Archives ofDisease in Childhood 1991; 66: 1359-1363

CURRENT TOPIC

Lyme disease in paediatrics

B Cryan, D J M Wright

The cluster of 39 children with a presumptivediagnosis of juvenile rheumatoid arthritis inConnecticut, around the town of Old Lyme,initiated the research which eventually led to thearthritis being named Lyme arthritis.' Theassociation with a multisystem illness thencaused the syndrome to be redefined as Lymedisease. Epidemiologically, an infectious aetio-logy appeared probable, with ticks as likelyvectors for the disease. Clinically, a skin rash,erythema migrans, preceded the developmentof the arthritis. The skin rash had been firstdescribed in Scandinavia in 1913 and associatedwith tick bites.2 Some patients presented withBannwarth's syndrome, a meningopolyneuritisassociated with tick bites, described in Europeas early as 1941.3 A possible causative organism,Borrelia burgdorferi, was isolated from a tick in1982.4 This hypothesis was substantiated whenthe spirochaete was grown from blood, skinlesions, and cerebrospinal fluid (CSF) ofpatients with Lyme disease, and the diseasewas reproduced in rabbits, fulfilling Koch'spostulates.5 6 Since then the disease has beenrecognised in many American states, NorthernEurope, Russia, China, and Japan.7 It has alsobeen reported in the southern hemisphere inAustralia.7

Department ofMedical Microbiology,Charing Cross Hospital,Fulham Palace Road,London W6 8RFCorrespondence to:Dr Wright.

Transmission

B burgdorferi is transmitted by the ticks of thegenus ixodes. Ixodes dammini is most frequentlyimplicated in the eastern USA. In the westernUSA and Japan I paciftcus predominates.I ricinus acts as the vector in Europe withI persulcatus in Asiatic and European Russia.8 9The nymphal ticks become infected byB burgdorferi while feeding on small mammalssuch as the white foot mouse, which act as

reservoirs of the spirochaete. The natural hostsof mature ticks are large mammals such as deer.Humans are incidentally bitten when they entera tick infected area, usually forest or scrubland.'0 Children playing in such areas would beat particular risk of infection. The frequency ofinfection after a tick bite depends on a numberof factors, such as the prevalence ofB burgdorferiinfection in the ticks, the duration of tickattachment, and the number of tick bites. In an

English study of forestry workers, 25% hadserological evidence of B burgdorferi exposure,and 5% symptoms attributable to Lymedisease. ' These workers may have had multiplebites whereas positive Lyme serology was

recorded in 20% of Austrian and 7% of Irish

populations who reported single bites.'2 13Another English study reported positive serologyin 12% of a population comprising farmers,gamekeepers, and forestry workers. 14

EpidemiologySince 1982 the number of cases of Lyme diseasein the USA has increased from 500 to over 4500in 1988. The highest incidence is in children ofless than 15 years and in adults of 25-44 years.'5Females were infected in 51% of cases. Dataon file at the Charing Cross Lyme DiseaseReference Laboratory records approximately600 cases per year or 0-39 cases per 100 000people. Most seropositivity is detected in middleaged patients, and the number of males is equalto that of females. In the eastern USA mostcases occur in the summer months,'0 15 whichfollows the period of maximal host seekingactivity of nymphal ticks in May and June. 16 Inthe western USA, cases are most common inspring. 5 At Charing Cross there are two peaks,one in spring and one in autumn. Most Irishcases have come to medical attention in theautumn.

Clinical featuresTypically, the clinical features of Lyme diseaseoccur in three stages.'7 There is considerableoverlap between stages of the disease, andfeatures of late Lyme disease may developwithout any history of early disease. One of the'unsolved mysteries' of Lyme disease is thatthere have been no outbreaks of arthritisreported since the initial 'cluster' in Connecticut.The characteristic rash, erythema migrans,

develops 2-30 days after the tick bite. Thisexpanding lesion is centred on the insect biteand spreads outwards over a period of weeks.It is seen in up to 50% of patients and isoccasionally due to multiple bites. 18 Non-specificconstitutional symptoms may occur at thisstage: malaise, anorexia, lethargy, arthralgia,headache, roseola skin rash, abdominal pain,cough, and sore throat are common complaints.Patients, especially children, may be feverish.The second stage of Lyme disease develops

1-4 months after infection. These patients tendto come to medical attention in the autumn orwinter. Lymphocytosis cutis occurs in 1% ofEuropean cases but has not been reported in theUSA. This lesion is a firm red to purple noduleor plaque. It develops on the ear lobe inchildren and on the nipple and areola in

1359

on 2 July 2018 by guest. Protected by copyright.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.66.11.1359 on 1 Novem

ber 1991. Dow

nloaded from

Cryan, Wright

adults.'8 Pathologically there is a dense dermalnodular lymphocytic infiltrate involving B andT lymphocytes. Infrequently, similar lympho-cyte infiltrates, at the base of the brain in Lymeborreliosis, have been misdiagnosed in childrenas B-cell lymphomas.

Central nervous involvement is common.Lymphocytic meningitis, cerebellar signs,cranial polyneuropathy, and radiculopathy havebeen reported. The differentiation from Guillain-Barre syndrome may be difficult, but in neuro-borreliosis the lymphocyte count is usuallygreater than 50 cells/mm3 and oligoclonal bandsmay be seen. Encephalitis, often clinicallyindistinguishable from herpes simplex encepha-litis, may occur. Outside the UK one mustconsider other arthropod-borne encephalo-pathies in such cases. In an Austrian study of 21children seropositive for Lyme disease, 17 hadcentral nervous system involvement, three hadarthritis, and one had skin symptoms.'9 In thisstudy and in a Swedish report aseptic meningitisand seventh nerve palsies were the commonestneurological features described.'9 20 Pain injoints and soft tissues and rarely asymmetricallarge joint arthritis may occur at this stage.These episodes are brief and recurrent. 17 A selflimiting myocarditis, which may necessitate theinsertion of a transvenous pacemaker, infre-quently presents during this phase of the illness.Bradycardia is not otherwise a feature of thedisease. Iritis or panophthalmitis has also beenreported, as has isolated inflammation of peri-pheral muscles. 17

In the third stage, joint disease is mostcommon. As many as 10% of patients mayprogress to a chronic erosive arthritis.' In anAmerican study of Lyme arthritis in 43children2' the mean age was 7 8 years and themale:female ratio 7:1. The knee was involved in41 patients, the elbow in eight, the hip or anklein four, and the wrist in two.2'At this stage in older patients demyelinating

encephalopathy, polyneuritis, and memoryimpairment may become manifest.' Acro-dermatitis chronica atrophicans is the character-istic skin rash of this phase of the disease.Initially this is a violaceous and often oedematousarea of skin in plaques or nodules, which lateratrophies. It resembles but is distinct fromscleroderma.'7 18 Acrodermatitis is unusual inchildhood Lyme borreliosis but at least twocases have been described.22 Eosinophilicfasciitis is a rare complication. 17As many as 30% of patients do not demon-

strate erythema migrans or other signs of earlydisease before presenting with late complica-tions.23 In European Lyme disease, joint mani-festations are reported to be less common thanin the USA. Conversely, in Europe nervousinvolvement tends to be more prominent.2425In North American Lyme disease, erythemamigrans is more often multiple than in Europe.Recent reports suggest that these differencesmay be less extreme.'3 26The isolation of B burgdorferi from the blood

has raised the possibility of transplacentaltransfer of the organism. To date, borreliaehave been isolated from one stillbirth and onenewborn infant but congenital abnormalities

resulting from Lyme disease during pregnancyhave not been unequivocally demonstrated.27The abnormalities that have been recorded wereassociated with serological evidence of infectionin retrospective studies.28

Laboratory diagnosisThe laboratory diagnosis of Lyme disease isdifficult. The organism has been grown frominfected tissues such as skin and body fluids,including CSF.29 Culture is an insensitivetechnique in most hands and not suitable forroutine diagnostic use. Visualisation of thepathogen in infected tissue has similar limita-tions.30 For these reasons serological techniqueshave been employed as routine diagnostic aids.

Indirect immunofluorescent assay (IFA)techniques were initially used to detect anti-bodies B burgdorferi but recently, as demand fortests has increased, enzyme linked immunosor-bent (ELISA) methodologies have becomepopular. Such tests can be automated, allowobjective test reporting, and are easier to stan-dardise than IFA methods.3'

Initial antibody responses are directed againstthe unsheathed flagellar proteins."' A rise inIgM antibodies directed against a 41 kDaflagellar antigen can be detected in most patientstwo to four weeks after infection. This responsepeaks at six to eight weeks. Assays which detectIgM response to such proteins are very usefulearly in the disease. ELISAs which use purifiedflagellae or flagella enriched preparations ofwhole B burgdorferi as antigens have increasedsensitivity in early disease.32 3 Later in thecourse of the illness antibodies directed againstproteins in the outer surface membranes (OSP),such as the 34 kDa OSP B or the 31 kDa OSPA, are detected.3' As B burgdorferi shares manyantigenic determinants with other bacteria, seramay contain antibodies which cross react in theIFA or ELISA assays described.3' 32 False posi-tive results can also occur in patients with con-nective tissue disorders.34 Therefore furtherdefinition of the immune response byimmunoblotting is a necessary confirmatoryprocedure if such tests provide positive results.Immunoblot patterns in children are similar tothose of adults.35 At the Charing Cross LymeDisease reference laboratory 11% of the ELISAspecimens that are weakly positive and a thirdof those that are strongly positive are confirmedby immunoblot. False negative results have alsobeen reported: as many as one in three patientswith erythema migrans are seronegative oninitial testing,23 and may become seropositiveon further testing. Antibiotic treatment early inthe illness may abrogate the antibody responseto B burgdorferi.36A suitable test system would include a sensi-

tive screening test, such as an ELISA whichuses whole B burgdorferi or a flagella enrichedpreparation as the antigen. All positive resultswould then be checked by immunoblotting.Where clinically indicated, specimen testingwould have to be repeated. In most instancesserum would be a suitable specimen, but ifneurological syndromes are being investigatedCSF should also be examined. In one study nine

1360

on 2 July 2018 by guest. Protected by copyright.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.66.11.1359 on 1 Novem

ber 1991. Dow

nloaded from

Lyme disease in paediatrics

patients with neurological syndromes suggestiveof Lyme disease had serum which was positiveby ELISA, but only one had a CSF positive byELISA and immunoblot and also evidence ofintrathecal antibody production. 3

In interpreting the laboratory results onemust be aware that a proportion of the popula-tion will have serological evidence of previousexposure to B burgdorferi which may be un-related to their current symptoms. Seropositiv-ity rates of approximately 25% are detected inhigh risk populations such as forestry workersor orienteers." 37 Clinical symptoms attribut-able to Lyme disease were detected in 5% of theforestry workers and 0-8% of the orienteers.Modern advances in molecular genetics and

the application of techniques such as the poly-merase chain reaction (PCR) may resolve someof these diagnostic problems. A PCR assayusing a highly conserved area of B burgdorferichromosomal DNA was able to detect fewerthan five copies of complementary DNA. Theassay did not cross react with B hermsi oreukaryotic DNA.38 The same authors expandedtheir primer range and were able to differentiatebetween B burgdorferi of European or Americanorigin.39 Using a PCR method, B burgdorferiDNA was detected in the urine of four of eightpatients with clinical syndromes suggestive ofLyme disease.' If there is a high prevalence ofB burgdorferi genetic material in the urine ofinfected patients, detection may simplify thelaboratory diagnosis of Lyme borreliosis. Bysuch techniques we may be able to detect theseorganisms in body tissues and further define thepathogenesis of Lyme disease. Classifyingborrelia by PCR may provide valuable epide-miological information which would haveespecial relevance to vaccine developmentstrategies.

TreatmentThe aims of treatment in Lyme borreliosis are(a) to speed the resolution of the patientspresenting illness and (b) to stop the progres-sion to major long term sequelae.

It is probably best to avoid prophylacticantibiotics for people who have been bitten byticks, as the likelihood of an individual develop-ing Lyme disease is small and there is always arisk of an unexpected allergic reaction toantibiotics.4' Insect repellents to prevent theattachment of the tick are more practical.Nursing and medical staff attending patientswith Lyme disease do not require prophylaxis asperson to person transmission is unknown.A particular problem of Lyme disease is that

the clinical outcome cannot always be predictedfrom the sensitivity of the causative bacteriumto antimicrobial agents in vitro. B burgdorferi issensitive to a variety of common antibiotics,such as benzylpenicillin, erythromycin, tetra-cycline, amoxycillin, ceftriaxone, andcefotaxime. Antibiotics such as metronidazole,rifampicin, and sulphonamides have no role inthe therapy of Lyme disease. The new quino-lones and the aminoglycosides are likewiseineffective at clinical dosageS.42-44An oral agent is preferable in the treatment

of early Lyme disease. Erythema migrans andits associated symptoms resolved significantlyearlier in patients given tetracycline or penicillinthan in those treated with erythromycin.45None of 39 patients treated with tetracyclinedeveloped major late complications but 8% ofthe 40 patients given penicillin, and 14% of the29 erythromycin treated patients did. All drugswere used in a dosage of 250 mg four times dailyfor 10 days. In a more recent study, no statisti-cally significant differences were detected in theresponses to tetracycline or penicillin therapy inerythema migrans. The tetracycline treatedpatients, however, developed fewer long termmajor complications, and fewer ofthem requiredfurther courses of antibiotic therapy.46

Oral doses of tetracycline, 250 mg four timesa day, produce peak serum concentrations fourtimes the minimum inhibitor concentration(MIC), whereas phenoxymethylpenicillin in oraldoses of 500 mg four times a day producesserum concentrations equal to the MIC.47Doxycycline in a dosage of 100 mg twice dailymay be a convenient alternative. Amoxycillincombined with probenecid has also been usedsuccessfully to treat early Lyme disease. Arecent study compared amoxycillin and pro-benecid (500 mg of each three times) to doxy-cycine (100 mg twice daily). All patients hadearly skin manifestations of Lyme disease andthe duration of therapy was 21 days in eachcase.48 In all patients the erythema migransresolved and none developed major late sequelae.Symptoms after treatment such as fatigue,headache and arthralgia, developed in 15% ofpatients and persisted for up to three months.Such symptoms have been noted by otherauthors and resolve. They are twice as likely tooccur in patients with disseminated early dis-ease than in those with single lesions.49The duration of treatment depent n the

severity of the illness: 10 days is thougnt to beadequate in mild cases with skin manifestationsbut 21 days may be more appropriate in cases ofdisseminated disease. Longer periods of treat-ment do not reduce the incidence of major orminor late sequelae.49 In children less than 9years old tetracyclines are contraindicated. Insuch cases amoxycillin and probenecid orally for10 to 21 days are the drugs of choice.Erythromycin is a less effective alternative forpenicillin sensitive children. In children olderthan 9 years, oral doxycycline is the mostappropriate therapy. Amoxycillin with pro-benecid is an alternative.The classical treatment for the neurological

signs of Lyme disease was high dose parenteralpenicillin.50 In patients who mainfest anuncomplicated Bell's palsy the outcome isusually favourable, and unless symptoms ofmeningitis or radiculitis are present oral admi-nistration is probably adequate. In a Swedishstudy of doxycycline treatment for neurologicalLyme disease, doxycycline concentrations of0-2 to 1 mg/l were obtained in the CSF during aregimen of 100 mg twice daily, all patients had afavourable outcome.5' On a treatment scheduleof 200 mg twice daily, CSF concentrations of0-6-1 9 mg/l were achieved. The MIC ofB burgdorferi was exceeded in the CSF after the

1361

on 2 July 2018 by guest. Protected by copyright.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.66.11.1359 on 1 Novem

ber 1991. Dow

nloaded from

1362 Cryan, Wright

first day of this dosage, but three to five days oftreatment was needed to achieve comparablelevels on the lower dosage regimen.52Treatment for patients with meningitis or

radiculitis should be more aggressive. Parent-eral penicillin in doses of 20 megaunits daily for10 days has been used as standard treatment forneuroborreliosis. Results have been disappoint-ing and the isolation of B burgdorferi from theCSF of a patient being treated with penicillinunderlines the inability of penicillin to killspirochaetes which may survive under sub-optimal conditions.53 From the results of invitro studies, third generation cephalosporinsappear effective.42 4 Ceftriaxone has a half lifeof 8-5 hours.54 In the absence of meningitis itcrosses the blood-CSF barrier to give CSFconcentrations 1L5% of those in serum. Two tofour hours after an intravenous infusion of 2 g ofceftriaxone, CSF concentrations three to fourtimes the MIC for B burgdorferi may beobtained.55 It has been used successfully andproved superior to parenteral penicillin in lateneuroborreliosis.56Though experience with ceftriaxone in Lyme

meningitis or radiculitis is limited it appears tobe a useful agent.57 58 An association has beendemonstrated between the use of ceftriaxoneand the development of biliary concretions.This antibiotic is as yet unlicensed in theUK.59 As an alternative cefotaxime may be usedand is equally effective.58 The duration of treat-ment depends on the clinical response andusually varies between 21 and 30 days. In apatient treated for 10 days with parenteralceftriaxone B burgdorferi was isolated from theCSF 7 5 months after treatment.56 Recoveryfrom meningitis is relatively rapid but radicularsigns may persist for up to 24 months.49Lyme encephalopathy is probably best

treated by parenteral ceftriaxone, or by benzyl-penicillin as a less effective alternative.56 60There is a case report of the successful use ofcefotaxime in Lyme encephalitis.6'

In patients with Lyme chronic synovitis, treat-ment failure rates of up to 50% have beenreported, especially if erosive changes arepresent.62 Ceftriaxone is superior to penicillin,with therapeutic success rates of over 90%.56 60Oral doxycycline or, alternatively amoxycillinwith probenecid, have been used to treat thesepatients, with success rates of 72% and 61%respectively.63The use of corticosteroids in patients with

chronic synovitis increases the risk of antibioticfailure and they are best avoided.56 The develop-ment of chronic synovitis with a poor responseto therapy has been associated with HLA DR4and HLA DR2 alleles.64

Jarisch-Herxheimer reactions are encounteredin 14% of patients treated for early Lymedisease. They occur within 2-4 hours of startingtreatment and are more common in severedisease.45 Meptazinol unlike adrenocortico-steroids, has been shown to ameliorate suchreactions in relapsing fever.65 Meptazinol maybe useful in cases of Lyme disease involvingvital organs.

In conclusion, Lyme disease is a seriousdisorder with potential immediate and long

term morbidity and mortality. It occurs in manyparts of the world and affects all ages. Thediagnosis of Lyme borreliosis should be con-sidered in any unusual neuropathy or arthro-pathy. Laboratory tests may confirm the diag-nosis but must be interpreted with care. Longterm complications may occur if it is notrecognised and treated effectively. Oral tetra-cycline, if the patient is older than 9 years, isrecommended for early Lyme disease. Youngerchildren should receive amoxycillin and pro-benecid. In the late stages, ceftriaxone orcefotaxime is more effective than penicillin.

In future, vaccines may provide protectionagainst Lyme borreliosis. Recent evidence inmurine models of B burgdorferi: infectionsuggests that outer membrane proteins may beimportant in this regard.66

1 Steere AC, Malawista SE, Snydman DR, et al. Lyme arthritis:an epidemic of oligoarticular arthritis in children and adultsin three Connecticut communities. Arthritis Rheum 1977;20:7-17.

2 Afzelius A. Report to the Verhandlungen der DermatologishenGessellschaft zu Stockholm. Arch Dermatol Syph 1910;101:405-6.

3 Bannwarth A. Chronische lymphocytare meningitis,enzundliche polyneuritis und 'rheumatismus' ein beitragzum problem 'allergie und nervensystem'. Archiv furPsychiatrieNervenkrankheiten 1941;113:284-376.

4 Burgdorfer W, Barbour AG, Hayes SF, Benach JL,Grunwald E, Davis JP. Lyme disease-a tick borne spiro-chetosis? Science 1982;216:1317-9.

5 Benach JL, Bosler EM, Hanrahan JP, et al. Spirochetesisolated from the blood of two patients with Lyme disease.N EnglJr Med 1983;308:740-2.

6 Steere AC, Grodzicki RL, Kornblatt AN, et al. The spiro-chetal aetiology of Lyme disease. N Engl J Med 1983;308:733-40.

7 Schmid GP. The global distribution of Lyme disease. RevInfect Dis 1985;7:41-50.

8 Parke A. From new to old England: the progress of Lymedisease. BM3r 1987;294:525-6.

9 Kawabata M, Bara S, Iguchi K, Yamaguti N, Russell H.Lyme disease in Japan and its possible incriminating tickvector Ixodes persulcatus. J Infect Dis 1987;5:854.

10 Lastavica CC, Wilson ML, Berardi VP, Spielman A,Deblinger RD. Rapid emergence of a focal epidemic ofLyme disease in coastal Massachusetts. N EnglJ Med 1989;320:133-7.

11 Guy EC, Martyn CN, Bateman DE, Heckels JE, LawtonNF. Lyme disease: prevalence and clinical importance ofborrelia burgdorferi species specific IgG in forestryworkers. Lancet 1989;i:484-6.

12 Schmutzhard E, Stane K, Pletchette M, et al. Infectionsfollowing tick bites. Tick-borne encephalitis and Lymeborreliosis-a prospective study from Tyrol. Infection 1988;16:269-72.

13 Cryan B, Cutler SJ, Wright DJM. Lyme disease in Ireland. IrMedJ (in press).

14 Baird AG, Gilies JCM, Bone FJ, Dale BAS, Miscambell NT.Prevalence of antibody indicating Lyme disease in farmersin Wigtownshire. BMJr 1989;299:836-7.

15 Centers for Disease Control. Lyme disease-United States1987 and 1988. MMWR 1989;38:668-72.

16 Piesman J, Mather TN, Dammin GJ, Telford SR, LastavicaCC, Spielman A. Seasonal variation of transmission risk ofLyme disease and human babesiosis. AmJ Epidemiol 1987;126:1187-9.

17 O'Neill PM, Wright DJM. Lyme disease. Br J Hosp Med1988;40:284-9.

18 Malane MS, Grant-Kels JM, Feder HM, Luger SW. Diagno-sis of Lyme disease based on dermatologic manifestations.Ann IntMed 1991;114:490-8.

19 Millner M, Schimek MG, Spork D, Schnizer M, Stanek G.Lyme borreliosis in children. Eur J Pediatr 1989;148:527-30.

20 Naglo A, Wide K. Borrelia infection in children. ActaPaediatr Scand 1989;78:918-22.

21 Culp RW, Eichenfield AH, Davidson RS, Drummond DS,Christofersen MR, Goldsmith DP. Lyme arthritis in child-ren. J Bone3Joint Surg [Am] 1987;69-A:96-9.

22 Nadal D, Gundelfinger R, Flueler U, Boltshauser E. Acro-dermatitis chronica atrophicans. Arch Dis Child 1988;63:72-4.

23 Steere AC, Bartenhagen NH, Craft JE, et al. The early clini-cal manifestations of Lyme disease. Ann Int Med 1983;99:76-82.

24 Schmidt R, Kabatzki J, Hartung 5, Ackermann J. Erythemachronicum migrans disease in the Federal Republic ofGermany. Zentralbi Bakteriol Mikrobiol Hgy [A] 1986;263:

25 Banach JL, Coleman JL. Clinical and geographic features ofLyme disease in New York. Zentralbi Bakteriol MikrobiolHgy [A] 1986;263:477-82.

on 2 July 2018 by guest. Protected by copyright.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.66.11.1359 on 1 Novem

ber 1991. Dow

nloaded from

Lyme disease in paediatrics 1363

26 Bigaignon G, Gubau P, Desmyter J, Vandepitte J. Lymeborreliosis in Belgium. Lancet 1987;i:557.

27 Schlesinger PA, Duray PH, Burke BA, Steere AC, StillmanMT. Maternal fetal transmission of Lyme disease spiro-chaete Borrelia burgdorferi. Ann IntMed 1985;103:67-8.

28 Nadal D, Hunziker UA, Bucher UA, Hitzig WH, Duc G.Infants born to mothers with antibiotics against Borreliaburgdorferi at delivery. EurJ7 Pediatr 1989;148:426-7.

29 Preac-Mursic V, Wilske B, Schierz G. European Borreliaburgdorfer isolated from humans and ticks cultural condi-tions and antibiotic susceptibility. Zentralbi BakteriolMikrobiol Hgy [Al 1986;263:112-8.

30 Magnarelli LA. Laboratory diagnosis of Lyme disease.Rheum Dis Clin NorthAm 1989;15:735-45.

31 Dattwyler RJ, Luft BJ. Immunodiagnosis of Lyme borrelio-sis. Rheum Dis Clin North Am 1989;4:727-33.

32 Karlsson M, Stiernstedt G, Granstrom M, Asbrink E,Wretlind B. Comparison of flagellum and sonicate antigensfor serological diagnosis of Lyme borreliosis. Eur J ClinMicrobiol Infect Dis 1990;3:169-77.

33 Hansen K, Asbrink E. Serodiagnosis of erythema migransand acrodermatitis chronica atrophicans by Borrelia burg-dorferi flagellum enzyme-linked immunosorbent assay.J ClinMicrobiol 1989;3:545-51.

34 Cutler SJ, Wright DJM. Comparison of immuofluorescentand enzyme linked immunosorbent assays for diagnosingLyme disease. J Clin Pathol 1989;42:869-71.

35 Nadal D, Taverna Ch. Hitzig WH. Immunoblot analysis ofantibody binding to polypeptides of Borrelia burgdorferi inchildren with different clinical manifestations of Lymedisease. PediatrRes 1989;4:377-82.

36 Dattwyler RJ, Volkman DJ, Golightly MG, Falldorf PA,Thomas J. Seronegative Lyme borreliosis after early anti-biotic treatment. Arthritis Rheum 1987;30:S36.

37 Fahrer H, van der Linden SM, Sauvain M, Gem L, ZhiouaE, Aeschlimann A. The prevalence and incidence of clinicaland asymptomatic Lyme borreliosis in a population at risk.J Infect Dis 1990;163:305-10.

38 Rosa PA, Schwan TG. A specific and sensitive assay for theLyme disease spirochete Borrelia burgdorferi using thepolymerase chain reaction. J Infect Dis 1989;160:1018-29.

39 Rosa PA, Schwan TG. Polymerase chain reaction analysesidentify two distinct classes of Borrelia burgdorferi J ClinMicrobiol 1991;29:524-2.

40 Goodman JL, Jurkovich P, Kramber JM, Johnson RC.Molecular detection of persistent Borrelia burgdorferi inthe urine of patients with active Lyme disease. InfectImmunol 1991;59:269-78.

41 Ewen PW, Ackroyd JF. Allergic reactions to drugs. In:Wright DJM, ed. Immunology of sexually transmitteddiseases. Lancaster: MPT Press, 1978:244-6.

42 Johnson RC, Kodner C, Russell M. In vitro and in vivosusceptibility of the Lyme disease spirochaete, Borreliaburgdorferi to four antimicrobial agents. Antimicrob AgentsChemother 1987;31:164-7.

43 Preac-Murisc V, Wilske B, Schierz G, Holmburger E. In vivoand in vitro susceptibilities of Borrelia burgdorferi. EurJ ClinMicrobiol 1987;6:424-6.

44 Preac-Murisc V, Wilske B, Schierz G, Sub E, Grob B.Comparative antimicrobial activity of new macrolidesagainst Borrelia burgdorferi. EurJ Clin Microbiol Infect Dis1989;8:65 1-2.

45 Steere AC, Green J, Hutchinson GJ, et al. Treatment of Lymedisease. Zentralbl Bakteriol Mikrobiol Hgy [A] 1987;263:352-6.

46 Weber K, Preac-Murisc V, Neubert U, et al. Antibiotictherapy of early European Lyme borreliosis and acroderma-

titis chronic atrophicans. Ann N Y Acad Sci 1988;539:324-46.

47 Luft BJ, Volkman DJ, Halperin JJ, Dattwyler RJ. Newchemotherapeutic approaches in the treatment of Lymeborreliosis. Ann NYAcad Sci 1988;539:352-61.

48 Dattwyler RJ, Volkman D, Conaty S, Gorevic P, Luft B.Ampxycillin plus probenecid compared to doxycycline forthe treatment of erythema migrans. Arthritis Rheum 1990;33(Suppl):S85.

49 Rahn DW, Malawista SE. Lyme disease: recommendationsfor diagnosis and treatment. Ann IntMed 1991;114:472-81.

50 Baumhackl U, Kristoferistsch W, Sluga E, Stanek G.Neurological manifestations of Borrelia burgdorferi infec-tions: the enlarged clinical spectrum. Zentralbl BakteriolMikrobiolHgy [A] 1987;263:334-6.

51 Dotevall L, Alestig K, Hanner P, Norkrans G, Hagberg L.The use of doxocycline in nervous system Borrelia burgdor-feri infection. ScandJ7 Infect Dis 1988;53:74-9.

52 Dotevall L, Hagberg L. Penetration of doxycycline intocerebrospinal fluid in patients treated for suspected Lymeneuroborreliosis. Antimicrob Agents Chemother 1989;33:1078-80.

53 Baranton G, Perolat P, Dufresne Y, Postic D, Quentin R,Fouquet B, Isolation of Borrelia burgdorferi in the spinalfluid of a patient treated with penicillin. Presse Med 1989;18:637.

54 Wise R. Antimicrobial agents: a widening choice. Lancet 1987;ii: 1251-4.

55 Chandrasekar P, Rolston K, Smith B, LeFrock J. Diffusionof ceftriaxone into the cerebrospinal fluid of adults.J Antimicrob Chemother 1984;14:427-30.

56 Dattwyler RJ, Halperin JJ, Volkman DJ, Luft BJ. Treatmentof late Lyme borreliosis-randomised comparison of ceftri-axone and penicillin. Lancet 1988;i: 1 191-4.

57 Pfister HW, Einhaupl KM. Neurological complications ofLyme-borreliosis: outpatient treatment with once dailyceftriaxone. Abstract IV. International Conference on LymeBorreliosis. Stockholm, 1990 Book A: 165.

58 Pfister HW, Prec-Mursic V, Wilske B, Schielke E, Sorgel F,Einhaupl KM. Randomised comparison of ceftriaxone andcefotaxime in Lyme neuroborreliosis. Jf Infect Dis 1991;163:311-8.

59 Schaad UB, Wedgewood-Krucko J, Tschaeppeler H. Rever-sible ceftriaxone-associated biliary pseudolithiasis inchildren. Lancet 1988;ii: 1411-3.

60 Dattwyler RJ, Volkman DJ, Goliath MG, Falldorf PA,Thomas J. Seronegative Lyme boreliosis after early anti-biotic treatment. Arthritis Rheum 1987;30(suppl 4):S36.

61 Pal GS, Baker JT, Wright DJ. Penicillin resistant borreliaencephalitis responding to cefotaxime. Lancet 1988;i:50-5 1.

62 Steere AC, Green J, Schoen RT, et al. Successful parenteralpenicillin therapy of established Lyme arthritis. N EnglJMed 1985;312:869-74.

63 Liu NY, Dinerman H, Levin RF, et al. Randomised trial ofdoxycycine versus amoxicillin probenecid for the treatmentof Lyme arthritis: treatment of non responders with ivpenicillin or ceftriaxone (abstract). Arthritis Rheum 1989;32(suppl):S32.

64 Steere AC, Dwyer E, Winchester R. Association of chroniclyme arthritis with HLA-DR4 and HLA-DR2 alleles.N EnglJ Med 1990;323:219-23.

65 Teklu B, Habte-Michael A, Warrell DA, White NJ, WrightDJM. Meptazinol diminishes the Jarisch-Herxheimerreaction of relapsing fever. Lancet 1989;i:835-9.

66 Simon MM, Schaible UE, Wallich R, Kramer MD. A mousemodel for Borrelia burgdorferi infection: approach to avaccine against Lyme disease. Immunol Today 1991;12:11-6.

on 2 July 2018 by guest. Protected by copyright.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.66.11.1359 on 1 Novem

ber 1991. Dow

nloaded from