1 pneumocystis carinii · inthis mitochondrial ribosomal rnafrom pcarinii article, the first of two...

Thorax 1992;47:305-314

AIDS and the lung: update 1992 1

Pneumocystis carinii pneumonia

Robert F Miller, DavidM Mitchell

Department of

Medicine, University

College and Middlesex

School ofMedicine,Middlesex Hospital,London WIN 8AA

F Miller

Department of

Respiratory Medicine,

Mary's Hospital,London W2 INYM Mitchell

Reprint requests to:

F Miller

During the first 10 years of the AIDSpandemic the clinical features of thesyndrome have been extensively reported. ByJanuary 1992 there had been 5451 cases ofAIDS in the United Kingdom, 3391 patientshaving died, and 16 828 people were known tohave antibodies to the human immuno-deficiency virus (HIV), with the potential toprogress to AIDS. Pneumocystis cariniipneumonia, despite widespread use of effectiveprophylaxis, rema

opportunist pneumocystis pneumonia inpatients immunosuppressed by malignancy,chemotherapy, corticosteroids, or HIV resultsfrom reactivation from a focus of organismsin the lungs that may, by analogy withtuberculosis, be the residuum of asymp-tomatic, self healed childhood infection. Thisassumption has recently been challenged bythe application of molecular biology to thestudy of P carinii.

opportunist infection in patients with AIDS. Cloning ofDNA from PneumocystisIn the two years since the articles in the carin'i

series AIDS and the lung appeared in Thorax The unique properties of organisms arethere have been important developments in precisely expressed at a genetic level, beingthe clinical management of respiratory defined by the sequence ofnucleic acid bases incomplications of HIV infection, and there is their DNA (or RNA in certain viruses). Part ofalso a clearer understanding of the molecular the gene coding for the large subunit of thebiology of the infectious complications. In this mitochondrial ribosomal RNA from P cariniiarticle, the first of two updates, we review has been cloned and sequenced. P cariniirecent advances in the molecular biology and pneumonia was induced in the rat and DNAclinical management of patients with was extracted and cloned from parasite en-Pneumocystis carinii infection. riched fractions from the lungs.3 Identification

of recombinants of non-rat origin (that is frompneumocystis) was achieved by negative colony

Molecular biology ofPneumocystis hybridisation and probing with genomic ratcarinii pneumonia DNA. Recombinants of interest were tested inTwo morphological forms of P carinii are an in situ hybridisation assay with sequencesseen in infected lungs: thin walled, irregularly obtained from human and rat lungs infectedshaped, single nuclee ltropBiozoiteasT2- and not infected with P carinii.3 Sequences5 im), which adhere firmly to the external specific for P carinii were confirmed by theirsurface of typ pneumocytes, andhicK characteristic in situ hybridisation patterns,walecysts77- gm) containing four to eight which were typical of the morphology andsing e nucleated sporozoites, which lie freely distribution of P carinii infection in lungwithin the alveoli. Uncertainties surround the tissue.3 Recombinant plasmid PAZ102 wastaxonomy of P carinii. Its morphological selected as a candidate mitochondrial sequencefeatures, including the irregular shaped because of strong signals derived in hybridisa-trophozoites and the absence of motility tion studies on infected lung samples. Theorganelles; the silver staining of the cyst walls; plasmid PAZ102 (an insert of 570 base pairs)the chemotherapeutic response to anti- was sequenced and compared with several dataprotozoal drugs, including pentamidine, bases. From sequence data on PAZ102 andthough not to antifungal drugs such as comparative analysis of data bases the fragmentamphotericin B; and the failure to culture the was identified as a portion ofthe gene coding fororganisms in various media have led to the the large subunit of mitochondrial ribosomalview that the organism is either a protozoon or RNA ofP carinii.a fungus. Oligonucleotide primers used for the

Serological studies using indirect fluores- polymerase chain reaction were constructedcence, enzyme linked immunosorbent assays, from two sequences of moderate conservationand Western blotting (immunoblotting) have that were specific to P carinii-PAZ102-E:shown that healthy children and adults, as 5'-GATGGCTGTTCCAAGCCCA-3' andwell as those with pneumocystis pneumonia, PAZ102-H: 5'-GTGTACGTTGCAAAG-have a high frequency of serum antibodies to ,TACTC-3'. These oligonucleotides producedthe organism.2 This tas led to the suggestion amplification ofspecific sequences from both ratthat asymptomatic infection occurs in child- and human bronchoalveolar lavage sampleshood. It has also been widely assumed that infected with P carinii, but not from samples

305

on June 17, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.47.4.305 on 1 April 1992. D

ownloaded from

http://thorax.bmj.com/

Miller, Mitchell

PAZ1 02-Li :5' ATAAGGT GAGG AGTCGA G AG -3'

PAZ102-L2:5' ATAAGGT AGAT AGTCGA A AG -3'

Comparative sequences ofPAZ 102-Li rat specific oligonucleotide (top) andPAZ 102-L2, which is human specific. The sequences of base pairs within the boxed areasare identical. Outside the boxes there are five base pair differences in the twooligonucleotides.

infected with a range of other organisms,including Candida albicans, Mycobacteriumtuberculosis, Cryptococcus neoformans, andAspergillus nidulans.An internal oligonucleotide, PAZ102-L1,

derived from rat P carinii produced stronghybridisation signals from Southern hybridisa-tion with amplified products from infected ratlungs but only very weak signals with productsfrom infected human lungs.4 Direct sequencingof the amplified products from both rat andhuman samples showed limited but consistentdifferences between tfrom these two hosts,5 which included fivebase chanLges in the sequence of the internaloligonucleotide PAZ102-LI (figure). Anoligonucleotide specific to the human derivedorganisms was constructed, PAZ102-1 2. Thisshowed strong hybridisation with the amplifiedproduct from human P carinii but weakhybridisation with the rat P carinii amplifiedproduct,4 indicating that natural infections inman and in the rat are due to different strains of1 carznuz. A -ubh4u~11tudyhas aerivedestimates of the sensitivity of DNA amplifica-tion methods by seeding homogenised humanlung with P carinii organisms derived from theimmunosuppressed rat model. A band ofamplified DNA, made visible by ethidiumbromide staining, was found to indicate thepresence of 100 organisms or more. A bandvisible only with Southern hybridisationdenoted smaller numbers of organisms, downto a lower limit of sensitivity of one or twoorganisms per sample.6

IsP carinii a reactivation or a reinfection?Studies using monoclonal antibody bindingto screen postmortem lungs from immuno-competent and immunosuppressed patientswho had died of causes other than pneumoniafailed to detect P carinii and challenged thetheory of carriage of latent parasites in normallungs.7 The theory was tested further by usingDNA amplification with PAZ102-E andPAZ102-H and Southern hybridisation usingPAZ102-L2 to search for P carinii inpostmortem lung samples from 15 non-immunosuppressed individuals, who haddied from various causes; oneprimary lung disease.8 N P carinii specificDNA amplification product was detected eitherby ethidium bromide staining or by Southernhybridisation in any of the specimens.8Both silver staining and DNA amplification-

oligoblotting were used to detect P carinii frombronchoalveolar lavage fluid in 47 patients; 37were immunosuppressed by HIV or by treat-ment for lymphoma, leukaemia, or vasculitisand 10 were immunocompetent and under-

going bronchoscopy for investigation ofvarious respiratory diseases. Of 15 inmmuno-suppressed patients who had diagnoses otherthan P carinii pneumonia (diagnosed byconventional silver staining), including pul-monary Kaposi's sarcoma and mycobacterial orbacterial infection, n had P carinii DNAdetected by amplification stainingand only three had low levels ofDNA6 detectedby oligoblotting.9 No P carinii DNA wasdetecteQ by eitheroligobtting in samples from the immuno-ccTmipetent group. A further study of 51episodes of acute respiratory illness in 47 HIVpositive patients undergoing bronchoalveolarlavage included 20 episodes of P cariniipneumonia and 31 respiratory episodes withother diagnoses. Silver staining gave negativeresults in all 31 episodes with other diagnoses.Seven of these patients had low levels ofP carinii DNA, detected by oligoblotting alone,and one patient with Staphylococcus aureuspneumonia had high levels of P carinii DNAdetectable in lavage fluid by ethidium staining.This patient had a past history of P cariniipneumonia and presented again 10 weeks laterwith a second episode of pneumocystispneumonia. Thus in these studies only 10 of46immunosuppressed patients had WeaR DNAsignals, implying low levels of ' carini,andstrong signals were seen almost e.aoithose wh P carznAt pne2 Qma

dettable P carinii in thesefour studies7'l0 does not support the view thatthere is longterm_a e of theorganism, aoing reactiution duringimmunosuppwssion. Unless there is someother site for dormant organisms (and no datasupport such a hypothesis), reinfection wouldseem a more likely cause ftorpneumocystispneumonia. Reports o ase clusters ofP cariniipneumonia MgnesttIldL nrizontal trns-

_PTM -Imission mnaybe rat ese data are

important for epidemiological studies andemphasise the possibility of an environmentalsource of infection.

Is Pneumocystis carnini a fungus or aprotozoon?Two studies have suggested a degree ofhomology between fungi and P carinii. Thesewerebased on analysis or 16S anc 5S riDosomalRNA sequences.'3"4 Chance similarities maybe observed, however, and misleadingconclusions are likely if only one DNAsequence is studied. More recently a large (6-8kilobase) fragment of P carinii mitochondrialDNA, which contains seven continuous genesequences coding for six proteins (parts ofapocytochrome b and NADH dehydrogenasesubunit I and all of subunits 2, 3, and 6 andcytochrome oxidase subunit 11) and aribosomal RNA gene (the small subunitof ribosomal RNA) have been cloned andsequenced.'5 Similarities between derivedamino acid sequences for these six proteins andfor sequences from three fungi (Neurosporacrassa, Aspergillus nidulans, and Leishmaniatarentolae) showed that P carinii had the same

306



/T


ownloaded from



degree of homology with the fungi as they didwith each er. An average of 60% similarltyto the fungi was seen, the closest being toNeurospora crassa. Protozoa were seen as adistinct and disparate group, showing only20% similarity.5 This study indicatesunequivocally that P hnb a greate

99mTc DTPA was shown by this study to be auseful diagnostic test. If further studies inlarger numbers ofpatients confirm that the scantime can be reduced to seven minutes, thistechnique could provide a rapid non-invasivescreening test for pneumocystis pneumonia.

degree of homology with fungi than with GALLIUM-67 SCANNING AND INDIUM-Illprota LABELLED HUMAN POLYCLONALplanning strategies for culture of the g nism IMMUNOGLOBULIN SCANNINGin vitr s. In view of Gallium-67 (67Ga) scanning has been used inthese close homologies with fungi it is interest- HIV positive patients with two distincting and perhaps paradoxical that antifungal problems-firstly, in patients with dyspnoea ordrugs are ineffective. cough, or both. and normal chest radingraphs

in whom pneumocystis pneumonia is suspectedInvestigation of Pneumocystis carinii and, secondly, in patients with abnormal chestpneumonia radiographs known to have bronchopulmonaryEmpirical treatment for pneumocystis iYaposi's sarcoma in whom infection is suspec-pneumonia in patients with typical symptoms, ed iaddition. Gallium-67 is useful in thischest radiographic abnormalities, and hypox- regard as it doesmntaccuuae in-Kaposi'saemia16 may no longer be justifiable as sarcoma, but does accumul rous infec-pneumocystis pneumona becomes less tive and inframmatory cond9itions.'81 The usec rmmon in relation to bacterial infectisai technique is limited, however,malignancy (including Kaposi's sarcoma) and by the prolonged protocol of' upto72 hours.as prior prophylaxis may modify its presenta- Human pojycional immunoglobulin labelledtion. Diagnosis of pneumocystis pneumonia with indium-i 11 ("'In) has been shown tocurrently relies on idencation of P carinii in localise sites of infection and inflammation. Inrespiratory tract samles, either broncno- rats immunosuppressed with steroid "'Inalveolar lavage fluid or induced sputum. We labelled immunoglobulin was better than 67Ganeed a good screening test to idenTify NiIV scanning for detecting pneumocystis pneu-positive patients with respiratory symptoms monia. "'In labelled immunoglobulin scans inwho may have pneumocystis pneumonia before rats with earlv pneumonia were abnormalbronchoscopy is performed. whereas most gallium scans remained normal;

rats with more advanced pneumonia shovmd aTECHNETIUM-99M LABELLED more intense diffuse intrapulmonary accumula-DIETHYLENETRIAMENEPENTA-ACETIC ACID tion with "'lIn labelled immunoglobulin thanTechnetium-99m labelled diethylenetriamene- with 67Ga. In rats with bacterial or fungalpenta-acetic acid (99mTc DTPA) scanning was superinfection "'In labelled immunoglobulincompared with other non-invasive scans showed patchy focal accumulation ininvestigations including chest radiography, addition to the diffuse accumulation whereasrespiratory function tests (forced expiratory gallium scans showed no additional foalvolume in one second, forced vital capacity, changes.20transfer factor, and carbon monoxide transfer "'In labelled immunoglobulin scanaing wascoefficient) and arterial oxygen saturation at used to i'iivse 13H positive atientsrest and during exercise in 13 HIV positive presenting with acute respiratory eDisodes.2' Inpatients with pn mocys e ie wwithout (12 had bronchoscopy and 10-with m a d Pieudomonas aerugznosabronchitis-did not; three had endobronchial peneumona)whohanost raiographsKaposi's sarcoma, six bacterial pneumonia, one lI- labelled immunoglobulin scans showedtuberculosis, and three negative results in the diffuse intrapulmonary accumulation. Of eightinvestigations and self limiting episodes). 99mTc patients with interstitial changes on the chestDTPA aerosol was inhaled by patients sitting radiograph, "'In labelled immunoglobulinin front of a gamma camera for 45 minutes scans were negative in two with pulmonarybefore bronchoscopy. 17 Kaposi's sarcoma and showed diffuse intra-There wereno significant differences between pulmonary accumulation in six with pneumo-

the two groups in terms of lung function tests cystis pneumonia. Two further patients withor arterial blo s Lnsions at rest. inhaled pneumocystis infection (one had coinfection99MTc DTPA had a fastermej1earance4waf- with Staphylococcus aureus), both with diffusetime in t with pneumocystis mania interstitial and alveolar radiographic abnor-(7 2 minutes) an in t w (22 malities, had diffuse and focal intrapulmonaryminutes). The median arterial oxygen accumulation of "'In labelled immuno-desaturation during exercise was 5% in those globulin.with pneumocystis pneumonia and 2% in those "'In labelled immunoglobulin like 67Gawithout. 99mTc DTPA lung clearance was scanning can distinguish between infection andbetter than the other non-invasive tests in Kaposi's sarcoma, but takes less time and givesdiscriminating pneumocystis pneumonia from a lower radiation dose to the patient.other pulmonary problems; and scanning foronly seven minutes was just as specific and SPUTUM INDUCTIONsensitive for the diagnosis of pneumocystis Sputum induction is a well establishedpneumonia as the longer time of 45 minutes. technique for diagnosing pneumocystis

307



/T


ownloaded from


Miller, Mitchell

pneumonia22 and other infections23 in HIVpositive patients. The success rate variesconsiderably and not all centres have been ableto es'talbTih-iittfiqu2 Set'e-is e-pendson seveal factors. Selection of patients isimportant; if only those with a high index ofsuspicion or pneumocystis pneumonia arestudied then the yield is high."" Patientsneed careful preparation an! exvcriencedte'c nicians increase the success rate (so thatsamples representative of the lower respiratorytract are obtained).23Adverse events have been'reported. Arterial

desaturation to 92% or less occurred andpersisted for up to 20 minutes aterthieprocedure in 11 of 20 HIV positive patientswith pneumocystis pneumonia and nine of 21patients with other diagnoses.24 Neither thedegree of theraRdigrfic abnormality, thealveolar-arterial oxygen gradient, nor baselineoxygen saturation identified patients whodeveloped desaturation.24 Saturation fell tobelow 84% less than 10 minutes after the startof induction in two patients; the procedure wasabandoned in these two and in two others, whodeveloped nausea and retching. The fall inoxygen saturation was attributed to abnor-malities of ventilation and perfusion ratiosrather n to hypov-enilation. 'l'he absenceof desaturation seen in the control groupsuggested inpatients with abnormal lungs. The persistenceof desatfor 0 minutes after jbeproc ioceds,such as exercise testing, are to be carried outsoon after spu tion.

TECHNIQUES TO INCREASE THE DIAGNOSTICYIELD FOR PNEUMOCYSTIS CARINII FROMINDUCED SPUTUM AND BRONCHOALVEOLARLAVAGE FLUIDMonoclonal antibodiesSeveral monoclonal antibodies specific forP carinii are now available. Some recognisecysts (R13/3G4-6, 5E12) and others both cystsand trophozoites (2E3, 3FB, 2G2). Used inimmunofluorescence tests, either singly or incocktails, these antibodies have proved to bepowerful diagnostic tools. Ng et al comparedGrocott-Gomori, toluidine blue-O, Diff-Quick(a Giemsa like stain), and immunofluorescencetests on 182 specimens (128 of induced sputumand 56 of bronchoalveolar lavage fluid, onetransbronchial biopsy specimen, and one openlung biopsy specimen). Use of indirect fluores-cence antibodies (IFA) was better than all fourconventional stains. P carinii was detected byIFA in 17 cases where the four other stains gavea negative result.25

Several diagnostic kits incorporatingdifferent monoclo 'bodies are

commercially available. Use of e tsincreases the sensitivity and specificity ofinduced sputum or bronchoalveolar lavage fluidfor diagnosis and reduces the time spent by staffscanning samples. Grocott staining may mis-identify Histop scoccus neoformans as Pneumocystis carinii,whereas immunofluorescence tests sub-stantially reduce this risk.26

DNA amplificationBronchoalveolar lavage fluid Silver stainingand DNA amplification were compared for thediagnosis of Pneumocystis carinii in 37immunosuppressed patients (33 of whomwere HIV positive) presenting with acuterespiratory episodes.9 DNA amplification withethidium bromide staining detected P cariniiDNA inai stainpositive. In six other patients, with negativesilver stains, whose presentation and chestradiographs were typical of pneumocystispneumonia DNA amplification gave a positiveresult in four; the other two patients werethought to have non-specific or lymphocyticpneumonitis.Induced sputum DNA amplification and silverstaining were used with bronchoalveolar lavagefluid and oducedspumdofrespiratory illness in 47 HIV positivepatients.'0 Silver staining gave a positive resultin 14 (70%) bronchoalveolar lavage specimensand only seven (35%) induced sputumsamples, whereas DNA amplification withethidium bromide staining was positive in 19(95%) lavage samples and 18 (90%) inducedsputum samples. Interestingly, DNA amplifi-cation gave a positive result for severalind at were"inaate' as they containedondyorai andpharyngeal debris that was silver stainnegative. This raises the possibility that DNAamplification applied to saliva may provide ameans of diagnosing P carinii pneumonia.

Treatment ofPneumocystis cariniipneumoniaIntravenous high dose co-trimoxazole remainsthe "gold standard" treatment for pneumo-cystis pneumonia. Conventionally used dosesof co-trimoxazole (sulphamethoxazole 100 mgand trimethoprim 20 mg/kg a day) m eexcessive. A study has shown that dose reduc-tion to 75% of conventional doses, to maintainserum nrimeloprim wgl wasassociated with a be'ter txicty profile but noreduction inefficawith co-trinrxixazole, dapsone-trimethoprim,and the substituted naphthaquinolone BW566have been shown to be effective in mild caseswhereas routine use of high dose adjunctiveglucocorticoids, in addition to parenteral anti-pneumocystis treatment, is now advocated forsevere episodes.The criteria used to stratify cases into mild,

moderate, and severe Pneumocystis cariniipneumonia are shown in table 1 andappropriate treatment is shown in table 2.

THERAPY FOR PNEUMONIA OF MILD TO

MODERATE SEVERITYDapsone-trimethoprimA double blind study of 60 patients with a firstepisode of pneumocystis pneumonia (mild to

moderate) compared 21 days' oral treatmentwith trimethoprim-sulphamethoxazole (20/100 mg/kg a day) with trimethoprim-dapsone(20 mg/kg a day and 100 mg a day).28

Oral treatment failed because of progressive

308



/T


ownloaded from



Table 1 Grading of severity of Pneumocystis carinii pneumonia

Mild Moderate Severe

Symptoms and Increasing exertional Dyspnoea on minimal Dyspnoea at rest, tachypnoea at rest,signs dyspnoea with or exertion, occasional persistent fever, cough

without cough and dyspnoea at rest,sweats fever with or without

sweats

Blood gas tensions Pao2normal, Sao2 Pao2 8 1-11 kPa Pao2 < 80 kPa(room air) falling on exercise

Chest radiograph Normal or minor Diffuse interstitial Extensive interstitial shadowingperihilar shadowing shadowing with or without diffuse alveolar

shadowing ("white out") sparingcostophrenic angles and apices

Pao2-partial pressure of oxygen; Sao,-arterial oxygen saturation, measured with a transcutaneous oximeter.

Table 2 Treatment of Pneumocystis carinii pneumonia

Mild andmoderate Severe

Co-trimoxazole:Intravenous V VOral x

Dapsone-trimethoprim, oral xPentamidine:

Intravenous V VNebulised x

Clindamycin-primaquine xBW566 xGlucocorticoids x

-/Treatment feasible; x -treatment with this drug notshown to be effective, or data to support its use notavailable.For definitions of mild, moderate, and severe see table 1.

pneumonitis in only three of 30 patients giventrimethoprim-sulphamethoxazole and in onlytwo of 30 given trimethoprim-dapsone. Majorside effects necessitated a change to intravenouspentamidine in 17 treated with trimethoprim-sulphamethoxazole and only nine treated withtrimethoprim-dapsone. Severe biochemicalhepatitis occurred in six and severeneutropenia in five given trimethoprim-sulphamethoxazole and in one each receivingdapsone-trimethoprim; rash and nausea andvomiting occurred with equal frequency inthe two treatment groups. Asymptomaticmethaemoglobinaemia occurred in mostpatients receiving trimethoprim-dapsone,reaching 20% in only one patient; mild hyper-kalaemia (5- 1-61 mmol/l) occurred in 16 of thepatients treated with dapsone-trimethoprim.

In this study the two oral treatmentsappeared equally effective for the treatment ofmild to moderate pneumocystis pneumonia,dapsone-trimethoprim being better tolerated.This combination is now widely used in bothNorth America and the United Kingdom forthe treatment of mild to moderate pneumo-cystis pneumonia.

BW566The AIDS Clinical Trials Group (ACTG)protocol 167 compared oral BW566 at a dose of750 mg three times daily with oral co-trimoxazole (sulphamethoxazole 1600 mg andtrimethoprim 320 mg-that is, 75% ofconventionally used doses) three times dailyfor 21 days in 408 patients with mild tomoderate pneumocystis pneumonia; 322 had

cytologically confirmed disease and 86 had aclinical diagnosis (13th ACTG meeting,Washington, December 1991; unpublisheddata).One hundred and sixty patients received

BW566 (114 of whom had mild disease) and162 received co-trimoxazole (115 had milddisease). Response rates were identical at60% (96 responders to BW566 and 98 to co-trimoxazole). Toxicity occurred in 11 patientsreceiving BW566 and 31 receiving co-trimoxazole. Abnormalities in liver functiontests occurred in 7 5% of those receiving co-trimoxazole versus 0 5% of those receivingBW566, rash in 7% versus 3%, and nausea orvomiting in 6-5% versus 1%. The drug fever(5%) and leucopenia (3%) seen with co-trimoxazole did not occur with BW566. Thisstudy showed that BW566 had a better toxicityprofile than co-trimoxazole while showingequal efficacy.

Other second line drugsOther second line drugs that have shownefficacy in the treatment of pneumocystispneumonia, including piritrexin, trimetrexate-folinic acid, and clindamycin-primaquine,have not gained widespread use; there are fewdata comparing them with high dose co-trimoxazole or intravenous pentamidine.

TREATMENT FOR SEVERE PNEUMONIAGlucocorticoidsSome clinicians have suggested that anti-pneumocystis treatment may cause an initialrapid acceleration ofintrapulmonary inflamma-tion.Fuare often seen in the first three to five days oftreatment, even in patients with mild disease atpresentation.2930 Possibly the death ofP cariniiorganisms exacerbates intrapulmonary inflam-mation.0 If this is so, beneficial effects ofcorticosteroids might be expected if they arebegun before further lung inflammation hasoccurred.

Adjunctive treatment with corticosteroidsfor severe AIDS associated Pneumocystiscarinii pneumonia was initially advocated as ameans of preventing death in patients whodeveloped respiratory failure.29 Othersbelieved that corticosteroids were contra-indicated because other life threatening

309



/T


ownloaded from


Miller, Mitchell

opportunist infections would increase andKaposi's sarcoma would spread more rapidly.In May 1990 the United States NationalInstitutes of Health convened a consensuspanel that reviewed data from five randomisedtrials.3' In four studies steroids were part ofinitial treatment and in the fifth study they wereadded only when treatment failed. One studywas larger32 and so was given greater weightand the other studies were used to support orcontrast with data from this study. The detailsof the studies are given in table 3.

Study 1 The Californian CollaborativeTreatment Group study enrolled patients withproved (n = 220) or presumed (n = 31)pneumocystis pneumonia in an open, random-ised study. Entry criteria were a hypoxaemiaratio (the ratio of arterial oxygen pressure to thefraction of inspired oxygen) of 75 or more andtreatment for pneumocystis pneumonia of upto 36 hours. Prednisolone was started at 40 mgtwice daily and continued at reducing doses for21 days. Three end points-failure of oxygena-tion (a hypoxaemia ratio below 75), the use ofmechanical ventilation, and death-werereduced by half in those receiving cortico-steroids. In patients who had the lowest levelsof oxygenation overall mortality after 31 dayswas reduced from 43% to 19%. Them faction appeared to e prevention of furtherdecline in oxygenation during the first threedays of treatment. Adverse effects of cortico-steroids included higher rates of oralcandidiasis and herpes simplex infection.

Study 2 This was an open, randomised studythat compared adjunctive intravenous methyl-prednisolone (0-5 mg/kg four times daily for 10days) with standard treatment alone for a firstepisode of pneumocystis pneumonia. Patientshad an arterial oxygen pressure (Pao2) of less

than 67-5 mm Hg (9 0 kPa) or a carbon dioxidepressure (Paco2) of less than 30 mmHg(4 0 kPa) at rest and had been having anti-pneumocystis treatment for up to 24 hours.Fifty three patients had been studied by thetime of the National Institutes of Healthconsensus meeting, when this study wasterminated. Gastrointestinal haemorrhage andtwo episodes of acute psychosis occurred in thegroup treated with steroid. There were twodeaths (7%) in those given steroids and eightdeaths (32%) in those receiving standard treat-ment alone.

Study 333 This was a randomised, doubleblind study which compared reducing doses ofintravenous methylprednisolone (beginningwith 60 mg every six hours for eight days) withstandard treatment alone. Patients wereenrolled if their Pao2 was less than 51 mm Hg(6-8 kPa). Eleven patients were enrolled within48 hours of starting antipneumocystis treat-ment, and 30 others entered after 48 hours. Inthe 41 patients enrolled (19 received steroidsand 22 placebo) no benefit in survival after21 and 51 days was found. There were nodifferences in arterial oxygen pressure orradiographic appearances between thosereceiving steroid and placebo. Three oppor-tunist infections occurred in patients takingsteroid.

Study 434 This randomised, double blind,placebo controlled study enrolled 38 patientswho had either an oxygen saturation of85-90%at rest or a 5% fall in oxygen saturation onexercise. Prednisolone (60 mg daily) or placebowas given for 21 days, beginning within 48hours of the start of antipneumocystis treat-ment. In 42% of patients given placebo and 6%of those given prednisolone deterioration inoxygenation reached the study end point (a

Table 3 Studies of adjunctive corticosteroid treatment in Pneumocystis carinii pneumoniaj31~~~ ~ ~~~~~~~33 4,

No of patients

Randomised

Placebo controlled

Oxygenation entry criteria

251

Yes

No

Hypoxaemiaratio >75

53

Yes

No

Pao2 < 9 kPa (air)

41

Yes

Yes

Pao2 < 6 8 kPa (air)

,J144Y338

Yes

Yes

Sao, 85-90% on airor 5% decreaseon exercise

23

Yes

Yes

Pao2 < 10 kPa at Fio2= 35%

Maximum intervalbetween initiation ofantipneumocystis drugand corticosteroid(hours)

Corticosteroid treatment:Drug (route)Initial dose

Duration (days)

36

Prednisolone (oral)40mg bd

21

24

Methylprednisolone(IV) 0-5 mg/kg qds

10

Unlimited

Methylprednisolone(IV) 60 mg qds

8

48

Prednisolone (oral)60 mg daily

21

72

Methylprednisolone(IV) 40 mg qds

7-10

Outcome (°/.. given corticosteroid v % given no steroid):Death 14/29 7/32 31/37 6/10 25/82

IPPV 15/33 11/44 NA NA 25/82

Subsequent fall in 17/37 NA NA NA NA

oxygenation

IPPV-intcrmittcnt positive pressure ventilation; bd-twice daily; Pao2-arterial oxygen tension; qds-four times daily; NA-not available; Sao2-

arterial oxygen saturation; Fio2-fractional inspired oxygen.

310

332 4 33 5 -1

1 31zStudv:



/T


ownloaded from



10% decrease in oxygen saturation). Patients of prednisolone by mouth and reduced to zeroinitially treated with adjunctive prednisolone over 10 days.also had longlasting improvements in exercisetolerance.

Study 535 This randomised, double blindstudy compared 40 mg of methylprednisoloneevery six hours with placebo for seven to 10days in patients with severe pneumocystispneumonia (defined by a respiratory rate ofmore than 30/min, an alveolar arterial gradientofmore than 30mm Hg (4-0 kPa), and a Pao2 ofless than 75 mm Hg (10-0 kPa) with 35%oxygen but higher than 60 mm Hg (8-0 kPa)with 100% oxygen). Steroids were startedwithin 72 hours of initial antipneumocystistreatment. The study was stopped after 23patients had been enrolled, because nine of the12 patients taking steroid survived until hos-pital discharge but only two of the 11 receivingplacebo did.

The conclusions of the consensus panel werethat adjunctive corticosteroid treatment clearlreduce , respiratoryfaiure, or deterioration in-oxyRgenation inpatients with moderate to severe pneumocystispneumonia. Only one study (study 3) did notshow benefit and this differed in that it allowedlate corticosteroid treatment (that is, rescuetreatment) whenever respiratory failureoccurred and most of the patients re;eivedcorticosteroid treatment more than 72 hoursafter specific antipneurnocystis treatment hadbeenstarted. The panel suggested, firstly, thatadjunctive steroids should be given to allpatients who presented with a Pao2 of 70mm Hg (8 3kPa)nr less or an alveolar-arterialoxygen gradient of more than 35 mm Hg (4 7kPa). The panel also felt that those patientswith mild hypoxaemia (Pao2 above70 mm Hg-9K K'a) t also benefit fromstQids, though it would be difficult to showthis benefit as there was a good outcome in thisgroup anyway. The panel suggested, secondly,that adjunctive glucocorticoid treatmentshould begin at the start of specific anti-pneumocystis treatment. In some patients thiswould be on a presumptive diagnosis ofpneumocystis pneumonia and clearly there wasa need to confirm the diagnosis rapidly in thesecircumstances. Thirdly, it was suggested thatprednisolone 40 mg twice daily should be usedfor the first five days, reducing to 40 mg on'edaily for the sixth to the 10th day and then20 mg daily for the 11th to the 9ays.ruternaiveuy sntraveous methylprednisoloneat 75% of the above doses was suggested. Thepanel emphasised that these regimens mightnot be better than others that had beenreported.

In the United Kingdom a regimen common-ly used is to give 1 g of methylprednisoloneintravenously on three consecutive days with-out tapering the dose. In those responding to

EFLORNITHINE FOR SALVAGE TREATMENTThirty one patients with pneumocystispneumonia (first episode in 27 patients andsecond episode in four) who had failed torespond to treatment with pentamidine, co-trimoxazole, or both were treated witheflornithine at a dose of 400 mg/kg a dayti-rnnously in divided doses. Twenty one ofthe 31 patients (68%) responded.36 Failure ofresponse to co-trimoxazole or pentamidine wasdefined as three or more of the following:(1) Pao2 below 9-0 kPa with the patient breath-ing air; (2) an increasing respiratory rate; (3) adeteriorating chest radiograph; (4) persistentfever after at least four days of treatment.Response to eflornithine was defined as survivalto discharge with an improved chest radio-graph and a normal respiratory rate and arterialblood gases during the breathing of air with nofever. Of the 31 patients receiving eflornithine,21 survived and 10 died. Three ofthe survivorsalso received methylprednisolone, as did two ofthose who died. All three patients who wereventilated and received eflornithine died.Adverse reactions included bone marrow sup-pression in 15 patients, severe enough in five torequire discontinuation of the drug. Sixteenpatients had infusion phlebitis. One patienthad an early relapse (within one month) despiteinstigation of secondary prophylaxis.

CONTINUOUS POSITIVE AIRWAYS PRESSUREVENTILATIONContinuous positive airways pressure ventila-tion by either nasal mask or face mask is usedas a means of improving oxygenation inhypoxaemic patients with pneumocystispneumonia. It acts, as in other conditions, byimproving ventilation-perfusion relationshipswitin the lungs, probably y nolding narrow,poorly compliant airways open.37 In patientswith deteriorating arterial oxygen tensions con-

Finuous positive airways pressure ventilationvia face mask improves oxygenation and "buystime"whilecttimiebialdaIsandalso ables discussion about whether totinue treatment to be undertaken with thepatient and next of kin.37

Prophylaxis ofPneumocystis cariniipneumoniaIn 1989 the United States Centres for DiseaseControl made recommendations for theprimary and secondary prophylaxis ofPneumocystis carinii pneumonia. These were

that primary prophylaxis should be given toHIV positive patients with CD4 (T helperlymphocyte) counts below 0 2 x 106/1 or toHIV positive patients with persistent, un-

explained fever or oral candidiasis irrespectiveof their CD4 count.38 Secondary prophylaxisshould be given to all HIV positive patientsafter an episode ofpneumocystis pneumonia. Itwas suggested at the time that prophylaxisshould be with either oral co-trimoxazole 960mg daily or nebulised pentamidine. The data

this regimen no rebound deterioration has beenreported.29 In patients showing only a partialresponse further doses may be beneficial-0*5 gmethylprednisorone may be given for fourdays, with the dose tapered thereafter to 40 mg

311



/T


ownloaded from


Miller, Mitchell

were insufficient at that time to support alter-native regimens such as the use of dapsone,dapsone and trimethoprim, Fansidar (sul-fadoxine and pyrimethamine), or intermittentintravenous pentamidine.The dose of pentamidine suggested was 300

mg of pentamidine isethionate given once amonth via a Respirgard II nebuliser. The basisfor this dosage came r en FranciscoCommunity Group prophylaxis trial.39 Thistrial recruited 8 patients for sec dary pro-phylaxis. wetomisdto reeive30 or 150 mg of pentamidine once a fortnight,or 300 mg once a month. All patients receivedpentamidine via a Respirgard II nebuliser.Eighteen months after randomisation there hadbeeoy22 neeofpn st eunniain the group treated with 30 mg ofpentamidineand 16 in those treated with 150 mg fortnightly,and in those treated with 300 mg once a monthonly eight patients had relapsed. The differencein relapse rate between those receiving 30 mgfortnightly and those receiving 300 mgmonthly was significant.At the 13th AIDS Clinical Trials Group

meeting in Washington in December 1991 theresults of protocol 021 were presented inadvance of publication. This protocol looked atsecondary prophylaxis and compared co-trimoxazole 960 mg once a day with nebulisedpentamidine 300 mg once a month via aRespirgard II nebuliser in 310 patients, whowere all receiving zidovudine. The studydesign was such that patients who were in-tolerant to either treatment were automaticallyswitched to the alternative for ongoing pro-phylaxis.:lpatients we.rerutd. dfweeks of an episode of pneumocystis pneu-monia and all had received less than two weeksof alternative prophylaxis before entering thestudy. Some patients who had previously re-acted adversely to high dose co-trimoxazolewere also included. The end point of the studywas time to relapse of pneumocystis pneu-monia, time to development ofextrapulmonarypneumocystosis (both cytologically confirmed),or survival time. Of the 310 patients recruited,154 received co-trimoxazole and 156 receivednebulised pentamidine. When the study endedthere had been 50 relapse the. gruPtrea -trimoxazole and 36 inthe group receiving nebuisepnm dine.Seven of the 14 patients in the co-trimoxazolearm had changed treatment because of adversereactions to co-trimoxazole and so were takingnebulised pentamidine when they relapsed.Reanalysis of the data according to the drugthat patients were taking at the time of relapseshowed that seven patients were receiving co-trimoxazole and 43 were receiving nebulisedpentamidine. In addition, the relapses thatoccurred were seen earlier in those receivingnebulised pentamidine. Mortality in the twogroups was similar: 43 patients in the co-trimoxazole arm died and 47 in the nebulisedpentamidine arm. Five ofthese deaths were dueto pneumocystis pneumonia (three patientswere having co-trimoxazole and two penta-midine). Change of treatment was due to

haematological or dermatological toxicity. The

study showed that there was no increase inzidovudine toxicity caused by simultaneoustreatment with co-trimoxazole at a dose of 960mg daily. Furthermore, patients who had hadadverse reactions to high dose co-trimoxazolefor treatment of pneumocystis pneumonia didnot have increased side effects while taking thedrug at a lower dose for prophylaxis.Some prophylaxis against cerebral toxoplas-

mosis in those patients VXX&_ying...co-trimoxazole was also apparent from this study.Te n lle inthe study, six in the group originally treatedwith co-trimoxazole and four in those receivingnebulised pentamidine. At the time the toxo-plasmosis developed only one patient wasreceiving co-trimoxazole; five other patientshad changed to nebulised pentamidine becauseof co-trimoxazole toxicity, so that nine patientswith toxoplasmosis were receiving nebulisedpentamidine at the time of presentation. Therewas an equal incidence of bacterial infections inthe two groups, but bacterial infectionsoccurred earlier in the patients treated withnebulised pentamidine.On the basis of this study the United States

Public Health Service Task Force on AIDSrecommended co-trimoxazole 960 mg daily forfirst line primagl secndary prophylaxis atthe end of 1991. on y neuipentamidine was recommended as second linetreatment, to be reserved for patients intolerantto the first line treatment. The task forceconsidered that the data were insufficient tosupport the use of alternative regimens, thoughthey acknowledged that drugs such as dapsone,with or without trimethoprim, were widelyused for both primary and secondaryprophylaxis.

Other concerns have been- raised about theuse of nebulised pentamidine for prophylaxisof pneumocystis pneumonia. Firstly, there areseveral reportsofaouin health care ,_rrsupervising patients receiving nebulisedpentamidine; these adverse reactions includecircumoral paraesthesia, bronchospasm, andgough.4I In addition, a progressive decline inthe transfer factor for carbon monoxide hasbeen reported in a respirator functiontechnician who worked in a room adjacent tothe one used for nebulising pentamidine.4'Yet serial transfer factor measurements inpatients having nebulised pentamidine showno sequential decline. These problems may bereduced by ensuring that nebulisation occurs inan isolated area,pre y in a dedicated rwith an extra fan, , onceinstructed in the technique, are allowed tonebulise pentamidine in their own home.

In areas of high prevalence of tuberculosisthere is concern that nebulisation may increasethe risk of the socomial transmission oftuberculosis becau ringeeprocedure, thus g ting droplet aerosol thatis dispersed into the environment.As this form of prophylaxis is not systemic

and there is minimal or no systemic absorptionof the drug there is a risk that it may not

suppress the development of extrapulmonary

312



/T


ownloaded from



pneumocystosis. In the largest reported seriesof patients with extrapulmonary pneumo-cystosis seven of the eight patients hadreceived prior prophylaxis with nebulisedpentamidine.42

Concerns have been expressed about thelong term toxicity to the lungs of patientsreceiving this drug by the aerosol route. Whenused for nebulisation in water pentamidine issupersaturated, acidic, and hypo-osmoiar.43There are reports of the drug crystallising outwiRthin the lung,' inducing a granulomatousresponse,18 or being associated with apicalbullous change and accelerated emphysema.45Compliance with long term treatment with

nebulised pentamidine is also thougt to bepoor.-" ln a study of primary and secondary-prophylaxis for pneumocystis pneumonia withnebulised pentamidine at a dose of 150 mgfortnightly compliance (defined as attendancewithin three days of the prescribed date fortreatment) was only 23%.' This study did notcompare the compliance rate with that ofpatients receiving co-trimoxazole, dapsone, orFansidar.

EFFECTS OF NEBULISED PENTAMIDINE ON THEDIAGNOSTIC YIELD FROM BRONCHOALVEOLARLAVAGE FLUID AND INDUCED SPUTUMA retrospective study of52 patients undergoingfibreoptic bronchoscopy to diagnosepneumocystis pneumonia analysed the yieldfrom lavage in those who had and had notreceived prior nebulised pentamidine prophy-laxis. Twehtty-M-patients- iornebulised pentamidine.47 The yield oTpneumocystsis from bronchoalveolar lavagefluid was in these patients and 100% inthe 31 pareceived priornebuisecd pentamine. 1ransronchial biopsyspecimens were positive for P carinii in 13of 16 patients (81%) who had received priorne _d ine and in 16 of 19 patients(e4g) who had not. Five clumps of P carinii

de were seen on average for those whohad received prior nebulised pentamidine and20 for those with no prior prophylaxis withpentamidine.47This study also compared the radiographic

appearances in those patients who had receivedprior nebulised pentamidine and in those whohad not. Radiographic changes were "typical"in 28 patients (90%) who had not received priorpentamidine but in only 11 (52%) who hadreceived pentamidine. Two patients (7%) withno prior prophylaxis and eight patients (38%)who had received pentamidine had predomin-antly upper lobe infiltrates. In addition, cysticchanges were seen in five patients receivingprior prophylaxis and in only one patientwith no prior prophylaxis; two patients whopresented with pneumothorax had receivedpentamidine. The authors concluded thatnebulised pentamidine prophylaxis reducedthe diagnostic yieId irom bronchoalveolarlavage and increased the frequency of atypicalradiographic appearances.A retrospective analysis was carried out on

the results of 348 induced sputum-proceduresperformed during the itial invesgations tor

possible pneumocystis pneumonia in a mixedpopulation of HIV positive adults and chil-dren. Twenty nine patients were excluded asthey were having prophylaxis with oral orintravenous regimens. Of the remaining 319patients, 126 were receiving prophylaxis withnebulised pentamidine prophylaxis at the timeof the sputum induction and 193 patients werereceiving no prophylaxis at the time of theprocedure.48 Seventy two of the 319 episodeswere ultimately found to be due to pneumo-cystis pneumonia, on the basis of sputuminduction in 54, bronchoalveolar lavage in 16,percutaneous lung biopsy in one, and necropsyin one. Of these 72 episodes, 44 occurred inpatients receiving no prophylaxis and 28 inpatients receiving prior prophylaxis withnebulised pentamidine. These two groups weresimilar in age, duration of symptoms, andarterial-alveolar oxygen gradient at the time ofsputum induction. The mean CD4 counts weresignificantly lower (0-2 x 106/1) in those who hadreceived prior nebulised pentamidine than inthose with no prior prophylaxis (0-68 x 106/1).48

Five patients receiving no prophylaxis failedto produce a-sample otsputum. Thirty six ofthe remaining 39 patients (92%) produced asamp positive for P carinii, compared withonly 18 of 28 patients (64%) who had receivedprior nebulised pentamidine. In those patientswho had received no prior prophylaxis chestradiographs were normal or typical in 65%;showed upper lobe predominance ot intiltrare in14%; and showed other appearances, includingf6cal (but not upper lobe) changes or cystic orcavitary abnormalities, in 2%. In those whohad received prior nebulised pentamidineradi(graphs were normal or typical of pneu-mocystis pneumonia in 41% and showed upperlobe predominance of infiltrates in 37% andfocal or cystic-cavitary changes in 22%. Theconclusions of this study were that there was asignificant reduction in the yield from inducedsputum for the diagnosis of pneumocystiispneumonia in patients who had received priorpentamidine and in this group a larger propor-tion of patients had upper lobe predominanceof infiltrates on the chest radiograph.48

ConclusionsThe techniques of molecular biology havealready begun to change much of ourunderstanding of the basic biology and lifecycle ofP carinii. Application of the techniqueof DNA amplification to specimens ofbronchoalveolar lavage fluid and inducedsputum have been shown to give a significantlybetter yield from these specimens for thediagnosis of pneumocystis infection thanconventional silver staining. Possibly in thefuture even less "invasive" samples, such assaliva, will be used to diagnose pneumocystispneumonia.

Patients with mild to moderate pneumoniamay be given oral treatment. Dapsone-trimethoprim appears to be as effective as oralco-trimoxazole and better tolerated bypatients. The new agent BW566 also appearspromising, with efficacy similar to that of co-trimoxazole and with a better toxicity profile.

313



/T


ownloaded from


Miller, Mitchell

Glucocorticoid treatment is now well estab-lished for patients with severe pneumocystispneumonia and should be offered to all patientswith suspected or proved pneumocystispneumonia who have an arterial oxygen tensionof 8-0 kPa or less in addition to parenteral co-trimoxazole or pentamidine. Primary prophy-laxis for pneumocystis infection should beoffered to HIV positive patients with CD4counts below 02 x 106/1 and to those withunexplained persistent fever or oral candidiasisregardless of their CD4 count; and secondaryprophylaxis should be offered to all patientsafter an episode of pneumocystis pneumonia.The prophylactic agent of choice is daily oralco-trimoxazole.

1 Peglow SL, Smulian AG, Linke MJ, Pogue CL, Nurre S,Crisler J, et al. Serologic responses to Pneumocystis cariniiantigens in health and disease. J Infect Dis 1990;161:296-306.

2 Wakefield AE, Steward TJ, Moxon ER, Marsh K, HopkinJM. Infection with Pneumocystis carinii is prevalent inhealthy Gambian children. Trans R Soc Trop Med 1990;84:800-2.

3 Wakefield AE, Hopkin JM, Burns J, Hipkiss JB, Stewart TJ,Moxon ER. Cloning ofDNA from Pneumocystis carinii. JInfect Dis 1988;158:859-62.

4 Wakefield AE, Pixley FJ, Banerji S, Sinclair K, Miller RF,Moxon ER, et al. Amplification of mitochondrialribosomal RNA sequences from Pneumocystis cariniiDNA of rat and human origin. Molec Biochem Parasitol1990;43:69-76.

5 Sinclair K, Wakefield AE, Banerji S, Hopkin JM.Pneumocystis carinii organisms derived from rat andhuman hosts are genetically distinct. Molec BiochemParasitol 1991;45:183-4.

6 Peters SE, Wakefield AE, Banerji S, Hopkin JM. Quantifica-tion of the detection of Pneumocystis carinii by DNAamplification. Molecular and Cellular Probes (in press).

7 Millard PR, Heryet AR. Observations favouring Pneumo-cystis carinii pneumonia as a primary infection: a mono-clonal antibody study on paraffin sections. JPathol 1988;154:365-70.

8 Peters SE, Wakefield AE, Sinclair K, Millard PR, HopkinJM. A search for Pneumocystis carinii in post mortem lungsby DNA amplification. JPathol 1992;166:195-8.

9 Wakefield AE, Pixley FJ, Banerji S, Sinclair K, Miller RF,Moxon ER, et al. Detection of Pneumocystis carinii withDNA amplification. Lancet 1990;336:451-3.

10 Wakefield AE, Guiver L, Miller RF, Hopkin JM. DNAamplification on induced sputum samples for diagnosis ofPneumocystis carinii pneumonia. Lancet 1991;337:1378-9.

11 Jacobs JL, Libby DM, Winters RA, Gelmont DM, FriedED, Hartman BJ. A cluster of Pneumocystis cariniipneumonia in adults without predisposing illnesses. NEngl JMed 1991;324:246-9.

12 Rucbush TK, Weinstein RA, Bachner RL. An outbreak ofpneumocystis pneumonia in children with acute lympho-cytic leukaemia. Am J Dis Child 1978;132:143-48.

13 EdmanJC, Kovacs JA, Masur H, Santi D, Elwood H, SoginM. Ribosomal RNA sequences show Pneumocystis cariniito be a member of the fungi. Nature 1988;334:519-22.

14 WatanabeJI, HoriH, Tanabe K, Nakamura Y. 5S ribosomalRNA sequence ofPneumocystis carinii and its phylogeneticassociation with Rhizopoda/Myxomycota/Zygomycotagroups. JProtozool 1989;36:17-9S.

15 Pixley FJ, Wakefield AE, Banerji S, Hopkin JM. Mito-chondrial gene sequences show fungal homology forPneumocystis carinii. Molec M.icrobiol 1991;5:1347-51.

16 Miller RF, Miller AB, Weller IVD, Semple SJG. Empiricaltreatment without bronchoscopy for Pneumocystis cariniipneumonia in the acquired immunodeficiency syndrome.Thorax 1989;44:559-64.

17 Robinson DS, Cunningham DA, Dave S, Fleming J,Mitchell DM. Diagnostic value oflung clearance of'9Tc-DTPA compared with other non-invasive investigationsin Pneumocystis carinii pneumonia in AIDS. Thorax1991;46:722-6.

18 Birley HDL, Buscombe JR, Griffiths MH, Semple SJG,Miller RF. Granulomatous Pneumocystis cariniipneumonia in a patient with the acquired immuno-deficiency syndrome. Thorax 1990;45:769-71.

19 Miller RF. Nuclear Medicine and AIDS. Eur J Nucl Med1990;16: 103-18.

20 Fishman JA, Strauss HW, Fischman AJ, Nedelman M,Callahan R, Khaw B-A, et al. Imaging of Pneumocystiscarinii pneumonia with "'In-labelled non-specific poly-clonal IgG: an experimental study in rats. Nucl MedCommun 1991;12:175-87.

21 Miller RF, Buscombe J, Grant A, Oyen W, Lui D, ClaessensR, et al. Investigation of acute respiratory illness in HIVpositive patients using indium labelled polyclonalimmunoglobulin scans. Thorax (in press).

22 Lcigh TR, Parsons P, Hume C, Husain OAN, Gazzard B,

Collins JV. Sputum induction for diagnosis of Pneumo-cystis carinii pneumonia. Lancet 1990;ii:205-6.

23 Miller RF, Kocjan G, Buckland J, Holton J, Malin A,Semple SJG. Sputum induction for the diagnosis ofpulmonary disease in HIV positive patients. JInfect 1991;23:5-15.

24 Miller RF, Buckland J, Semple SJG. Arterial desaturation inHIV positive patients undergoing sputum induction.Thorax 1991;46:449-51.

25 Ng VL, Yajko DM, McPhaul LW. Evaluation of an indirectfluorescent antibody stain for detection of Pneumocystiscarinii in respiratory specimens. J Clin Microbiol 1990;28:975-9.

26 Cregan P, Yamamoto A, Lum A, Van der Heide T,MacDonald M, Pulliam L. Comparison of four methodsfor rapid detection of Pneumocystis carinii in respiratoryspecimens. J Clin Microbiol 1990;28:2432-6.

27 Sattler F, Cowan R, Neilsen DM, Ruskin J. Triethoprim-sulfamethoxazole compared with pentamidine for treat-ment of Pneumocystis carinii pneumonia in the acquiredimmunodeficiency syndrome: a prospective non-crossoverstudy. Ann Intern Med 1988;109:280-7.

28 NMedina I, Mills J, Leoung G, Hopewell PC, Lee B, ModinG, et al. Oral therapy for Pneumocystis carinii pneumoniain the acquired immunodeficiency syndrome. N Engi JMed 1990;323:776-82.

29 Miller RF, Mitchell DM. Management of respiratory failurein patients with the acquired immunodeficiency syndromeand Pneumocystis carinii pneumonia. Thorax 1990;45:140-6.

30 Miller RF, Roberts CM. Intensive care management ofHIVpositive patients and patients with AIDS. Clin Intens Care1991;2:17-25.

31 Consensus statement on the use of corticosteroids as adjunc-tive therapy for pneumocystis pneumonia in the acquiredimmunodeficiency syndrome. N Engl J Med 1990;323:1500-4.

32 Bozzette SA, Sattler FR, Chiu J, Wu AW, Gluckstein D,Kemper C, et al. A controlled trial of early adjunctivetreatment with corticosteroids for Pneumocystis cariniipneumonia in the acquired immunodeficiency syndrome.N Engl JMed 1990;323:1451-7.

33 Clement M, Edison R, Turner J, Montgomery B, Luce J,Feigal D, et al. Corticosteroids as adjuvant therapy insevere Pneumocystis cariniipneumonia [abstract]. Am RevRespir Dis 1989;139:A250.

34 Montaner JSG, Lawson LM, Levitt N, Belzberg A,Schechter MT, Ruedy J. Corticosteroids prevent earlydeterioration' in patients with moderately severe

Pneumocystis carinii pneumonia and the acquiredimmunodeficiency syndrome (AIDS). Ann Intern Med1990;1 13:14-20.

35 Gagnon S, Boota AM, Fischl MA, Baier H, Kirksey DW,LaVoie L. Corticosteroids as adjunctive therapy for severePneumocystis carinii pneumonia in the acquired immuno-deficiency syndrome. N Engl JMed 1990;323:1444-50.

36 Smith D, Davis S, Nelson M, Youle M, Gleeson J, GazzardB. Pneumocystis carinii pneumonia treated with eflornith-ine in AIDS patients resistant to conventional therapy.AIDS 1990;4:1019-21.

37 Miller RF, Semple SJG. Continuous positive airway pres-sure ventilation for respiratory failure associated withPneumocystis carinii pneumonia. Respir Med 1991;85:133-8.

38 Girard PM, Pocidalo JJ, Murray JF. Primary prophylaxisagainst common infectious diseases in person with humanimmunodeficiency virus infection. Am Rev Respir Dis1991;143:447-50.

39 Leoung GS, Feigal DW, Montgomery AB, Corkery K,Wardlaw P, Adams M, et al. Aerosolized pentamidine forprophylaxis against Pneumocystis carinii pneumonia.N EnglJMed 1990;323:769-75.

40 Green ST, Nathwani D, Christie PR, Kennedy DH.Aerosolized pentamidine. Lancet 1989;ii:1284.

41 Gude JK. Selective delivery of pentamidine to the lung byaerosol. Am Rev Dis 1989;139:1060.

42 Coker RJ, Clark D, Clayton EL, Ainsworth JG, Lucas SB,Miller R, et al. Disseminated Pneumocystis carinii infec-tion in AIDS. J Clin Pathol 1991;44:820-3.

43 Miller RF, Steel SJ. Nebulised pentamidine as prophylaxisfor Pneumocystis carinii pneumonia. J AntimicrobChemother 1991;27:153-60.

44 McGowan I, Potter M, George RJD, Michaels L, ScaravilliF, Miller RF. HIV encephalopathy presenting as

hypomania: clinico-pathological conference. Genito-urinary-Med 1991;67:420-4.

45 Miller RF, Semple SJG, Lucas SB. Premature bullouspulmonary damage in AIDS. Genitourinary Med 1991;67:4-9.

46 West D, Steel 5, Miller R. Audit of Pneumocystis cariniiprophylaxis in Bloomsbury Health Authority. Int JPharmac Pract 1991;1:23-6.

47 Jules-Elysee KM, Stover DE, Saman MB, Bernard EM,White DA. Aerosolized pentamidine: effect on diagnosisand presentation of Pneumocystis carinii pncumonia. AnnIntern Med 1990;112:750-7.

48 Levine SJ, Masur H, Gill VJ, Fuerstein I, Suffredini AF,Brown D, et al. Effect of aerosolized pentamidine prophy-laxis on the diagnosis of Pneumocystis carinii pneumoniaby induced sputum examination in patients infected withthe human immunodeficiency virus. Am Rev Respir Dis1991;144:760-4.

314



/T


ownloaded from


1 pneumocystis carinii · inthis mitochondrial ribosomal rnafrom pcarinii article, the first of two...

Documents