

aGravistat

a(tum

ors of the

liver

cholestasis anicteric hep

atitis)

2122f

23

120 mg of

Nervo

usn

ess

5 mon

ths

Necrosis com

plete

Max

alat

a(if

Trans

plantation hep

atic side-effects also

etha

nol-

anxiety states

destruc

tion

of

requ

ired

) Praminoa

know

n for Pr

aminoa

(tumors of the

extract

endog

enou

sthe paren

chym

a(beforeh

and V

alette

a )liv

er ch

olestasis anicteric hep

atitis)

dep

ression

fulm

inan

t liv

erfailu

re22

34f

120 mg

d of

Data missing

3 mon

ths

Hep

atitis increased

Jodthyrox

aRecov

ery after disco

ntinua

tion

of ka

vadr

y ex

tract

liver enz

ymes

med

ication sporad

ic notifications

of

with etha

nol

hepatic side-effects und

er Jod

throx

2334f

120 mg

d of

Data missing

1 mon

thIncrea

sed liver

paracetamol

Notifications

of he

patic side-effects

etha

nol

enzy

mes jaund

ice

und

er paracetam

olex

tract

( continued)

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

APPENDIX

2 (Con

tinu

ed)

Case Rep

orts as Analyz

ed by the German

Fed

eral Institute for D

rugs and M

edical Products

(the German

BfA

rM) C

ircu

lated in N

ovem

ber 200

1

Timeframe

Patient

(beginning of

(agegender)

treatment reactions

(f5female

to first occurrence

Identifierm

5male)

Dose

Indication

of symptoms)

Hepatotoxic adverse drug

Concomitant drugs

Notes

2447

f

Antares

a12

0Data missing

1 mon

thIncrea

sed liver

Fischo

lkap

seln

aRestored to he

alth after disco

ntinua

tion

etha

nol-

enzy

mes

ofco

ncom

itan

t med

ication an

dex

tract

continuationof A

ntares

a -med

ication

2535

f

Ethan

ol-

Data missing

3 mon

ths

Hep

atitis increased

Hyp

ericum

Restored to he

alth n

o he

patic side-

extract

liver enz

ymes

caps

ules

effectsk

nown for co

ncom

itan

tmed

ication

2638

m

Acetone

Data missing

2 weeks

Liver-cell

Penicillin-V

aNo he

patic side-effects kn

own for

extract

impairm

ent

conc

omitan

t med

ication

2739

m

70 m

gd of

Data missing

2 weeks

Liver-cell

Non

eData missing

aceton

e im

pairm

ent

extract

28Age

not

Kav

ain

Data missing

Hep

atitis

L-Thy

roxine

Recurren

ce of he

patic side-effects

provided

Lorza

araplus

hepatic side-effects also kno

wn for

f

Estrage

staPflastera

conc

omitan

t med

ications

Antra M

UPS

a

2960

f

Up to 48

0Dep

ressive

1 ye

arFu

lminan

t liv

eretile

frin-H

CL

Trans

plantation spo

radic notifications

mg

d of

emotiona

lfailu

repiretan

idof hep

atic side-effects und

er piretan

idetha

nol

deterioration

extract

3032

m

24

0 mg

dRestlessn

ess

3 mon

ths

Necrotizing

hep

atitis

Baldrian

aEva

luation of the

necessity for

of ethan

olwith insu

fficienc

y (occasiona

lly)

tran

splantation

extract

of the

liver m

etab

olic-

toxic-allergic dru

gdam

age

a Information on

gen

erics m

anufacturers a

nd lo

cation

s were no

t provided

for brand

-nam

e dru

gs

Sour

ce A

ppe

ndix of a letter sen

t to participan

ts in

a step-by-step

plan an

d cop

ied to the Med

icines C

ontrol A

genc

y w

hich

cop

ied the

letter to orga

niza

tion

s on

its co

n-su

ltation lis

t The

letter was entitled ldquoHea

ring

stage

II 71

71-A

-306

46 679

1800-339

0 dru

gs con

taining ka

va-kav

a ( Piper methysticum

) an

d kav

aine

inc

luding ho

meo

pathic

remed

ies with a fina

l con

centration

up to D6rdquo

IM intramuscular

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

APPENDIX 3

Executive Summary Issued January 18 2002 by the Bundesverband derArzneimittelndashHersteller e V (BAH) and Bundesverband der Pharmazeutischen

Industrie eV (BPI) Comments of BAHBPI on the BfArM Letter of November 8 2001 on Kava- and Kavain-Containing Medicinal Products

Executive Summary

On December 18 2001 the BAH and BPI the German pharmaceutical trade associations sub-mitted a statement to BfArM the German health authority on behalf of German kava manu-facturers subsequent to a letter from BfArM of November 8 2001 The industryrsquos statementcomes to the conclusion that the presented data on the benefitrisk assessment of kava- andkavain-containing medicinal products do not justify the withdrawal of marketing authorisationsThe companies already included risk information in their package leaflets and expert informa-tion at their own responsibility and consider control of patients applying kava products by aphysician as an appropriate means of ensuring safe usage

In particular in most of the reported cases the causality between kava intake and liver reac-tions is not clear because further medication was used which might have caused liver toxicityIn many cases detailed information on the patientsrsquo history co-medication consumption of al-cohol and further particulars are missing thus not permitting a sound evaluation of these casesRegarding clinical efficacy there are placebo-controlled and reference-controlled clinical stud-ies as well as open studies which demonstrate an improvement of symptoms in conditions ofnervous anxiety stress and restlessness

Data on the Risk Assessment

The report published by Kraft et al (2001) describes liver failure in a 60-year-old female patientwho took a kava preparation in an amount up to four times of the recommended daily dose Livertransplantation was required Due to further medication containing piretanid and etilefrin whichmight have caused liver-related side effects kava is unlikely to be the only cause of the side effect

The case report published by Brauer et al (2001) describes liver failure in a 22-year old femalepatient Liver transplantation was required Since the patient also took rizatriptan and oral con-traceptives which might have liver-related side effects kava is unlikely to be the only cause ofthe side effect

The case report published by Sass et al (2001) describes a 50-year old female patient who tookkava for seven months in a daily dose of 60 mg kavapyrones She experienced a hepatic comaliver transplantation was required Glimepirid and estradiol were used as co-medication Acausal connection between the liver effect and kava intake cannot definitely be excluded How-ever contribution by the co-medication to the side effect seems possible

A further case report on kava (Strahl et al 1998) describes a necrotising hepatitis in a 39-yearold female patient with positive re-exposition During the first application of the kava productan oral contraceptive as well as paroxetin were used A causal relationship with kava cannot beexcluded but the patientrsquos history and a potential pre-existing liver damage must be taken intoaccount In addition the kava preparation used was not identified by the physician

In additional reports from Switzerland Escher et al (2001) and Stoller (2000) describe twocases a liver transplantation in a 50-year old female patient using also evening primrose oil ayeast preparation and paracetamol as well as liver symptoms in a 33-year old patient who alsoapplied propyphenazon and paracetamol and consumed alcohol

KAVA WORK-IN-PROGRESS 261

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

DENHAM ET AL262

Most of the other case reports (not quoted by BfArM) cannot be assessed since essential dataare missing or they have to be evaluated as ldquoimprobablerdquo or ldquoquestionablerdquo

Data on the Benefit Assessment

According to a meta-analysis performed by Pittler and Ernst (2001) therapeutic equivalenceof kava and synthetic anxiolytics can be assumed

For an extract prepared with acetone as [a] solvent there are seven randomised placebo-con-trolled double blind studies as well as one randomised reference-controlled double blind studyperformed between 1991 and 2001 They demonstrate the efficacy of the kava extract in condi-tions of nervous anxiety stress and restlessness

On various ethanolic extracts the following data are available

A three-arm double-blind clinical study in 127 patients versus opipramol and buspirone inorder to prove efficacy in general conditions of nervous anxiety

A non-interventional study in 1187 patients confirming these results and demonstrating goodtolerability

A pharmacodynamic study versus bromazepam and placebo showing relaxing and anxiolyticeffects of a kava extract as well as better tolerability than bromazepam

An in-vivo experiment (elevated plus maze test in rats) showing a remarkable anxiolytic ef-fect of a kava extract comparable to that of diazepam

A randomised controlled double-blind study in 69 patients demonstrating improvement ofthe total Hamilton Anxiety Scale score as well as for the score for [psychologic] and somaticanxiety at a dose of 200 mg kavapyrones daily

A study demonstrating a tranquillising effect (comparable to benzodiazepines) in conditionsof anxiety prior to medical surgery

A non-interventional study in 3338 patients demonstrating a remarkable decrease in symp-toms of anxiety after an average duration of therapy of 25 months

An observational study in 52 patients showing a decrease of anxiety-related symptoms An observational study including 30 patients with psycho-somatic complaints which demon-

strated improvement Further experiments with a lower number of patients as well as a non-interventional study

currently being performed including 131 patients

As a result of their proven clinical efficacy kava preparations were included in the draft pos-itive list issued by the German ministry of health in July 2001 According to this draft list kavaproducts are reimbursable by the state health insurance like chemical substances used in this in-dication field

Conclusion

Conditions of nervous anxiety stress and restlessness must be regarded as wide-spread dis-orders which without treatment might have severe [psychologic] and social consequences andfor this reason require medical treatment In case kava products are withdrawn from the mar-ket only chemical substances would be available as therapeutic alternatives eg benzodi-azepines anxiolytics anti-depressants neuroleptics et cetera Yet all these substances have

Note added An indication field is a range of indications for example anxiety for reimbursement under the statemedical system in Germany

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263

many side-effects including liver toxicity Benzodiazepines as an alternative have a high po-tential of addiction

Available data on the benefitndashrisk assessment of kava and kava-containing medicinal prod-ucts do not justify the withdrawal of the respective marketing authorisations Inform[ing] the patient by package leaflets as well as expert information and [supervision] of patients by aphysician are regarded as an appropriate means [using kava]

REFERENCES

Brauer R Pfab R Becker K Berger H Stangl M Fulminan liver failure after taking the plant remedy kava-kava [PosterAbstract] 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash30 2001

Kraft M Spahn T Menzel J Senninger N Dietl K-H Herbst H Domschke W Lerch M Fulminant liver failure aftertaking the plant antidepressant kava-kava Deutsche Medizin Wochenschrift 2001126970ndash972

Pittler M Ernst E Efficacy of kava extract for treating anxiety Systematic review and meta-analysis J Clin Psy-chopharmacol 200020(1)84ndash89

Escher M Desmeules J Giostra E Mentha G Hepatitis associated with kava a herbal remedy for anxiety BMJ2001322139

Sass M Scnabel S Kroger J Liebe S Schareck W Acute liver failure caused by kava-kavamdasha rare indication for livertransplant 12th Workshop for Experimental and Clinical Liver Transplantation and Hepatology Wilsede GermanyJune 28ndash302001

Strahl S Ehret V Dahm H Maier K Necrotising hepatitis after taking medicinal plants Deutsche Medizin Wochen-schrift 19981231410ndash1414

Stoller R Liver damage whilst taking kava extracts Schweizerische Arztezeitung 200081(24)1335ndash1336

KAVA WORK-IN-PROGRESS 263