practice parameter: treatment of nervous system lyme disease (an

Practice Parameter Treatment of nervoussystemLymedisease (an evidence-based review)Report of theQuality Standards Subcommittee of theAmericanAcademyofNeurology

JJ Halperin MDED Shapiro MDE Logigian MDAL Belman MDL Dotevall MDGP Wormser MDL Krupp MDG Gronseth MDCT Bever Jr MD

ABSTRACT Objective To provide evidence-based recommendations on the treatment of nervous sys-tem Lyme disease and postndashLyme syndrome Three questions were addressed 1) Which antimicrobialagents are effective 2) Are different regimens preferred for different manifestations of nervous sys-tem Lyme disease 3) What duration of therapy is needed Methods The authors analyzed publishedstudies (1983ndash2003) using a structured review process to classify the evidence related to the ques-tions posed Results The panel reviewed 353 abstracts which yielded 112 potentially relevant arti-cles that were reviewed from which 37 articles were identified that were included in the analysisConclusions There are sufficient data to conclude that in both adults and children this nervous sys-tem infection responds well to penicillin ceftriaxone cefotaxime and doxycycline (Level B recommen-dation) Although most studies have used parenteral regimens for neuroborreliosis several Europeanstudies support use of oral doxycycline in adults with meningitis cranial neuritis and radiculitis (LevelB) reserving parenteral regimens for patients with parenchymal CNS involvement other severe neu-rologic symptomatology or failure to respond to oral regimens The number of children (8 years ofage) enrolled in rigorous studies of oral vs parenteral regimens has been smaller making conclusionsless statistically compelling However all available data indicate results are comparable to those ob-served in adults In contrast there is no compelling evidence that prolonged treatment with antibioticshas any beneficial effect in postndashLyme syndrome (Level A) NEUROLOGY 2007691ndash1

STATEMENT OF PURPOSE The Quality Stan-dards Subcommittee (QSS) develops scientificallysound clinically relevant practice parameters to aidin the practice of neurology This article addressesthe use of antibiotic treatments in patients with ner-vous system Lyme disease and post-Lyme syn-drome These recommendations address the needsof medical providers caring for patients with theseconditions

Lyme disease is a multisystem infectious diseasecaused by the tick-borne spirochete Borrelia burg-dorferi which frequently affects the nervous sys-tem Published guidelines are available to assist inthe diagnosis of nervous system Lyme disease1 andfor treatment of Lyme disease in general2 Howeverthere continues to be considerable controversy anduncertainty about the best approach to treatment ofneuroborreliosis In the United States Lyme diseaseaffecting the nervous system is generally treated

with parenteral antibiotics although several Euro-pean studies have demonstrated comparable effi-cacy with oral doxycycline a drug that achievesadequate levels in the nervous system Duration oftreatment varies widely with published recommenda-tions ranging up to 4 weeks despite a lack of compel-ling data supporting courses longer than 2 weeksSome practitioners treat with combinations of antimi-crobials for many months despite an absence of datato indicate this is rational or effective Finally there is alack of clarity as to which syndromes associated withLyme disease reflect nervous system infection whichare consequences of infection outside the nervous sys-tem and which are postinfectious

The relevant literature was reviewed in detail todetermine the following

1 Which antimicrobial agents have been shownto be effective or ineffective in the treatment of ner-vous system Lyme disease

This article was previously published in electronic format as an Expedited EndashPub at wwwneurologyorg

From the Department of Neurosciences (JJH) Overlook Hospital NYU School of Medicine Summit NJ Departments of Pediatrics andEpidemiology and Public Health (EDS) Yale University School of Medicine New Haven CT Department of Neurology (EL) University ofRochester School of Medicine and Dentistry NY Department of Neurology (ALB LK) SUNY Stony Brook NY Department of InfectiousDiseases (LD) Sahlgrenska University Hospital Gothenburg Sweden Division of Infectious Diseases (GPW) Department of Medicine NewYork Medical College Valhalla Department of Neurology (GG) University of Kansas Medical Center and Research Service VAMHCS and theDepartment of Neurology (CTB) University of Maryland School of Medicine

Approved by the Quality Standards Subcommittee on July 29 2006 by the Practice Committee on March 15 2007 and by the AAN Board ofDirectors on April 5 2007

Disclosure The authors report no conflicts of interest Received December 26 2006 Accepted in final form March 7 2007

Supplemental data atwwwneurologyorg

Address correspondence andreprint requests to the AmericanAcademy of Neurology 1080Montreal Ave St Paul MN55116guidelinesaancom

SPECIAL ARTICLE

Copyright copy 2007 by AAN Enterprises Inc 1

Published Ahead of Print on May 23 2007 as 10121201wnl00002655176697628

2 If different regimens are preferred for differentmanifestations of neuroborreliosis

3 What duration of therapy is needed

DESCRIPTION OF THE ANALYTIC PROCESS Inthe spring of 2004 the Quality Standards Subcom-mittee (QSS) of the American Academy of Neurol-ogy (AAN) convened an expert panel ofinvestigators from the United States and Europewho have published extensively in the field Thepanel was selected to represent a broad range of rel-evant expertise and opinion

In May 2004 a literature search was performed(all languages) using OvidMEDLINE Pubmed andEMBASE using search terms ldquoLyme Disease[DrugTherapy Therapy]rdquo ldquoBorrelia Infections[DrugTherapy Therapy]rdquo ldquoBorrelia burgdorferi groupand (borreliosis or Borrelia or neuroborreliosis)rdquoand ldquoAnti-Infective Agents[Therapeutic Use] and(antibiotic$ or antimicrob$ or anti-microb$)rdquo Thisresulted in 353 citations After elimination of dupli-cate citations each abstract was reviewed by at leasttwo members of the panel for relevance for furtherreview Any disagreements were arbitrated by athird reviewer This resulted in a list of 112 articleseach of which was then reviewed by at least twomembers of the panel Members of the panel recom-mended adding 10 additional references After de-tailed review of all 122 the panel decided 37 articlescontributed relevant assessable data Articles wereexcluded if they did not address treatment of neu-roborreliosis were not peer reviewed or were solelyreview articles The selected articles were then re-viewed in detail by all panel members to assess thequality of the evidence contained

Studies were divided into three groups adultLyme disease pediatric Lyme disease and post-Lyme syndrome Each article was reviewed to deter-mine if it specifically addressed treatment of neu-roborreliosis and if it contained original dataThose that were relevant were then graded as ClassI through IV using standard criteria as listed inAppendix 2 An evidence table was constructed list-ing each study its class the treatment regimens as-sessed whether it was prospective or retrospectivewhether it was blinded or open whether it was con-trolled or not whether it used explicit or objectiveresponse criteria the number of subjects the dura-tion of observation the completeness of follow-upand the outcomes

Overall four studies5-747 were Class I (three inpost-Lyme syndrome) One47 performed in chil-dren was considered Class I with regard to its pre-determined outcome measure CSF antibiotic levelsbut this study did not discuss clinical outcomes

Four studies were Class II (three in adults with neu-roborreliosis151819 one in children48) All were ratedClass II with regard to at least one of their predeter-mined objective measures of disease activity ELISACSF cell count or culture all of which were appar-ently measured in masked fashion All four of thesestudies would be considered Class III with regard toclinical outcomes for which assessments were notmasked All other studies were Class III or IV

ANALYSIS OF THE EVIDENCE When Lyme borre-liosis affects the nervous system it typically presentswith (a) all or part of a triadmdashmeningitis cranialneuritis and radiculoneuritis (known in Europe asGarin-Bujadoux-Bannwarth syndrome) (b) paren-chymal inflammation of the brain or spinal cord (c)mild radiculoneuropathy presenting as a more dif-fuse predominantly sensory peripheral neuropa-thy34 or (d) encephalopathy (alteration of cognitivefunction of varying severity with or without evi-dence of brain infection) Most well performedstudies have focused on (a) the group in which thediagnosis is most clear-cut and treatment responseis most straightforward to assess

Parenchymal CNS involvement is quite rare andstudies of treatment of these individuals havelargely been anecdotal (Class IV) Similarly only alimited number of small studies have addressed (c)or (d) all are Class III or IV

A separate entity defined differently by differentauthors often referred to as ldquopost-Lyme syn-dromerdquo occurs in patients who have had Lyme dis-ease but after treatment that would normally beexpected to be effective have continued to have re-sidual chronic symptoms including one or more ofthe following musculoskeletal pain (without frankarthritis fibromyalgia-like) fatigue and ldquoneuro-psychiatricrdquo symptoms The latter typically consistof perceived memory or cognitive difficulty irrita-bility sleep disturbance depression headache limbor other paresthesiasmdashall in the absence of clinicalor laboratory evidence of focal or inflammatorycentral or peripheral nervous system involvement5-7

Thus this entity is often included in discussions ofneuroborreliosis even though there is no evidenceof CNS infection in such individuals Although onlythree published studies have addressed this disor-der all have been Class I

ADULT NEUROBORRELIOSIS Long before thecharacterization of Lyme disease anecdotal re-ports89 indicated that erythema migrans-associatedmeningitis was responsive to penicillin The firsttreatment trial10 published in 1983 compared out-comes in 12 US patients with Lyme meningitis

2 Neurology 69 July 3 2007

treated with high dose IV penicillin to those in agroup of patients evaluated previously and treatedonly with prednisone Symptoms resolved far morequickly in penicillin-treated patients No penicillin-treated patients had relapses following treatmentalthough several had residual symptoms One even-tually was retreated for persistent rheumatologicsymptoms The same year a Swedish study11 re-ported 21 patients with chronic meningitis subse-quently shown to be due to B burgdorferi whoresponded to IV penicillin No control group wasincluded since these patientsrsquo illnesses had been un-remitting prior to treatment the authors consideredthe patientsrsquo pretreatment course as equivalent to anhistorical control Although these studies weresmall and Class III they indicated that neuroborre-liosis was antibiotic responsive just as other studieshad shown that Lyme disease in general was antibi-otic responsive12 Therefore no studies of treatmentof neuroborreliosis have included placebo-treatedpatients

Subsequent studies have compared the efficacyof various regimens In 1988 ceftriaxone and IVpenicillin were first shown13 to be comparably effec-tive Ceftriaxone doses of 2 gday were shown to beas effective as 4 gday (Class III) The same year aconsecutive series of 113 patients14 with neuroborre-liosis (Class IV) including 15 with encephalitistreated with parenteral penicillin doxycycline orcefuroxime was shown to have generally excellentresponses (no comparisons among regimens) Sub-sequent studies compared parenteral doxycycline(200 mgday for 2 days then 100 mgday for 8 daysn 39) to parenteral penicillin (20 MUday for 10days n 36) (Class II with regard to CSF abnor-malities Class III clinical)15 cefotaxime (2 g 3 timesdaily for 10 days) vs penicillin (5 MU 4day for 10days) (Class III)16 cefotaxime (3 g twice daily 10days n 69 of whom 49 had neuropathy) vs peni-cillin (20 MUday 10 days n 66 44 with neu-ropathy) (Class III)17 and ceftriaxone (2 gday 10days n 17) to cefotaxime (2 g every 8 hours x 10days n 16) (Class II with regard to CSF abnor-malities Class III clinical)18 In each study bothregimens demonstrated comparable rates ofefficacy19-22

Although most studies have focused on patientswith Lyme meningoradiculitis two have assessedtreatment response in patients with Lyme encepha-lopathy defined as objectively demonstrable cogni-tive abnormalities on formal mental status testingor neuropsychological testing In the first23 (ClassIII) 27 adults with Lyme encephalopathy polyneu-ropathy or both were treated with ceftriaxone 2 gIV daily for 2 weeks Response to therapy as mea-

sured by clinical signs and symptoms CSF analysesand neuropsychologic testing was gradual and typ-ically was not apparent until several months follow-ing completion of treatment Six months followingtreatment 17 (63) had improved 6 (22) im-proved but relapsed and 4 (15) were unchangedSince symptoms had been prolonged and unremit-ting prior to treatment this was believed to be dueto the effect of treatment even though no controlgroup was included for direct comparison A secondstudy in which CSF abnormalities were present in89 demonstrated efficacy of ceftriaxone (2 g dailyfor 30 days) in 18 adult patients with Lyme enceph-alopathy (Class III)24 In this study at 12 to 24months follow-up all patients were somewhat im-proved (2 patients 11) greatly improved (9 pa-tients 50) or normal (7 patients 39)

While most studies demonstrated excellent re-sponses to a wide range of antimicrobial regimensseveral have raised the possibility that a significantnumber of patients may have residual difficulties Inone25 (Class IV) 50 patients were identified whohad intrathecal production of anti-B burgdorferi an-tibody Of the 44 who were studied 31 had cranialneuropathies 12 radicular weakness 12 other formsof weakness 29 had pain 11 had headache (multi-ple findings in most) At post-treatment follow-uphalf the patients reported headache and concentra-tion problems although their neurologic deficitswere better The authors believed this representedan increase in subjective symptoms despite im-provement in objectively demonstrable abnormali-ties similar to findings reported earlier26 Howeverin the earlier study fewer than 10 had residualcognitive complaints and the remaining individualshad symptoms suggestive of sequelae of diseaserather than ongoing infection In another study27

(Class IV) 25 of the patients assessed 5 years fol-lowing treatment reported persistent ldquoneurologicrdquodifficulties However in this study sequelae ap-peared to reflect neurologic damage at the time ofinfection not ongoing infection or antibiotic treat-ment failure Similarly in a study of 36 patients28

(Class IV) treated for Lyme meningitis many de-scribed continued problems at 1 year follow-upHowever none appeared to develop new neurologicor arthritic signs or symptoms Rather symptomsranged from nonspecific (headaches myalgias) toresidua of prior cranial or peripheral neuropathies

Importantly none of these follow-up studies in-cluded a control population When this was done ina large case control study of patients (primarily withnon-neurologic Lyme disease) in Connecticut29 pa-tients who had been diagnosed with Lyme diseasereported having increased difficulties and multiple

Neurology 69 July 3 2007 3

symptoms but in fact these symptoms occurred justas frequently in uninfected age-matched controls

Finally two Finnish studies argued that longercourses of treatment might be necessary One30

(Class (III) described 60 patients treated for 100days half receiving 2 weeks of ceftriaxone followedby oral amoxicillin the other half with oral ce-fixime for the entire period Clear case definitionswere not provided and the rationale for prolongedtreatment was not clear at 1 year follow-up 2930patients in each group were considered cured Nocomparison group treated with conventionalcourses was included A second study31 describedfailure of conventional courses of ceftriaxone Diag-nostic criteria were not explicitly described in thisarticle and a major criterion of treatment successwas disappearance of symptoms perhaps resultingin the conclusion that conventional treatment wasineffective based on premature assessment of symp-tom resolution

Although numerous studies have demonstratedefficacy of parenteral antibiotic regimens in neu-roborreliosis there is also substantial evidence sup-porting the role of oral agents particularlydoxycycline Pharmacologic studies32 have demon-strated that in patients taking 200 mg of oral doxycy-cline daily the CSF concentration exceeds theMIC formany but not all B burgdorferi strains At least 10studies have addressed the outcomes of treatingneuroborreliosis with doxycycline12141519-222733-35

All but two1434 used oral doxycycline since bloodlevels achieved are comparable whether this drug isgiven orally or IV data from these studies of IVdoxycycline have been included All studies wereperformed in Europe most patients underwent CSFexamination and were shown to have a CSF pleocy-tosis No studies in US patients with neuroborrelio-sis have explicitly compared treatment response tooral doxycycline vs parenteral regimens In the vastmajority of studied patients the clinical manifesta-tion of neuroborreliosis was meningitis facial nervepalsy or radiculitis A very few patients had evi-

dence of parenchymal spinal cord or brain involve-ment None of the studies was blinded not all wererandomized or prospective To compensate for thislack of blinding many used sequential lumbarpunctures using a decline in CSF mononuclearpleocytosis as an objective marker of treatment suc-cess In one study19 in which outcome was judgedby CSF improvement as well as clinical criteriadoxycycline (200 mg orally daily for 14 days n

31) was shown to be comparable to IV penicillin (3 gevery 6 hours for 14 days n 23) in patients withLyme meningitis with or without other associatednervous system involvement (Class II) A non-randomized comparison of patients with meningitisdemonstrated comparable CSF and clinical re-sponses to ceftriaxone (2 gday 10 to 14 days n

29 and oral doxycycline 400 mgday 10 to 14 daysn 36) (Class III)20

Other studies without comparison groups haveshown high rates of efficacy with doxycycline (200to 400 mgday 9 to 17 days 34 patients with menin-gitis facial palsy no comparison group) (ClassIII)21 doxycycline (200 mgday in most one with100 mgday one with two initial days of 400 mgtreatment duration 10 to 28 days 6 with encephalo-myelitis 63 with meningitiscranial neuritisradicu-litis) (Class IV)22 and doxycycline (200 to 400 mgday 10 to 19 days 37 patients with meningitiswith or without other abnormalities including 7with myelopathy in a manuscript submitted forpublication but not yet accepted) (Class III)36

Altogether these studies provide data on 300 pa-tients with definite neuroborreliosis treated withdoxycycline In no study were outcomes demonstra-bly different whether patients received doxycyclineor parenteral beta-lactam regimens Fifteendoxycycline-treated patients were retreated for per-sistent symptoms none developed late neurologicsequelae even though follow-up for some was aslong as several years Aggregating the data from theeight studies that compared doxycycline to paren-teral regimens (figure) the overall response rate todoxycycline was 986 of the response rate to par-enteral penicillin or ceftriaxone (95 CIs 948 to1025) Given the very narrow CI it is highly un-likely that there is a clinically or statistically signifi-cant difference between these regimens

Although only two of these studies were Class IIand none Class I they do appear to demonstrate thelow probability of developing late neurologic se-quelae following treatment with oral doxycyclineSimilarly at least one US study37 demonstrated nolong-term adverse health outcomes in childrentreated for Lyme disease associated facial palsy(84 treated with oral doxycycline or amoxicillin

Figure Relative efficacy of doxycycline vsparenteral treatment

Relative efficacy ofdoxycycline vs parenteraltreatment (ratio of responserate to doxycycline toresponse rate to parenteralpenicillin or ceftriaxone RRof 10 indicating identicalresponse rates with theagents being compared) ineight studies and inaggregate Responses inmost studies were judgedclinically in study 6 CSFcriteria were used as well Forsummed data RR is 0986(95 CIs 0948 to 1025)Additional analyses ofdoxycycline vs parenteralpenicillin or ceftriaxoneindividually and of parenteralpenicillin vs ceftriaxonesimilarly showed nosignificant differences Keyto studies 127 219 315434 514 620 733 835


16 with ceftriaxone) Interpretation of such nega-tive findings in relatively small studies must be tem-pered though with an appreciation of the generallybenign long-term outcome in many patients withacute neuroborreliosis A 1990 German study34

compared outcomes in 66 antibiotic-treated patientswith neuroborreliosis to outcomes in 57 patientsevaluated before the identification of the infectiouscause of this syndrome and therefore never treatedThe majority of patients in both groups (59 ofuntreated 62 treated) were asymptomatic atlong-term follow up no patients in either group hadevidence of recurrent or progressive disease

Although it is likely that US patients with thesame manifestations of neuroborreliosis will simi-larly be doxycycline responsive there are differ-ences between the B burgdorferi strains and speciesprevalent in the United States and Europe hence thedata may not be fully applicable However Euro-pean studies3839 assessing the susceptibility of differ-ent Borrelia strains to multiple antimicrobialsincluding doxycycline have found no significantdifferences in minimal inhibitory concentrationsamong the different species Minimal bactericidalconcentrations (MBCs) have been more variable39

but for doxycycline were comparable for all speciesIn light of these observations treatment responsesmight be expected to be comparable in US and Eu-ropean patients however this remains untested

In the United States in particular there has beena general reluctance to treat CNS B burgdorferi in-fections with oral regimens Since several studieshave shown4041 that many patients with nervoussystem Lyme disease have a vigorous CSF pleocyto-sis without symptoms of meningitis some recom-mend routinely examining CSF in patients withsuspected neuroborreliosis (eg Lyme disease asso-ciated facial nerve palsy) and treating those with apleocytosis with parenteral regimens Since no ClassI studies with good long-term follow-up prove thatoral treatment of Lyme meningitis is as effective asparenteral therapy with beta-lactam agents particu-larly in patients infected with US strains of B burg-dorferi treatment with parenteral beta lactamagents is reasonable However given the absence ofevidence of late CNS complications following oralantibiotic treatment of Lyme meningitis in all trialsperformed to date coupled with the potential forgreater morbidity associated with parenteral regi-mens initial treatment with oral doxycycline with-out a lumbar puncture also appears to be a safe andvalid approach to treatmentmdashat least of facial nervepalsy42

Notably the only oral regimen that has beenshown to be effective in neuroborreliosis is doxycy-

clinemdasha drug with good CNS penetration Amoxi-cillin and cefuroxime axetil which are useful innon-neurologic Lyme disease may be useful in neu-roborreliosis patients who cannot take doxycyclinebut data in support of this are purely inferentialmdashnamely the absence of observed long-term sequelaein individuals treated with these medications37 anobservation subject to the previously notedlimitations

The role of corticosteroids in patients with neu-roborreliosis remains unclear No prospective trialshave addressed this question The issue arises mostfrequently in patients with facial nerve palsy sincesome guidelines for treatment of idiopathic facialnerve palsy43 but not others44 recommend their rou-tine use Early anecdotal observations suggestedthat patients with Lyme arthritis who received ste-roids were more difficult to cure1345 however ste-roids may well have been used in these patientsbecause they already had more severe disease Insome animal models nervous system disease ismore pronounced if corticosteroids are adminis-tered46 In contrast anecdotal studies have sug-gested outcomes in patients with severe radicularpain47 or encephalomyelitis48 may be improved ifcorticosteroids and antibiotics are administeredconcurrently One large retrospective review49 of101 patients with Lyme disease-associated facialnerve palsy found no significant difference in out-come regardless of treatment (antibiotics in 37 ste-roids alone or in combination with antibiotics in 44no treatment in the remainder) although the onlypatient in the group with severe residua had re-ceived steroids In a 10- to 20-year follow-up studyof patients with Lyme disease50 including 31 pa-tients with facial nerve palsy there was no differ-ence in long-term outcome between those who hadreceived steroids and those who had not In sum thelimited available data suggest no clear beneficial orharmful effect of steroids in patients with neurobor-reliosis who are treated with appropriate antimicro-bial therapy

PEDIATRIC NEUROBORRELIOSIS A wide rangeof Lyme disease-associated neurologic disorders hasbeen described in children including cranial neu-ropathies headache seizures meningitis meningo-encephalitis encephalopathy focal neurologicsigns ataxia vertigo chorea and transverse myeli-tis Information comes from case reports small se-ries of patients and from studies mainly focused onrheumatologic aspects of the disease conductedshortly after Lyme disease was recognized as a dis-tinct clinical entity51-53 More recent studies in bothEurope and the United States addressing neurologic


involvement in cohorts of larger size show facialnerve palsy and meningitis to be the most frequentneurologic syndromes in children415154 In contrastto adult Lyme disease patients Bannwarthrsquos syn-drome (meningoradiculitis) and mild radiculoneuri-tis are uncommon and encephalopathy rare51

As in adults there are relatively few studies spe-cifically assessing efficacy of treatment of neurobor-reliosis in children Only one study identified in theliterature review was rated as Class I (but did notaddress clinical treatment response) In this study75 children with Lyme neuroborreliosis were ran-domized to receive IV treatment with either penicil-lin G or ceftriaxone55 On the 10th day of treatmentpaired samples of serum and CSF were tested to de-termine the antibiotic concentration in each sampleCSF concentrations of both drugs exceeded the min-imal inhibitory concentration though not surpris-ingly the duration for which this was true wassubstantially greater for ceftriaxone (24 hours)than for penicillin The authors concluded that bothdrugs likely would be effective for treating Lymeneuroborreliosis but no specific treatment responsedata were provided In the only Class II study56 23children with Lyme neuroborreliosis were random-ized to receive 14 days of IV treatment with eitherpenicillin G or ceftriaxone All children did well andnone had sequelae at follow-up 6 months laterThe investigators concluded that treatment with ei-ther drug was highly effective

All of the remaining studies were categorized aseither Class III or Class IV Most were case seriesalthough there was one cross-sectional survey withcontrols an observational study of 169 childrenwith Lyme neuroborreliosis54 (facial nerve palsy55 meningitis 272 Bannwarthrsquos syndrome36 meningoencephalitis 36 ldquoGuillain-Barrerdquo 18) in Lower Saxony in Germany Al-though the focus of the study was on theepidemiology and the diagnosis of neuroborreliosisvirtually all of the children were treated IV withpenicillin for 10 to 14 days and all had excellent out-comes In a Swedish study 203 children with Lymeneuroborreliosis were treated IV with penicillin (53children) ceftriaxone (109 children) or cefotaxime(19 children) or orally with doxycycline (22 chil-dren)35 At follow-up symptoms and signs resolvedin 58 of the children by the end of treatment in92 by 2 months and in all children by 6 monthsafter treatment (three were lost to follow-up) In aretrospective Austrian study of 160 children withLyme disease (45 [28] with meningitis facialpalsy or both33) 33 received ceftriaxone IV 7 re-ceived benzylpenicillin IM and 5 received doxycy-cline orally All were treated for 10 to 21 days All

151 children seen at 3-month follow-up recoveredcompletely In a series of 187 Danish patients withLyme neuroborreliosis 40 (21) were children40

Most received penicillin G IV and all apparently didwell at follow-up months to years later In a USstudy children with Lyme neuroborreliosis57 (facialnerve palsy n 6 meningitis n 4 overlap notspecified) were identified prospectively from a co-hort of 201 children with Lyme disease The chil-dren were treated with a variety of antibioticregimens Ceftriaxone IV was administered to threechildren with meningitis two with facial nervepalsy and one with facial nerve palsy and meningi-tis The other children received orally administeredantibiotics (amoxicillin doxycycline erythromycinor penicillin) At follow-up none reported sequelae

Nine Swedish children with Lyme neuroborrelio-sis were treated with penicillin G IV58 All but one (achild whose facial palsy persisted for 3 months) hadprompt improvement and cure of their illnesses Atfollow-up 3 months later all were asymptomaticand appeared to be cured Although one report offollow-up of 63 children with erythema migranswas identified by our search53 only one of the chil-dren had neuroborreliosis (facial nerve palsy) Inany case all of the children in the report were wellat follow-up Finally 43 children with facial nervepalsy due to Lyme disease were assessed 7 to 161months (mean 49 months) after infection37 Ofthese 84 had been treated orally with either doxy-cycline or amoxicillin the remainder had receivedceftriaxone Twenty of the patients underwent neu-ropsychological testing and all had average or aboveaverage scores on a large battery of tests Althoughchildren with facial nerve palsy were more likelythan normal matched controls to report musculo-skeletal pain changes in behavior and numbnessreports of problems with activities of daily livingwere similar among affected patients and matchedcontrols The investigators concluded that the long-term neuropsychological and health outcomes ofchildren with facial nerve palsy due to Lyme diseasewere comparable to those who did not have Lymedisease

Conclusions Based on four Class II studies antibi-otic regimens have been established as probably safeand effective for both children and adults OneClass I and one Class II study suggest that paren-teral regimens are probably safe and effective forsevere neurologic disease but two Class II studies1519

and numerous Class III and IV studies suggest thatoral treatment particularly with doxycycline iscomparably safe and effective in many situationsnot involving parenchymal CNS involvement Al-though the evidence is stronger in adults than chil-


dren all available evidence indicates that theresponses to oral treatment are comparable inadults and children However it must be empha-sized that no definitive data exist to establish thesuperioritymdashor lack thereofmdashof either oral or par-enteral treatment Specific regimens are listed in ta-bles 1 and 2

POSTndashLYME DISEASE Post-Lyme syndrome Asdiscussed above patients who have received ac-cepted antibiotic regimens for various forms ofLyme disease sometimes have residual chronicsymptoms referred to variably as post-Lyme syn-drome (PLS) post-Lyme disease syndrome post-treatment chronic Lyme disease (PTCLD) or evenchronic Lyme disease There has been controversyas to whether PLS is a form of active infection inwhich the organism is difficult or impossible toeradicate from various ldquoprivilegedrdquo sites vs a postin-fectious or noninfectious type of chronic fatiguesyndrome in which there is no ongoing infectionArguments offered to support the possibility of per-sistent active infection derive from the apparentsimilarity between these symptoms and patientsrsquoperceptions of the cognitive difficulty and fatiguenoted with untreated or partially treated Lyme dis-ease however in patients with untreated or incom-pletely treated Lyme disease these symptoms aretypically associated with objective abnormalities onphysical or laboratory examination and symptomsand abnormal findings clearly respond frequently

with symptom resolution to a 2- to 4-week courseof IV antibiotics2459

Most available data argue against persistent Bburgdorferi infection in patients who have receivedwhat are normally curative courses of antimicrobialtherapy First antibiotic resistance has not beendemonstrated in this genus60-62 Second persistentsymptoms do not correlate with any objective mea-sure of nervous system disease or with laboratorymeasures of inflammation563 Third there is no pre-cedent for such a phenomenon in other spirochetalinfections64 Fourth anti-B burgdorferi antibodyconcentrations often decline even to undetectablelevels despite persistent symptoms56365 Such a de-cline in antibody in the face of persistent infectionappears to be without precedent in other bacterialinfections Fifth Lyme disease lacks characteristicsof other infections that justify longer treatmentcourses such as infections in which available anti-microbials have poor in vitro activity against the or-ganism infections caused by an intracellularpathogen or infections involving a biofilm Finallypatients with PLS do not respond to a further courseof IV antibiotics Anecdotally some experience asubjective improvement while on antibiotics withsymptoms recurring rapidly following medicationdiscontinuation suggesting a placebo effect Stillbased on a transient improvement in symptomssome physicians have treated patients with PLS withvarious antibiotic regimens for months to years

Table 1 Antimicrobial regimens used in treatment of nervous system Lyme disease

Medication Adult dose Pediatric dose Classification

Oral regimens

Doxycycline (preferred) 100 (ndash200) mg BID 8 yo 4 (ndash8) mgkgd in 2 divided dosesmax 200 mgdose

B

Amoxicillin (whendoxycyclinecontraindicated)dagger

500 mg TID 50 mgkgd in 3 divided doses max 500mgdose

C

Cefuroxime axetil (whendoxycyclinecontraindicated)dagger

500 mg BID 30 mgkgd in 2 divided doses max 500mgdose

C

Parenteral regimens

Ceftriaxone 2 g IV daily 50ndash75 mgkgd in 1 dose max 2 g B

Cefotaxime 2 g IV Q8H 150ndash200 mgkgday in 3ndash4 divided dosesmax 6 gday

B

Penicillin GDagger 18ndash24 MUddivided doses Q4H

200ndash400000 UKgd divided Q4H max18ndash24 MUday

B

For all recommended duration is 14 days although published studies have used courses ranging from 10 to 28 days without sig-nificantly different outcomesTetracyclines are relatively contraindicated in children 8 years of age or in pregnant or lactating womendaggerThese two oral regimens are effective in non-nervous system Lyme borreliosis There are no data demonstrating efficacy in neu-roborreliosis but large numbers of patients have been treated with these regimens for other forms of Lyme disease without obvi-ous subsequent onset of nervous system involvement As such they may be an oral alternative in individuals who cannot takedoxycyclineDaggerThe antibiotic dosage should be reduced for patients with impaired renal function


To address this controversy three randomizeddouble-blind placebo-controlled trials of antibiotictherapy in PLS have been published in adults withwell-documented Lyme disease who had previouslyreceived accepted initial courses of antibiotics foracute disease but who had residual symptoms typi-cal of PLS

The first pair of trials with a combined total of129 patients was reported initially in a single arti-cle in which patients were divided into two groupsone seropositive one seronegative Treatment re-sponse reported in each group as a separate trialwas measured using the physical and mental health-related quality of life components of the MedicalOutcomes Study 36-item short form general healthsurvey (SF-36) as the primary outcome measure5 Asecond article7 assessed whether neurocognitivechanges occurred in the same population measur-ing attention memory and executive functioningusing a battery of neuropsychological tests as oneset of primary outcome measures Mood and otherpsychiatric symptoms were assessed using the BeckDepression Inventory (BDI) and the Minnesota

Multiphasic Personality Inventory (MMPI-2) as anadditional set of primary outcome measures En-trance criteria were symptoms of classic Lymedisease physician-documented recommended treat-ment of Lyme disease and typical musculoskeletalor cognitive PLS symptoms often accompanied byfatigue that had begun within 6 months of the initialinfection and that had been present for at least 6months (but less than 12 years) Patients could beeither seropositive or seronegative but were ex-cluded if they had previously received IV antibioticsfor 60 or more days had known hypersensitivity tostudy medications had active synovitis or positivePCR for B burgdorferi gene segments in CSF orblood Patients were evaluated at baseline wererandomized to receive either placebo or IV ceftriax-one 2 gday for 30 days followed by oral doxycy-cline 200 mgday for 60 days and then re-evaluatedat 1 3 and 6 months

For the combined group of 129 patients at base-line the physical component of the SF-36 score wasabout 15 SD below and the mental componentabout 05 SD below that of age-matched norms in-dicating significant impairment in health-relatedquality of life5 By contrast although the baselineBDI showed mild to moderate depression there wasno significant difference in neuropsychological testmeasurements in the combined patient group vsage-matched norms7 There was also no significantdifference in baseline SF-36 BDI or neuropsycho-logical test measurements in seropositive vs sero-negative patients

At 6 months there was no significant differencein the SF-365 neuropsychological test and BDImeasurements7 between patients who had receivedplacebo vs antibiotic therapy Notably about 40of patients in each group (ie placebo and antibi-otic) improved in the total SF-36 summary score at 6months while about 30were unchanged and 30were worse Similarly several neuropsychologicaltest and BDI measurements improved at 6 months inboth placebo- and antibiotic-treated patients

The other trial6 investigated potential changes infatigue cognitive function and CSF clearance of Bburgdorferi antigen in 55 patients with PLS Theprimary outcome measures were (a) improvementin fatigue on a global fatigue severity measure (FSS-11) (b) improvement in mental speed on the A-Atest a computerized reaction time task of cognitiveprocessing speed and (c) CSF clearance of B burg-dorferi OspA antigen Entrance criteria werephysician-documented erythema migrans or aCDC-defined late manifestation of Lyme diseaseconfirmed by positive ELISA and Western blotcompletion of a recommended course of antibiotic

Table 2 Syndromes and treatment options

Syndrome Treatment options

Meningitis Parenteralparticularly ifsevere

Doxycycline POdagger

Any neurologic syndrome with CSFpleocytosis

Parenteralparticularly ifsevere


Peripheral nerve (radiculopathydiffuse neuropathy mononeuropathymultiplex cranial neuropathy normalCSF)


Parenteral iftreatment failure orif severe

Encephalomyelitis Parenteral

Encephalopathy Parenteral

Post-treatment Lyme syndrome No antibioticsindicatedsymptomaticmanagement only

Available data in European neuroborreliosis indicate that oraldoxycycline and parenteral ceftriaxone are equally effective inLyme meningitis and anecdotal data from the United Statesindicate that in patients with Lyme diseasendashassociated facialpalsy response to oral treatment is sufficient that CSF exami-nation may be unnecessary Although none of these studies isClass I it was the consensus of the panel that in the absenceof brain or spinal cord involvement oral treatment of neurobor-reliosis is an acceptable option in appropriate circumstancesdaggerStudies assessing oral treatment of neuroborreliosis have onlyused doxycycline Other agents such as amoxicillin or cefu-roxime axetil may be effective in individuals who cannot toler-ate doxycycline but relevant data are lacking


treatment and severe persistent fatigue (typicallyaccompanied by other musculoskeletal or neuropsy-chiatric symptoms of PLS) coincident with initial in-fection Patients were excluded if they had knowncephalosporin hypersensitivity or a confoundingmedical or severe psychiatric disorder Patients wereevaluated at baseline were randomized to receiveeither placebo or IV ceftriaxone 2 gday for 28 daysand then re-evaluated at 1 and 6 months

At 1 month fatigue improved in both antibiotic-and placebo-treated groups At 6 months there wasa significant improvement in the FSS-11 fatiguescore in patients treated with ceftriaxone comparedto placebo with no improvement in mental speed(or any of the other secondary neuropsychologicaltests) or in clearance of CSF antigen (although anti-gen was detected in CSF of only 16 of the patientsat baseline) in the antibiotic vs placebo group Sub-group analysis showed that improvement in the fa-tigue score after ceftriaxone particularly occurred inpatients who had received an initial oral rather thanIV course of antibiotics for acute Lyme disease Ofnote patients in the ceftriaxone group more oftenguessed that they were on active treatment than didthe placebo patients suggesting that blinding of pa-tients may have been compromised and that im-provement in fatigue in the ceftriaxone group mayhave been due to placebo effect Importantly theauthors point out that the adverse events associatedwith parenteral antibiotic therapy do not justifysuch therapy for fatigue a problem which might re-spond to safer symptomatic management

An additional controlled treatment trial66 pre-sented to date only in abstract form evaluated 10weeks of ceftriaxone treatment vs placebo in a totalof 37 patients Preliminary findings appear to indi-cate lack of sustained improvement in cognitivefunction in the antibiotic-treated patients at 14weeks post-treatment follow-up Although this is amuch anticipated NIH-funded trial since it has notyet been published its methodology detailed con-clusions and validity cannot be assessed and it hasnot been included in the analysis

In summary published antibiotic treatment tri-als of PLS provide compelling Class I evidence thatPLS is not due to active Borrelia infection and is notresponsive to further antibiotic therapy particularlywith respect to overall health-related quality of lifeand cognitive and depressive symptoms Whetherprofound fatigue in PLS is antibiotic responsive isstill an open question but the data on this point arepartially confounded and as noted above the au-thors of the study of fatigue in PLS do not endorsefurther antibiotic therapy for this symptom Theavailable studies confirm the importance of placebo

effect that obscures interpretation of symptom im-provement in individual patients with PLS treatedwith antibiotics

Finally in PLS there is dissociation between per-ceived problems with cognition and memory (whichare often also present in association with depressivesymptoms) and normal neuropsychological testing7

In contrast to individuals with PLS patients withprior untreated or partially treated Lyme diseasewith late cognitive symptoms (which typically aremilder than in PLS) often have abnormalities onneuropsychological testing CSF or imaging stud-ies The latter patients may well have active infec-tion (eg Lyme encephalopathy) and often respondto IV ceftriaxone with resolution of symptoms

Conclusions Several Class I studies indicate that thedisorder referred to as post-Lyme syndrome doesnot respond to prolonged courses of antibiotics andthat such treatment can be associated with seriousadverse events (see below)

RECOMMENDATIONS

1) Parenteral penicillin ceftriaxone and cefo-taxime are probably safe and effective treat-ments for peripheral nervous system Lymedisease and for CNS Lyme disease with or with-out parenchymal involvement (Level Brecommendation)

2) Oral doxycycline is probably a safe and effectivetreatment for peripheral nervous system Lymedisease and for CNS Lyme disease without pa-renchymal involvement (Level B recommenda-tion) Amoxicillin and cefuroxime axetil mayprovide alternatives but supporting data arelacking

3) Prolonged courses of antibiotics do not improvethe outcome of post-Lyme syndrome are poten-tially associated with adverse events and aretherefore not recommended (Level Arecommendation)

COMMENT ON TREATMENT SAFETY Althoughthe antimicrobial regimens discussed are widelyused and generally well tolerated none is withoutpotential side effects In one of the studies of post-Lyme syndrome6 12 of the 28 patients receivingceftriaxone developed diarrhea while 4 developedallergic reactions (1 anaphylaxis 3 minor) Becauseof its biliary excretion ceftriaxone tends to causepseudolithiasis (precipitation of the drug in the gallbladder) and may be associated with pseudomem-branous colitis more frequently than other antimi-crobials Of the 55 patients (treated and placebo)who had indwelling IV access catheters 3 developed


line sepsis (1 of 28 on ceftriaxone) One othertreated patient in this study developed anaphylaxiswhile 10 developed less severe adverse events In theother published pair of long-term treatment trials5

27 of 129 patients developed adverse effects (16 of64 receiving ceftriaxone) 2 of which (both patientson ceftriaxone) were life threatening (1 pulmonaryembolism 1 fever and GI bleed) Combining treatedpatients in these three studies life-threatening com-plications occurred in 1 per 23 while overall ad-verse events occurred in about 1 of every 3 treatedpatients

Although oral doxycycline avoids issues relatedto line infections the drug is associated with gastricirritation and with photosensitization The latter isparticularly problematic since most acute manifes-tations of Lyme disease occur in summer and au-tumn Tetracyclines also cause abnormalities ofdeveloping bones and teeth in the fetus and in chil-dren under age 8

RECOMMENDATIONS FOR FUTURE RESEARCHQuestions remain regarding the preferred therapeu-tic approach in early neuroborreliosis particularlyamong US patients The efficacy of oral doxycyclinecompared to a parenteral regimen such as ceftriax-one needs to be clearly established as does the pre-dictive value of CSF abnormalities Assuming oraldoxycycline is shown to be effective it would thenbe helpful to assess the relative efficacy of other oralregimens such as amoxicillin or cefuroxime axetil

Similarly the optimal approach to the very rareentity of parenchymal CNS neuroborreliosis re-mains undefined An assessment of preferred treat-ment duration as well as a clear determination ofthe correct metrics of successful treatment wouldbe very helpful

Although it is clear that prolonged antimicrobialtherapy is not helpful in the treatment of post-Lymesyndrome this entity will remain problematic untilits pathophysiology is better understood

MISSION STATEMENT OF QSS

The mission of the QSS is to prioritize develop and publish evidence-based practice parameters related to the diagnosis treatment and prog-nosis of neurologic disorders The QSS is committed to using the mostrigorous methods available within our budget in collaboration withother available AAN resources to most efficiently accomplish thismission

DISCLAIMER

This statement is provided as an educational service of the American Acad-emy of Neurology It is based on an assessment of current scientific andclinical information It is not intended to include all possible propermethodsof care for a particular neurologic problem or all legitimate criteria forchoosing to use a specific procedure Neither is it intended to exclude anyreasonable alternativemethodologies The AAN recognizes that specific pa-

tient care decisions are the prerogative of the patient and the physician car-ing for the patient based on all of the circumstances involved

CONFLICT OF INTEREST STATEMENT

The American Academy of Neurology is committed to producing inde-pendent critical and truthful clinical practice guidelines (CPGs) Signifi-cant efforts are made to minimize the potential for conflicts of interest toinfluence the recommendations of this CPG To the extent possible theAAN keeps separate those who have a financial stake in the success orfailure of the products appraised in the CPGs and the developers of theguidelines Conflict of interest forms were obtained from all authors andreviewed by an oversight committee prior to project initiation AANlimits the participation of authors with substantial conflicts of interestThe AAN forbids commercial participation in or funding of guidelineprojects Drafts of the guidelines have been reviewed by at least threeAAN committees a network of neurologists Neurology peer reviewersand representatives from related fields The AAN Guideline AuthorConflict of Interest Policy can be viewed at wwwaancom

APPENDIX 1Quality standards subcommittee members Jacqueline FrenchMD FAAN (Co-Chair) Gary S Gronseth MD (Co-Chair)Charles E Argoff MD Stephen Ashwal MD FAAN (ex-officio) Christopher Bever Jr MD MBA FAAN John DEngland MD FAAN Gary M Franklin MD MPH (ex-offi-cio) Gary H Friday MD MPH FAAN Larry B GoldsteinMD FAAN Deborah Hirtz MD (ex-officio) Robert G Hollo-way MDMPH FAAN Donald J Iverson MD FAAN LeslieA Morrison MD Clifford J Schostal MD David J Thur-man MD MPH William J Weiner MD FAAN SamuelWiebe MD

APPENDIX 2AAN classification of evidence for therapeutic interventionClass I Prospective randomized controlled clinical trial withmasked outcome assessment in a representative populationThe following are required

(a) primary outcome(s) clearly defined(b) exclusioninclusion criteria clearly defined(c) adequate accounting for drop-outs and cross-overs with

numbers sufficiently low to have minimal potential for bias(d) Relevant baseline characteristics are presented and sub-

stantially equivalent among treatment groups or there is appro-priate statistical adjustment for differencesClass II Prospective matched group cohort study in a represen-tative population with masked outcome assessment that meetsandashd above OR a RCT in a representative population that lacksone criterion andashdClass III All other controlled trials (including well-defined nat-ural history controls or patients serving as own controls) in arepresentative population where outcome is independently as-sessed or independently derived by objective outcomemeasurementClass IV Evidence from uncontrolled studies case series casereports or expert opinionObjective outcome measurement an outcome measure that isunlikely to be affected by an observerrsquos (patient treating physi-cian investigator) expectation or bias (eg blood tests admin-istrative outcome data)

APPENDIX 3Classification of recommendations

A Established as effective ineffective or harmful for thegiven condition in the specified population (Level A rat-ing requires at least two consistent Class I studies)


B Probably effective ineffective or harmful for the givencondition in the specified population (Level B rating re-quires at least one Class I study or at least two consistentClass II studies)

C Possibly effective ineffective or harmful for the givencondition in the specified population (Level C rating re-quires at least one Class II study or two consistent Class IIIstudies)

U Data inadequate or conflicting given current knowledgetreatment is unproven

REFERENCES1 Halperin J Logigian E Finkel M Pearl R Practice pa-

rameters for the diagnosis of patients with nervous systemLyme borreliosis (Lyme disease) Neurology 199646619ndash627

2 Wormser GP Dattwyler RJ Shapiro ED et al The clini-cal assessment treatment and prevention of Lyme dis-ease human granulocytic anaplasmosis and babesiosisClinical Practice Guidelines by the Infectious Diseases So-ciety of America Clin Infect Dis 2006431089ndash1134

3 Halperin JJ Luft BJ Volkman DJ Dattwyler RJ Lymeneuroborreliosis peripheral nervous system manifesta-tions Brain 19901131207ndash1221

4 Logigian EL Steere AC Clinical and electrophysiologicfindings in chronic neuropathy of Lyme disease Neurol-ogy 199242303ndash311

5 Klempner M Hu L Evans J et al Two controlled trialsof antibiotic treatment in patients with persistent symp-toms and a history of Lyme disease N Engl J Med 200134585ndash92

6 Krupp L Hyman L Grimson R et al Study and treat-ment of post Lyme disease (STOP-LD) a randomizeddouble masked clinical trial Neurology 2003601923ndash1930

7 Kaplan R Trevino R Johnson G et al Cognitive func-tion in post-treatment Lyme disease do additional antibi-otics help Neurology 2003601916ndash1922

8 Hellerstrom S Erythema chronicum migrans Afzeliuswith meningitis Acta Derm Venereol 195131227ndash234

9 Hollstrom E Successful treatment of erythema migransAfzelius Acta Derm Venereol 195131235ndash243

10 Steere AC Pachner ARMalawista SE Neurologic abnor-malities of Lyme disease successful treatment with high-dose intravenous penicillin Ann Intern Med 198399767ndash772

11 Skoldenberg B Stiernstedt G Garde A Kolmodin GCarlstrom A Nord C Chronic meningitis caused by apenicillin-sensitive microorganism Lancet 1983II75ndash78

12 Steere AC Green J Hutchinson GJ et al Treatment ofLyme disease Zentralbl Bakteriol Mikrobiol Hyg [a]1987263352ndash356

13 Dattwyler R Halperin J Volkman D Luft B Treatmentof late Lyme borreliosisndashrandomised comparison ofceftriaxone and penicillin Lancet 198811191ndash1194

14 Skoldenberg B Stiernstedt G Karlsson M Wretlind BSvenungsson B Treatment of Lyme borreliosis with em-phasis on neurological disease Ann NY Acad Sci 1988539317ndash323

15 Kohlhepp W Oschmann P Mertens H Treatment ofLyme borreliosis Randomized comparison of doxycy-cline and penicillin G J Neurol 1989236464ndash469

16 Pfister H Preac-Mursic V Wilske B Einhaupl K Cefo-taxime vs penicillin G for acute neurologic manifestations

in Lyme borreliosis A prospective randomized studyArch Neurol 1989461190ndash1194

17 Hassler D Zoller L Haude M Hufnagel H Heinrich FSonntag H Cefotaxime versus penicillin in the late stageof Lyme disease prospective randomized therapeuticstudy Infection 19901816ndash20

18 Pfister H Preac-Mursic V Wilske B Schielke E Sorgel FEinhaupl K Randomized comparison of ceftriaxone andcefotaxime in Lyme neuroborreliosis J Infect Dis 1991163311ndash318

19 Karlsson M Hammers-Berggren S Lindquist L Stiernst-edt G Svenungsson B Comparison of intravenous peni-cillin G and oral doxycycline for treatment of Lymeneuroborreliosis Neurology 1994441203ndash1207

20 Borg R Dotevall L Hagberg L et al Intravenous ceftri-axone compared with oral doxycycline for the treatmentof Lyme neuroborreliosis Scand J Infect Dis 200537449ndash454

21 Dotevall L Hagberg L Successful oral doxycycline treat-ment of Lyme disease-associated facial palsy and meningi-tis Clin Infect Dis 199928569ndash574

22 Karkkonen K Stiernstedt S Karlsson M Follow-up ofpatients treated with oral doxycycline for Lyme neurobor-reliosis Scand J Infect Dis 200133259ndash262

23 Logigian E Kaplan R Steere A Chronic neurologic man-ifestations of Lyme disease N Engl J Med 19903231438ndash1444

24 Logigian EL Kaplan RF Steere AC Successful treatmentof Lyme encephalopathy with intravenous ceftriaxoneJ Infect Dis 1999180377ndash383

25 Treib J Fernandez A Haass A Grauer M Holzer GWoessner R Clinical and serologic follow-up in patientswith neuroborreliosis Neurology 1998511489ndash1491

26 Shadick N Phillips C Logigian E et al The long-termclinical outcomes of Lyme disease a population-basedretrospective cohort study Ann Intern Med 1994121560ndash567

27 Berglund J Stjernberg L Ornstein K Tykesson-JoelssonK Walter H 5-y Follow-up study of patients with neu-roborreliosis Scand J Infect Dis 200234421ndash425

28 Cimperman J Maraspin V Lotric-Furlan S Ruzic-SabljicE Strle F Lyme meningitis a one-year follow up con-trolled study Wiener Klinische Wochenschrift 1999111961ndash963

29 Seltzer E Gerber M Cartter M Freudigman K ShapiroE Long-term outcomes of persons with Lyme diseaseJAMA 2000283609ndash616

30 Oksi J Nikoskelainen J Viljanen M Comparison of oralcefixime and intravenous ceftriaxone followed by oralamoxicillin in disseminated Lyme borreliosis Eur J ClinMicrobiol Infect Dis 199817715ndash719

31 Wahlberg P Granlund H Nyman D Panelius J SeppalaI Treatment of late Lyme borreliosis J Infect 199429255ndash261

32 Karlsson M Hammers S Nilsson-Ehle I Malmborg AWretlind B Concentrations of doxycycline and penicillinG in sera and cerebrospinal fluid of patients treated forneuroborreliosis Antimicrob Agents Chemother 1996401104ndash1107

33 Krbkova L Stanek G Therapy of Lyme borreliosis inchildren Infection 199624170ndash173

34 Kruger H Kohlhepp W Konig S Follow-up of antibioti-cally treated and untreated neuroborreliosis Acta NeurolScand 19908259ndash67


35 Thorstrand C Belfrage E Bennet R Malmborg P Eriks-son M Successful treatment of neuroborreliosis with tenday regimens Pediatr Infect Dis J 200221142ndash145

36 Dotevall L Borg R Hagberg L Doxycycline treatment ofLyme neuroborreliosis with meningoradiculitis andormyeloencephalopathy Submitted

37 Vazquez M Sparrow S Shapiro E Long-term neuropsy-chologic and health outcomes of children with facialnerve palsy due to Lyme disease Pediatrics 200311293ndash97

38 Baradaran-Dilmaghani R Stanek G In vitro susceptibil-ity of thirty Borrelia strains from various sources againsteight antimicrobial chemotherapeutics Infection 19962460ndash63

39 Sicklinger M Wienecke R Neubert U In vitro suscepti-bility testing of four antibiotics against Borrelia burgdor-feri a comparison of results for the three genospeciesBorrelia afzelii Borrelia garinii and Borrelia burgdorferisensu stricto J Clin Microbiol 2003411791ndash1793

40 Hansen K Lebech A The clinical and epidemiologicalprofile of Lyme neuroborreliosis in Denmark 1985ndash1990A prospective study of 187 patients with Borrelia burgdor-feri specific intrathecal antibody production Brain 1992115399ndash423

41 Belman AL Reynolds L Preston T Postels D Grimson RCoyle PK Cerebrospinal fluid findings in children withLyme disease-associated facial nerve palsy Arch PediatrAdolesc Med 1998152928ndash929

42 Shapiro E Gerber M Lyme disease and facial nerve palsyMore questions than answers Arch Pediatr Adolesc Med19981521183ndash1184

43 Grogan PM Gronseth GS Practice parameter Steroidsacyclovir and surgery for Bellrsquos palsy (an evidence-basedreview) report of the Quality Standards Subcommittee ofthe American Academy of Neurology Neurology 200156830ndash836

44 Corticosteroids for Bellrsquos palsy (idiopathic facial paraly-sis) John Wiley and Sons 2006 Accessed at httpwww-cochraneorgreviewsenab001942html

45 Bentas W Karch H Huppertz H Lyme arthritis in chil-dren and adolescents outcome 12 months after initiationof antibiotic therapy J Rheumatol 2000272025ndash2030

46 Pachner AR Amemiya K Bartlett M Schaefer H ReddyK Zhang WF Lyme borreliosis in rhesus macaques ef-fects of corticosteroids on spirochetal load and isotypeswitching of anti-Borrelia burgdorferi antibody Clin Di-agn Lab Immunol 20018225ndash232

47 Pfister HW Einhaupl KM Franz P Garner C Corticoste-roids for radicular pain in Bannwarthrsquos syndrome a dou-ble blind randomized placebo controlled trial Ann NYAcad Sci 1988539485ndash487

48 Massengo SA Bonnet F Braun C Vital A Beylot J Bas-tard J Severe neuroborreliosis the benefit of prolongedhigh-dose combination of antimicrobial agents with ste-roids an illustrative case Diagn Microbiol Infect Dis200551127ndash130

49 Clark JR Carlson RD Sasaki CT Pachner AR SteereAC Facial paralysis in Lyme disease Laryngoscope 1985951341ndash1345

50 Kalish R Kaplan R Taylor E Jones-Woodward LWorkman K Steere A Evaluation of study patients withLyme disease 10-20-year follow-up J Infect Dis 2001183453ndash460

51 Belman AL Iyer M Coyle PK Dattwyler R Neurologicmanifestations in children with North American Lymedisease Neurology 1993432609ndash2614

52 Eichenfield A Goldsmith D Benach J et al ChildhoodLyme arthritis experience in an endemic area J Pediatr1986109753ndash758

53 Salazar J GerberM Goff C Long-term outcome of Lymedisease in children given early treatment J Pediatr 1993122591ndash593

54 Christen H Hanefeld F Eiffert H Thomssen R Epidemi-ology and clinical manifestations of Lyme borreliosis inchildhood A prospective multicentre study with specialregard to neuroborreliosis Acta Paediatr Suppl 19933861ndash75

55 Millner M Thalhammer G Neuroborreliosis in child-hood treatment with penicillin sodium and ceftriaxoneActa Dermatovenerologica Alpina Panonica et Adriatica19965169ndash172

56 Mullegger R Millner M Stanek G Spork K Penicillin Gsodium and ceftriaxone in the treatment of neuroborrelio-sis in childrenndasha prospective study Infection 199119279ndash283

57 Gerber M Shapiro E Burke G Parcells V Bell G Lymedisease in children in southeastern Connecticut PediatricLyme Disease Study Group N Engl JMed 19963351270ndash1274

58 Jorbeck H Gustafsson P Lind H Stiernstedt G Tick-borne Borrelia-meningitis in children An outbreak in theKalmar area during the summer of 1984 Acta PaediatrScand 198776228ndash233

59 Halperin JJ Luft BJ Anand AK et al Lyme neuroborre-liosis central nervous system manifestations Neurology198939753ndash759

60 Hunfield K-P Kraiczy P Kekoukh E Schafer V Brade VStandardized in vitro susceptibility testing of Borreliaburgdorferi against well-known and newly developed an-timicrobial agents- Possible implications for new thera-peutic approaches to Lyme disease Int J Med Microbiol2002291(suppl 33)125ndash137

61 Hunfield K-P Ruzic-Sabljic E Norris D Kraiczy P StrleF In vitro susceptibility testing of Borrelia burgdorferisensu lato isolates cultured from patients with erythemamigrans before and after antimicrobial chemotherapyAntimicrob Agents Chemother 2005491294ndash1301

62 Nowakowski J Wormser G Treatment of early Lymedisease infection associated with erythema migrans InCoyle P ed Lyme disease St Louis MO Mosby-YearBook 1993149ndash162

63 Nowakowski J Nadelman R Sell R et al Long-termfollow-up of patients with culture-confirmed Lyme dis-ease Am J Med 200311591ndash96

64 Wormser G Lyme disease Insights into the use of antimi-crobials for prevention and treatment in the context ofexperience with other spirochetal infections Mt SinaiJ Med 199562188ndash195

65 Asch E Bujak D Weiss M Peterson M Weinstein ALyme disease an infectious and postinfectious syndromeJ Rheumatol 199421454ndash461

66 Fallon B Sackheim H Keilp J et al Double-blindplacebo-controlled retreatment with IV ceftriaxone forLyme encephalopathy clinical outcome In 10th Interna-tional Conference on Lyme Borreliosis and Other Tick-Borne Diseases September 11ndash15 2005 Vienna Austria2005196


DOI 10121201wnl00002655176697628 published online May 23 2007Neurology

J J Halperin E D Shapiro E Logigian et al Neurology

review) Report of the Quality Standards Subcommittee of the American Academy of Practice Parameter Treatment of nervous system Lyme disease (an evidence-based

This information is current as of May 23 2007

Online ISSN 1526-632X1951 it is now a weekly with 48 issues per year Copyright All rights reserved Print ISSN 0028-3878

reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology

ServicesUpdated Information amp

97628citationfullhtmlhttpnneurologyorgcontentearly2007052301wnl000026551766including high resolution figures can be found at

Supplementary Material

697628DC1httpnneurologyorgcontentsuppl2007062701wnl00002655176Supplementary material can be found at

Citations

97628citationfullhtmlotherarticleshttpnneurologyorgcontentearly2007052301wnl000026551766This article has been cited by 12 HighWire-hosted articles

Subspecialty Collections

httpnneurologyorgcgicollectionperipheral_neuropathyPeripheral neuropathy

httpnneurologyorgcgicollectionmeningitisMeningitis

httpnneurologyorgcgicollectionencephalitisEncephalitis

httpnneurologyorgcgicollectioncranial_neuropathyCranial neuropathy

httpnneurologyorgcgicollectionbacterial_infectionsBacterial infections

httpnneurologyorgcgicollectionall_pediatricAll Pediatricfollowing collection(s) This article along with others on similar topics appears in the

Permissions amp Licensing

httpnneurologyorgmiscaboutxhtmlpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in

Reprints

httpnneurologyorgmiscaddirxhtmlreprintsusInformation about ordering reprints can be found online



2 If different regimens are preferred for differentmanifestations of neuroborreliosis

3 What duration of therapy is needed

DESCRIPTION OF THE ANALYTIC PROCESS Inthe spring of 2004 the Quality Standards Subcom-mittee (QSS) of the American Academy of Neurol-ogy (AAN) convened an expert panel ofinvestigators from the United States and Europewho have published extensively in the field Thepanel was selected to represent a broad range of rel-evant expertise and opinion

In May 2004 a literature search was performed(all languages) using OvidMEDLINE Pubmed andEMBASE using search terms ldquoLyme Disease[DrugTherapy Therapy]rdquo ldquoBorrelia Infections[DrugTherapy Therapy]rdquo ldquoBorrelia burgdorferi groupand (borreliosis or Borrelia or neuroborreliosis)rdquoand ldquoAnti-Infective Agents[Therapeutic Use] and(antibiotic$ or antimicrob$ or anti-microb$)rdquo Thisresulted in 353 citations After elimination of dupli-cate citations each abstract was reviewed by at leasttwo members of the panel for relevance for furtherreview Any disagreements were arbitrated by athird reviewer This resulted in a list of 112 articleseach of which was then reviewed by at least twomembers of the panel Members of the panel recom-mended adding 10 additional references After de-tailed review of all 122 the panel decided 37 articlescontributed relevant assessable data Articles wereexcluded if they did not address treatment of neu-roborreliosis were not peer reviewed or were solelyreview articles The selected articles were then re-viewed in detail by all panel members to assess thequality of the evidence contained

Studies were divided into three groups adultLyme disease pediatric Lyme disease and post-Lyme syndrome Each article was reviewed to deter-mine if it specifically addressed treatment of neu-roborreliosis and if it contained original dataThose that were relevant were then graded as ClassI through IV using standard criteria as listed inAppendix 2 An evidence table was constructed list-ing each study its class the treatment regimens as-sessed whether it was prospective or retrospectivewhether it was blinded or open whether it was con-trolled or not whether it used explicit or objectiveresponse criteria the number of subjects the dura-tion of observation the completeness of follow-upand the outcomes

Overall four studies5-747 were Class I (three inpost-Lyme syndrome) One47 performed in chil-dren was considered Class I with regard to its pre-determined outcome measure CSF antibiotic levelsbut this study did not discuss clinical outcomes

Four studies were Class II (three in adults with neu-roborreliosis151819 one in children48) All were ratedClass II with regard to at least one of their predeter-mined objective measures of disease activity ELISACSF cell count or culture all of which were appar-ently measured in masked fashion All four of thesestudies would be considered Class III with regard toclinical outcomes for which assessments were notmasked All other studies were Class III or IV

ANALYSIS OF THE EVIDENCE When Lyme borre-liosis affects the nervous system it typically presentswith (a) all or part of a triadmdashmeningitis cranialneuritis and radiculoneuritis (known in Europe asGarin-Bujadoux-Bannwarth syndrome) (b) paren-chymal inflammation of the brain or spinal cord (c)mild radiculoneuropathy presenting as a more dif-fuse predominantly sensory peripheral neuropa-thy34 or (d) encephalopathy (alteration of cognitivefunction of varying severity with or without evi-dence of brain infection) Most well performedstudies have focused on (a) the group in which thediagnosis is most clear-cut and treatment responseis most straightforward to assess

Parenchymal CNS involvement is quite rare andstudies of treatment of these individuals havelargely been anecdotal (Class IV) Similarly only alimited number of small studies have addressed (c)or (d) all are Class III or IV

A separate entity defined differently by differentauthors often referred to as ldquopost-Lyme syn-dromerdquo occurs in patients who have had Lyme dis-ease but after treatment that would normally beexpected to be effective have continued to have re-sidual chronic symptoms including one or more ofthe following musculoskeletal pain (without frankarthritis fibromyalgia-like) fatigue and ldquoneuro-psychiatricrdquo symptoms The latter typically consistof perceived memory or cognitive difficulty irrita-bility sleep disturbance depression headache limbor other paresthesiasmdashall in the absence of clinicalor laboratory evidence of focal or inflammatorycentral or peripheral nervous system involvement5-7

Thus this entity is often included in discussions ofneuroborreliosis even though there is no evidenceof CNS infection in such individuals Although onlythree published studies have addressed this disor-der all have been Class I

ADULT NEUROBORRELIOSIS Long before thecharacterization of Lyme disease anecdotal re-ports89 indicated that erythema migrans-associatedmeningitis was responsive to penicillin The firsttreatment trial10 published in 1983 compared out-comes in 12 US patients with Lyme meningitis


















































Oral regimens


B



C



C

Parenteral regimens



B



B
































RECOMMENDATIONS














DISCLAIMER
































































































Citations











Reprints




















































Oral regimens


B



C



C

Parenteral regimens



B



B
































RECOMMENDATIONS














DISCLAIMER
































































































Citations











Reprints

































RECOMMENDATIONS














DISCLAIMER
































































































Citations











Reprints












DISCLAIMER
































































































Citations











Reprints























































































Citations











Reprints




practice parameter: treatment of nervous system lyme disease (an

Documents