Lyme Disease - Europe PMC

6 downloads 0 Views 1MB Size Report
keratitis, or panophthalmitis have been re- ported at later stages. About 20% of patients also have evidence ofmild hepati- tis in the early disseminated stage of.
Lyme Disease Revuw from a Canadian perspectie

DAVID R. BURDGE, MD, FRCPC DAVID O'HANLON, MB, CHB, FRCPC

Lyme borreliosis is an infectious disease caused by the tick-transmitted spirochete Borrelia burgdorferi. To date, the only known endemic focus of Lyme disease in Canada is Long Point, Ont. A national case definition for surveillance purposes, consensus statement regarding laboratory diagnosis, and treatment guidelines have recently been developed in an attempt to standardize the approach to surveillance, diagnosis, and management of Lyme borreliosis in Canada. La borreliose de Lyme est une maladie infectieuse causee par le spirochete Borrelia burgdorferi, lequel est transmis par une tique. Au Canada, le seul foyer endemique connu jusqu'a maintenant est la region de Long Point, Ontario. On a recemment developpe une definition nationale des cas pour ameliorer la surveillance, un enonce de consensus pour le diagnostic de laboratoire et un guide therapeutique visant a standardiser l'approche a la surveillance, au diagnostic et au traitement de la borreliose de Lyme au Canada. (an Fain Ptysk 1992;38:1426-1432.

YME BORRELIOSIS IS AN IN-

E_

fectious disease caused by the tick-transmitted spirochete Borrelia burgdo?feri. The illness usually begins in summer and is characterized by a virtually pathognomonic skin lesion, which was initially called erythema chronicum migrans and is now more commonly termed erythema migrans (EM). Soon after infection the organism can disseminate to multiple organ systems and can cause acute or chronic clinical manifestations characteristically involving the cardiovascular, neurologic, musculoskeletal, and dermatologic systems."2 There are many similarities between Lyme borreliosis and syphilis: both are spirochete infections with the potential for multisystem disease and, like syphilis, Lyme disease responds to antibiotics at all stages. Lyme disease was initially recognized in 1975, when close geographic clustering of children who were thought to havejuvenile rheumatoid arthritis was observed in Lyme, Conn.3 The rural setting, the identification of EM as a feature of the illness, and the observation that patients with EM treated Dr Burdge is an Assistant Clinical Pmfessor in dte Divis of Infectious Diseases, Department of Medicine, University Hospita4 University of Bnitish Columbia,

Vancouver Dr O'Hanlon is a Clinical Instnstor in the Diision of Rezenatolog, Unierst of Britsh Cohnia Vancouver

1426 Canadian Family Physician VOL 38: June 1992

with penicillin showed more improvement than did untreated patients all suggested that this was an infectious disease, likely arthropod transmitted. The conclusive evidence came in 1982 when Burgdorfer and colleagues4 isolated what is now known as B burgdo?feri from Ixodes dammini ticks and when this organism subsequently was isolated from patients with Lyme disease. Borrelia burgdoferi has also been isolated from patients in Europe with EM, acrodermatitis chronica atrophicans, and Bannwarth's syndrome (radicular pain followed by chronic lymphocytic meningitis and sometimes cranial nerve palsy or

peripheral neuritis). Vectors of transmission and animal hosts Borrelia burgdorferi is transmitted by ticks of the Ixodes ricinus complex. In the northeastern and midwestern United States, Idammini is the vector, while in the western United States, it is Ixodespacficus. In Europe, I ricinues is the primary vector. While the organism has been demonstrated in other species of ticks and in mosquitoes and deer ffies, only ixodid ticks seem important in transmission of infection to humans. In Connecticut, the spirochete can be found in 1 0% to 35% of all recovered ticks, and in Shelter Island, NY, infection rate exceeds 50%.1"2 Ixodid ticks feed once during each of the three stages of their usual 2-year life cycle. Larval ticks take one blood meal in late sum-

mer; nymphs feed during the following spring and early summer; and adult ticks take their single blood meal that autumn. In the United States, I dwnmizi larvae and nymphs feed mainly on white-footed mice, and adults feed on large mammals, especially deer. The spirochete causes no illness whatsoever in the mouse, which has spirochetemia throughout the summer. The larval tick feeds on the infected mouse and thereby becomes infected. The spirochete remains in the digestive tract of the tick until the following summer, when as a nymph it attaches to another host. At that point, the organism migrates to the tick's salivary glands and is injected as it feeds. The nymphal stage is responsible for most transmissions. It should be noted, however, that experimental transmission studies in rodents indicate that 24 hours or more of tick attachment is needed before significant transmission of the spirochete occurs.5 The persistence of the spirochete in this cycle is largely dependent on the small mammal host, while deer appear to be important for survival of the adult tick and maintenance of the tick cycle. In the western United States, Ipacificus feeds predominantly on lizards rather than mice, and lizards are not susceptible to infection with this spirochete. This characteristic probably accounts in part for the low prevalence of Lyme disease on the west coast. The organism is found in only 1% to 3% of ticks surveyed in California and Oregon.6

Lyme disease is widespread outside of North America and is highly prevalent in northern and central Europe. It is also recognized in China andJapan, although the exact incidence of the disease in Asia is un-

known."2 In Canada, only one area is known to be endemic for Lyme borreliosis, the Long Point National Wildlife Area at Long Point, Ont. Sporadic cases have been reported from across the country, however, and a total of 140 cases was reported from 1984 through 1990; 100 of these reports were from Ontario, 40 from the rest of Canada. Forty-two percent of the Ontario cases and 97.5% of those reported from other provinces had a history oftravel to a recognized endemic area outside of the reporting province.8 Only recently has consensus been achieved on a national case definition for surveillance purposes; therefore, we still lack accurate information about the frequency of Lyme disease in Canada.8 Many of the previously reported cases, however, do not meet the present case definitions, and it appears that the disease is either not endemic or of extremely low prevalence in the rest of Canada. In addition to surveillance for human cases of Lyme disease, tick surveillance is also important. A consensus statement on the epizootiology of Lyme disease in Canada has recently been published.8 On the basis of current information, I dammini ticks are endemic only in Long Point. In all other areas of the country, I dammini are classified as being adventitious or absent.8 Ixodespacificus is endemic in British Columbia, and further investigation is under way to determine whether B burdorferi is present in that province.

Epidemiology and geographic distribution Lyme borreliosis occurs only in areas where Ixodes ticks are endemic. In North America, there are three geographic foci: the northeastern United States, the midwest (Wisconsin and Minnesota), and California and Oregon in the west.7 According to the Cen- Clinical characteristics ters for Disease Control, from 1980 to Lyme borreliosis usually starts with EM and 1988, 13 795 cases ofLyme disease were re- associated symptoms (acute localized infecported, the vast majority of which came tion), which is often followed weeks to from these three geographic regions. Peo- months later by acute cardiac, neurologic, ple of all ages and both sexes are affected. or musculoskeletal signs and symptoms The onset of illness is generally between (acute disseminated) and months to years May and November, with highest rate of later by chronic skin, nervous system, or onset in June andJuly. From 1984 to 1987 joint disease (chronic disseminated). The the Centers for Disease Control reported illness is characterized by remission and exabout 1500 new cases each summer in the acerbations with a trend to improvement United States.2 in time, even without therapy." 2

Canadian Family Physician VOL 38: June 1992 1427

Skin manifestations. Erythema migrans is the virtually pathognomonic skin lesion that is associated with Lyme disease. Erythema migrans occurs around the site of the usually painless tick bite. About two thirds of patients with Lyme borreliosis develop EM, usually 3 to 30 days after the bite.' ' The lesion normally begins as a red macule or papule, and an area of expanding erythema develops, usually over the course of several days. Usually it is flat and is not painful or itchy, but it does feel warm to touch. The common partial central clearing results in the classic target rash. Lesions can be located anywhere, but the thigh, groin, and axilla are particularly common sites. Erythema migrans lesions are characteristically at least 5 cm in diameter, and true EM lesions less than 2.5 cm in diameter often increase in size if observed over 24 to 48 hours. The lesions usually fade within 3 to 4 weeks, even without therapy, although occasionally they last longer. Differential diagnosis of EM includes superficial fungal infection, fixed drug eruptions, and erythema annulare. Skin biopsy has been reported to demonstrate B burgdoiferi in up to 50% of cases.9 About half of patients in the United States who are untreated with primary EM go on to develop multiple secondary annular lesions.'° These lesions are smaller and removed from the primary lesion and are not at the site of tick bites; they represent cutaneous dissemination of infection. The secondary annular lesions also usually fade within 3 to 4 weeks, even in the absence of therapy. The EM lesion is commonly associated with systemic symptoms of malaise, fatigue, headache, fever, myalgias, and regional lymphadenopathy. Some patients also have signs and symptoms suggestive ofmeningeal irritation and migratory musculoskeletal pains. These early signs and symptoms are typically intermittent and changing.'0 The chronic skin disease associated with B burgdoofe infection is acrodermatitis chronica atrophicans. This disease is seen much more commonly in Europe than in North America. The disease starts with a reddish-blue lesion that becomes sclerotic and atrophic. It should be noted that B burgdorferi can be isolated from the acrodermatitis chronica atrophicans skin lesion as long as 10 years after onset.",2 1428 Canadian Family Physician VOL 38: June 1992

Musculoskektal manifestations. Overall, about 80% of patients in North America with Lyme borreliosis will develop musculoskeletal signs and symptoms. " Onset is characteristically within a few weeks to 2 years after onset of infection. In most cases patients first complain of migratory joint or periarticular pain, without joint swelling. Frank arthritis begins about 6 months after onset of symptoms and occurs in 50% to 60% of patients. It consists of intermittent episodes, usually affecting one or two large joints asymmetrically (the knee most commonly). Attacks of arthritis generally last a few weeks to months, and then go into complete remission before recurring. Episodes of arthralgia or periarticular pain sometimes continue between episodes of frank arthritis. With time, even in the absence of therapy, frequency of relapses decreases; only about O0% of patients with arthritis go on to develop chronic arthritis (usually in one or two large joints), and less than 5% of patients develop erosive arthritis." Joint fluid aspirates reveal 500 to 110 000 white blood cells/mm3 consisting mainly of polymorphonuclear leukocytes. Immune complexes are present, and B burgdorferi has been isolated from synovial fluid.'2

Neurologic manifestations. During the early stages of B burgdorfeii infection, symptoms suggestive of meningeal irritation occur frequently. Steere et al'0 reported headache in 200 of 314 (64%) patients with EM, and stiff neck in 151 of 314 (48%). During this stage of the illness, there is no objective neurologic deficit, and cerebrospinal fluid is normal. In the United States, however, 15% to 20% of patients go on to develop definite neurologic abnormalities, usually several weeks to months after initial infection. The most common presentation is fluctuating symptoms of meningitis with superimposed cranial (usually facial palsy) or peripheral radiculoneuropathy. 3,14 Neck stiffness is usually subtle, but patients with objective neurologic abnormality have CSF pleocytosis, on the order of about 100 cells/mm3, which are predominantly lymphocytes. Cerebrospinal fluid protein is also often elevated, but the glucose level is usually normal. Patients with facial palsy alone can

have normal CSF, as the problem can be peripheral. Neurologic abnormalities typically last for weeks to months before resolving, but can recur and occasionally become chronic. The peripheral neuritis is often an asymmetric motor, sensory, or mixed radiculoneuropathy of the limbs or trunk. Results of electrophysiologic studies are usually abnormal. In Europe, the most common presentation of neuroborreliosis is Bannwarth's syndrome. This is characterized by lymphocytic pleocytosis without headache, neuritic pain, and sometimes, cranial nerve palsy. Cerebrospinal fluid typically shows lymphocytes, increased synthesis of B burgdofferi antibody in the CSF, and sometimes oligoclonal bands. More recently chronic syndromes of the central nervous system beginning months or years after disease onset have been associated with B burgdofferi. Progressive encephalomyelitis, subacute encephalitis, dementia, and syndromes suggestive of demyelinating disease have all been linked to late Lyme neuroborreliosis."2"5 The full extent of these late neurologic syndromes remains to be clarified.

Cardiac manifestations. About 8% of patients develop cardiac involvement, usually within several weeks of initial infection. The most common cardiac abnormality is a fluctuating degree of atrioventricular heart block. Heart block is usually of brief duration, and complete heart block rarely persists for more than a week. Insertion of a permanent pacemaker is, therefore, unnecessary, although the cardiac problem can recur. A few patients have developed myopericarditis, mild left ventricular dysfunction, or rarely, cardiomegaly. 16 Other systems. Although the skin, nerves, joints, and heart are most commonly affected in Lyme borreliosis, almost any organ system can be involved. The eye can be involved with conjunctivitis, but iritis, keratitis, or panophthalmitis have been reported at later stages. About 20% of patients also have evidence of mild hepatitis in the early disseminated stage of

infection.",2

Diagnosis Definitive diagnosis of Lyme borreliosis lies in culturing the organism from clinical specimens. This has been accomplished from skin biopsy specimens (both EM and acrodermatitis), blood, CSF, and synovial fluid."2 At present, however, routine diagnostic microbiology laboratories and, indeed, most reference laboratories in both Canada and the United States are not in a position to offer Borrelia culture as a routine diagnostic test. Furthermore, the yield on culture is generally low from clinical specimens. It is also possible to detect the spirochete in histologic sections, although staining artifact, low numbers of organisms in clinical specimens, and lack of specificity of routine stains (ie, specific monoclonal antibodies are necessary to differentiate B burgdoferi from other spirochetes) all make histologic sectioning and staining difficult. Despite the inherent difficulties ofboth spirochete culture and staining in clinical specimens, Canadian physicians are strongly encouraged to submit specimens for culture and staining (tissue, blood, CSF, synovial fluid) from any patient with possible Lyme borreliosis to provincial or national reference laboratories. Given the lack of definitively diagnosed cases in Canada, every attempt should be made to make a definitive diagnosis. Indeed, the recent Canadian consensus statement on laboratory investigation states that "isolation of B burgdorferi (from biopsies at the advancing edge of erythema migrans lesions) or identification by immunospecific staining is encouraged when possible."8

New methods. New methods of detection of B burgdo feri antigens or genetic material are being actively investigated. These research techniques include detection ofantigen in urine, DNA probes, and polymerase chain reaction and could add to our ability to make a definitive diagnosis of Lyme disease. At present, however, none of these techniques are available for routine diagnostic testing.

Serologic testing. Serodiagnosis, therefore, remains the most practical way of confirming a clinical diagnosis. Unfortunately, serologic tests suffer from lack of Canadian Family Physician VOL 38: June 1992 1429

Table1. CANADIAN TREATMENT GUIDELINES FOR LYME DISEASE: These guidelines could change as

fizrwr cont70lkd antibiotic studies are peifmed.

Eryth. Milrums; 10'1s p wm"ou CSF

Adults

rbmkhuilsu Amoxicillin

e

first-degre

hurt

100 mg orally BID

Doxycydineb OR

MOc

thryy of

Amxid pls hrohbned& 500 mg orally TID 500 mg orally TID

Probenecid

Cd re-

(NS dises wih CSF rmits; cd or trddwrwe hurt hlabe Artiis

coib.u. Amoxidcillin Probenecid

2-3

Of

ph.

2-3 2-3

Probmdi:

40 mg/kg/d 50 mg/kg/d

2-3 2-3

Aduls

Ceftriaxone

2 g IV daily

2

CWlidr

Ceftriaxone

50 mg/kg IV daily

2

aDuration of therapy is- guided by clinical response.

bDoxyycline should not be used dunpr Data fiom Health and Welfare Canada.8

or in children 9years oryoungei

5

standardization, from interlaboratory and intralaboratory variability, and from low specificity. 17"18

The indirect florescent antibody (IFA) test was the first serologic test used for confirming Lyme disease and is still widely used. To perform this test, spirochetes are fixed on to a glass slide and then incubated with human serum at one or more serial dilutions. Human anti-B burgdorferi antibodies present in the serum bind to the spirochete, and a second antibody raised against human immunoglobulin and labelled with fluorescein is then incubated on the slide. While the IFA test is relatively easy to do, it has problems with specificity, reproducibility, lack of quantification, and subjective interpretation. A titer of 1:256 or more can occur in patients exposed to other species of spirochete, or even more distantly related bacteria that share similar surface antigens. Therefore, the patient who has never been exposed to B burgdorferi, but who has had syphilis, leptospirosis, tick-borne relapsing fever, or periodontal disease or who has an autoimmune disease can have 1430 Canadian Family Physician V01

38: June

1992

antibodies that react positively in the IFA test. Other serologic methods include the enzyme-linked immunosorbent assay (ELISA) or enzyme immunoassay (EIA). This type of assay has advantages over the IFA in that it allows for tests of large numbers of specimens and uses a spectrophotometric determination that is quantifiable. However, it has similar problems in specificity to the IFA. The problem of false-positive results when using both IFA and ELISA, especially for patients with atypical clinical presentation from areas not known to be endemic for Lyme borreliosis, has required the development of a potentially more specific serologic test. The Western blot analysis is thought to have great potential for increasing the specificity of Borrelia serology.'9 Done properly, negative results from Western blot (not in a patient just recently infected) appear to rule out the presence ofanti-B burgdorferi antibodies with a high degree of confidence. Determination of what constitutes a positive test remains a matter of debate, however, and

authorities have still not reached a consensus on guidelines. After the first few days of infection, virtually all patients mount an IgM and IgG antibody response to B burgdorferr antigens. False-negative serologic tests can occur early in the course of infection, when detectable levels of antibody have yet to develop. Antibiotic therapy early in the course of infection can also delay or eliminate a diagnostic antibody response. However, virtually all patients with clinical disease related to B burgdorfrri will develop a detectable level of antibody in their serum. Indeed, false-negative results are not generally a problem, whereas false-positive results are a large concern, especially in areas such as Canada, where the prevalence of the disease appears to be low."7

Current consensus. Current consensus, both in Canada8 and the United States,20 is that the optimal approach to serodiagnosis involves use of a combination of an ELISA test with a confirmatory Western blot. Demonstration of seroconversion in sequential specimens can also be of diagnostic value.8 It is clear, however, that at present Lyme borreliosis remains a clinical diagnosis, and serologic testing should be used for confirmation of the clinical impression. 7 Although fatigue, malaise, and musculoskeletal pain can be features of Lyme disease, patients whose fatigue or musculoskeletal pain seem to indicate a diagnosis of Lyme disease will also have other objective clinical signs of the disease. Patients who have fibromyalgia or chronic fatigue syndrome but who have no objective skin, musculoskeletal, neurologic, or cardiac signs do not have Lyme borreliosis. ,2,8,20 Treatment guidelines Initial studies performed by Steere et a12' showed that patients with EM who were treated with penicillin had a better outcome than untreated patients. Subsequently, it was shown that patients treated with tetracycline had more rapid resolution of EM and associated symptoms than those treated with erythromycin. More important, however, none of 35 patients given tetracycline developed major late sequelae while three of 40 (8%) patients treated with

penicillin and four of 29 (14%) patients treated with erythromycin did.22 More recently, both doxycycline and combination therapy with amoxicillin and probenecid were shown to be highly effective in patients with EM, and treatment was associated with complete prevention of major sequelae.23 Antibiotic sensitivity testing done in vitro shows that B burgdorferi is highly sensitive to tetracycline, ampicillin, ceftriaxone, and imipenem. Penicillin is only moderately active in vitro, and the aminoglycosides (ciprofloxacin and rifampin) have no significant activity. Erythromycin is very active in vitro, but is not as effective in vivo." 2'22 Current Canadian treatment guidelines were recently developed at a national consensus conference on Lyme borreliosis, held in Guelph, Ont, in January 1991, sponsored by Health and Welfare Canada and the Canadian Infectious Disease Society.8 Guidelines for treatment are outlined in Table 1. Acute Lyme borreliosis is a very antibiotic-responsive infection, and treatment is 95% to I100% effective for EM and 90% to 95% effective for other acute manifestations. More chronic disease, including CNS disease and arthritis, is less responsive to antibiotics (80% to 85%). While some patients require retreatment, no firm data suggest that treatment of more than 2 to 4 weeks' duration results in lower failure rates or increased efficacy. Some authors24 suggest, however, that longer courses of therapy can be beneficial. A proportion of patients treated for Lyme disease continue to have minor symptoms after therapy, including headache, musculoskeletal pain, and fatigue. These symptoms usually resolve spontaneously within 6 to 12 months.23 Herxheimer-like reactions occur in 15% to 30% of treated patients, but are usually mild and can be managed with non-steroidal anti-inflammatory medications. Antibody titers to B burgdorferi characteristically decline within a year of successful treatment, but patients remain seropositive for an indefinite period.

Conclusion Lyme disease is an infectious disease caused by the tick-transmitted spirochete Canadian Family Physician V'OL 38: june 1992 1431

B burgdorferi. It is characterized by a pathognomonic skin lesion (EM), which can be followed by acute or chronic signs involving the cardiovascular, neurologic, musculoskeletal, or dermatologic systems. The disease is still uncommon in Canada. Diagnosis of Lyme disease remains clinical. Serologic testing is useful although problematic because of low specificity. Newer, more specific serologic testing is being developed, and newer methods of antigen detection and DNA probing are also being investigated, which could make laboratory diagnosis more practical. The illness is characterized by a relapsing and remitting course, with objective signs. The acute stages of illness are highly responsive to antibiotics, and chronic disease is slightly less responsive. Recent Canadian consensus statements regarding epizootiology, epidemiology, laboratory investigation, and clinical practice have been published in an attempt to standardize the approach to surveillance, diagnosis, and management of Lyme borreliosis in Canada. I * -9 O Is O -9

0 0 * * 0 .9 lb 0

0 * 0 0 * 0 0

9

-

- 0 0 - - - 0 * v

8. Health and Welfare Canada. Consensus Conference on Lyme disease. Can Dis Wkly Rep 1991; 17(13):63-70.

9. Berger RW. Dermatologic manifestations of Lyme disease. Rev Infect Dis 1989;2(Suppl 6):S 1475-81. 10. Steere AC, Bartenhagen NH, CraftJE, Hutchison GJ, Newman JH, Rahn DW, et al. The early clinical manifestations of Lyme disease. Ann Intern Med 1983;99:76-82.

11. Steere AC, Schoen RT, Taylor E. The clinical evolution of Lyme arthritis. Ann Intern Med 1987;107:725-3 1. 12. Snydman DR, Schenhein DP, Barardi VP, Lastavica CC, Pariser KM. Borrelia burgdorferi in joint fluid of chronic Lyme arthritis. Ann Intern Med 1986; 104:798-800. 13. Reik L, Steere AC, Bartenhagen NH, Shope RE, Malawista SE. Neurologic abnormalities of Lyme disease. Medicine (Baltimore) 1979;58:281-94. 14. Pachner AR, Steere AC. The triad of neurologic manifestations of Lyme disease: meningitis, cranial neuritis, and radiculoneuritis. Neurology 1985;35:47-53. 15. Pachner AR. Neurologic manifestation of Lyme disease, the new "Great Imitator." Rev Infect Dis 1989;2(Suppl 6):S 1482-6.

Requests for reprints to: Dr David R. Burdge, Room G611, Jean Matheson Pavilion, Shaughnessy Hospital, 4500 Oak St, Vancouver, BC V6H 3N1

16. Steere AC, Batsford WP, Weinberg M, AlexanderJ, Berger HJ, Wolfson S, et al. Lyme carditis: cardiac abnormalities of Lyme disease. Ann Intern Med 1980;93:8-16.

References

17. Barbour AG. The diagnosis of Lyme disease: rewards and perils. Ann Intern Med 1989;1 10:501-2.

1. Steere AC. Lyme disease. N Engl i Med

1989;321:586-96. 2. Steere AC. Borrelia burgdorferi (Lyme disease, Lyme borreliosis). In: Marschall GL, Douglas AG, BennettJC, editors. Principles and practice of infectious diseases. New York, NY: Churchill Livingstone, 1990:1819-27. 3. Steere AC, Malawista SE, Snydman DR, Shope RE, Andiman WA, Ross MR, et al. Lyme arthritis: an epidemic of oligoarticular arthritis in children and adults in three Connecticut communities. Arthritis Rheum 1977;20:7-17. 4. Burgdorferi W, Barbour AG, Hayes SF, Benach JL, Grunwaldt E, DavisJP. Lyme disease - a tick-borne spirochetosis? Science 1982;216:1317-9. 5. PlesmanJ, Mather TN, Sinsley RJ. Duration of tick attachment and Borrelia burgdorferi transmission. 7 Clin Microbial 1987;25:957-8. 6. Lane RS, Lavoie PE. Lyme borreliosis in California: acarological, clinical, and epidemiological studies. Ann NYAcad Sci 1988;539:192-203. 7. Steere AC, Malawista SE. Cases of Lyme disease in the United States: locations correlated with distrihution of Ixodes dammini. Ann Intern Med 1979;9 1:730-3.

1432 Canadian Family Physician VOL 38: June 1992

18. Barbour AG. Laboratory aspects of Lyme borreliosis. Clin Microbiol Res 1988; 1:399-414. 19. Grodzicki RL, Steere AC. Comparison of immunoblotting and indirect enzyme-linked immunosorbent assay using different antigen preparations for diagnosing early Lyme disease. ] Infect Dis 1988;157:790-7.

20. Rahn DW, Malawista SE. Lyme disease: recommendations for diagnosis and treatment. Ann Intern Med 1991;1 14:472-81.

21. Steere AC, Malawista SE, NewmanJN, Spieler PN, Bartenhagen NH. Antibiotic therapy in Lyme disease. Ann Intern Med 1980;93:1-8. 22. Steere AC, Hutchinson GJ, Rahn DW, Sigal LH, CraftJE, DeSanna ET, et al. Treatment of the early manifestations of Lyme disease. Ann Intern Med 1983;99:22-6. 23. Dattwyler RJ, Volkman DJ, Conaty SM, Platkin SP, Luft BJ. Amoxycillin plus probenecid versus doxycycline for treatment of erythema migrans borreliosis. Lancet 1990;336: 1404-6. 24. BurrascanoJ. Late-stage Lyme disease: treatment options and guidelines. Intern Med Specialist 1989; 10: 102-7.