Serological studies on the infection of dogs in

Serological studies on the infection of dogs in Ontario with Borrelia burgdorferi, the etiological agent of Lyme disease Harvey Artsob, Ian K. Barker, Richard Fister, Gregory Sephton, Daryl Dick, John A. Lynch, Doug Key Abstract A serological study was undertaken to determine whether dogs in Ontario are being exposed to Borrelia burgdorferi, the etiological agent of Lyme disease. This study consisted of a survey of randomly selected dogs and testing of diagnostic submissions from candidate Lyme disease cases. The survey of 1,095 dogs, bled between January 1988 and August 1989, revealed a total of 65 (5.9070) enzyme-linked immunosorbent assay (ELISA) reactors, of which 22 had immunofluorescent antibody assay (IFA) titers 2 1:32. All but one of the IFA-positive and 10 of the ELISA-positive, IFA-negative sera were further tested by western blot. Eight western blot positive and three equivocal reactors were obtained. Three of the eight confirmed reactors had visited areas known to be endemic for Lyme disease, leaving five reactors that might have been infected in previously undocumented areas for B. burgdorferi activity in Ontario. Diagnostic submissions of sera from 223 dogs were received between August 1987 and February 1992. Test results revealed 21 (9.47o) IFA reactors, of which only six had significant titers (> 1:256) and were reactive by an immunodot Borrelia test. All six dogs had travelled to known Lyme endemic areas. Based on results obtained from this study, it seems likely that the agent of Lyme disease is not widespread in Ontario. Zoonotic Diseases, National Laboratory for Special Pathogens, Bureau of Microbiology, Laboratory Centre for Disease Control, Tunney's Pasture, Ottawa, Ontario K1A OL2 (Artsob, Dick); Department of Pathology, Ontario Veterinary College, University of Guelph, Guelph, Ontario N1G 2W1 (Barker); Tufts Veterinary Diagnostic Laboratory, 200 Westboro Road, North Grafton, Massachusetts, USA 01536 (Fister); Veterinary Practitioner, P.O. Box 8, Powassan, Ontario POH 1Z0 (Sephton); Veterinary Laboratory Services, Ontario Ministry of Agriculture and Food, Guelph, Ontario NIH 6R8 (Lynch, Key). This work was supported in part by grants to Dr. I.K. Barker from Health and Welfare Canada and the Ontario Ministry of Health. R6sum6 etude s6rologlque de Borrella burgdorferl, agent causal de la maladle de Lyme chez les chiens en Ontarlo Une etude serologique a ete effectuee afin de determiner Ia prevalence de Borrelia burgdorferi chez les chiens en Ontario. Cette enquete comporte d'une part, un echantillonnage d'animaux choisis de facon non selective et, d'autre part, des demandes specifiques d'analyses provenant d'animaux susceptibles de presenter la maladie de Lyme. Les prelevements non selectifs ont ete effectues entre janvier 1988 et ao&t 1989 sur 1095 chiens. Les resultats indiquent que 5,9 % (65 chiens) ont reagi positivement aux analyses par methode enzymatique ELISA; 22 des animaux avaient un titre d'anticorps. 1:32. Des analyses par methode immunoblot western ont ete effectuees sur tous les cas, sauf un, ayant des titres d'anticorps significatifs et sur 10 des sujets positifs au test ELISA, mais n'ayant pas un titre d'anticorps significatif. Les r6sultats ont demontre huit reactions positives et trois equivoques. Trois des huit chiens presentant un resultat positif au test immunoblot western avaient visite des regions endemiques pour la maladie de Lyme. Il reste cinq animaux qui ont pu s'infecter dans des regions ontariennes non documentees pour B. burgdorferi. Deux cent vingt-trois serums de chiens, preleves entre le mois d'aofut 1987 et le mois de fevrier 1992, ont ete soumis pour diagnostic. Les resultats ont demontre un titre d'anticorps dans 9,4 % des cas (21 chiens). De ces animaux, seulement six avaient un titre d'anticorps significatif (>1:256) et etaient positifs au test par immunodot Borrelia. Tous les six avaient visite des regions endemiques. Les resultats de cette etude indiquent que l'argent causal de la maladie de Lyme n'est pas tres repandu en Ontario. (Traduit par Dr Thirkse Lanthier) Can Vet J 1993; 34: 543-548 Introduction L yme disease is a tick-transmitted zoonosis caused by the spirochete Borrelia burgdorferi. Lyme disease was initially recognized as a distinct clinical entity in Connecticut in 1975 (1). Although cases of 543

Lyme disease have been reported from 43 states in the United States, the disease has been documented primarily from three major areas: Massachusetts to Maryland in the Northeast, Wisconsin and Minnesota in the Midwest, and California and Oregon in the West (2). Lyme disease has also been reported from Europe (3), Asia (4,5), and Australia (6). More recent reports suggest that Lyme disease may occur in Africa (7,8) and, possibly, in South America (9). The extent of the occurrence of Lyme disease in Canada is presently unclear. As of June 1990, provisional data indicated that 103 human cases have been reported from seven Canadian provinces, but a large number of these cases were likely contracted outside the country (10). Borrelia burgdorferi has been isolated from whitefooted mice (Peromyscus leucopus) and ticks (Ixodes dammini, Dermacentor variabilis) collected at Long Point in southern Ontario (11). In addition, an isolate of B. burgdorferi was obtained from an I. dammini removed from a cat in Prince Edward Island (12). Two relevant tick vectors of Lyme disease have been documented in Canada (13). Ixodes dammini has been reported from six provinces: Manitoba, Ontario, Quebec, Prince Edward Island, New Brunswick, and Nova Scotia. All three active stages of the tick (larva, nymph, adult) have been demonstrated at Long Point, Ontario, and, more recently, larvae and nymphs were recovered in Marchand, Manitoba (14). Ixodes pacificus has been found in British Columbia. Borrelia burgdorferi has been associated with infection in dogs (15-21). Arthritis has been the most commonly reported clinical sign, but other manifestations have included anorexia, fever, fatigue, renal disorder, and heart block. Asymptomatic infections in dogs are apparently also common in areas where L dammini is endemic. Since dogs become infested with ticks more readily than humans, serosurveys on dogs have been useful in determining the possible presence of B. burgdorferi in different localities, and results have been reported from California (22), Connecticut (16), New Jersey (23), New York (18), North Carolina (24), Wisconsin (25), and the United Kingdom (26). Although field studies in Ontario have confirmed populations of L dammini only at Long Point, 76 human cases of Lyme disease have been reported in the province from 1984 to 1989, including 44 individuals who had no history of travel to an endemic area outside Ontario (27). Eight of these 44 cases involved people who had travelled to the known endemic area of Long Point. We report herein a study of dogs in Ontario consisting of a serosurvey of randomly selected dogs and serological testing of diagnostic submissions from candidate Lyme disease cases. This study was undertaken with the goal that dogs seropositive to B. burdgorferi might provide useful information concerning the distribution of this agent and its tick vector in the province. Materials and methods Serological survey One thousand and ninety-five canine sera were surveyed, of which 1,005 were obtained from the clinical pathology laboratories of the Ontario Veteri- 544 nary College, University of Guelph, and the Guelph Veterinary Services Laboratory, Ontario Ministry of Agriculture and Food, between January 1988 and August 1989, inclusive. These sera were arbitrarily chosen from specimens that had passed through these laboratories during that period. No selection bias was instituted based on the age, sex, breed, geographical location, or clinical signs of the dogs. The population examined was drawn from numerous localities within southern Ontario bounded on the north by a line drawn through Sault Ste-Marie, North Bay, and Ottawa. In addition, there were 14 sera from dogs from Thunder Bay in northwestern Ontario, and 30 from Timmins in northeastern Ontario. Ninety of the 1,095 sera were obtained from dogs by a veterinary practitioner in Powassan. Fifty-six of these were collected from dogs residing in or near Powassan from January to February 1988, while 34 were taken from dogs in Huntsville in December 1988. The latter location is near the town of Utterson, where a resident seroconverted to B. burgdorferi in 1986 (28). Diagnostic submissions Sera from 223 dogs having signs possibly consistent with Lyme disease or with exposure to B. burdgorferi were received between August 1987 and February 1992. Included were sera from 203 dogs submitted through the clinical pathology laboratories at the Ontario Veterinary College and Veterinary Services Laboratory, Guelph, and 20 were sent directly from veterinarians in Toronto, Ottawa, and Hamilton. These diagnostic submissions comprised 5 in 1987, 29 in 1988, 61 in 1989, 70 in 1990, 51 in 1991, and 7 in 1992. Case histories were not supplied for 53 dogs. Eightyseven (51.2%) of the remaining 170 animals -had arthritis or related conditions, such as lameness or joint pain. Less commonly reported signs included lethargy, anorexia, pyrexia, and cardiac or central nervous system manifestations. Forty-three (25.3Gb) of the dogs were, in fact, healthy, but tests were requested because of exposure to ticks, travel to a known Lyme disease endemic area, or a member of the family to which the pet belonged being seropositive for B. burgdorferi. Serological testings Strains used - Borrelia burgdorferi, strain B3 1, was used for antigen preparation for enzyme-linked immunosorbent assay (ELISA) and immunofluorescent antibody (IFA) tests. This strain, identified as ATCC 35210 was purchased from the American Type Culture Collection (ATCC), Rockville, Maryland, USA. Borrelia burgdorferi, strain 2591, kindly supplied by Dr. L. Magnarelli, Connecticut Agricultural Research Station, New Haven, Connecticut, USA, was used for western blot tests. The bacteria were cultured in Barbour-Stoemer-Kelly II (BSKII) medium (29). Enzyme-linked immunosorbent assay - Sera were tested for IgG antibodies to B. burgdorferi by ELISA using the procedure described by Wilkinson (30) for testing human sera but substituting an antidog conjugate, i.e., peroxidase-labelled goat antidog immunoglobulin (Cappel, Malvern, Pennsylvania, USA), for antihuman conjugate.

Table 1. Pattern of reactivity to Borrelia burgdorferi by immunofluorescent antibody and western blot assays of ELISA positive sera from the serological survey of dogs In Ontario Dog number IFA titer Western blot result DLS384 1:1024 Positive DLS464 1:512 Positive DLS6 1:256 Positive DLS456 1:256 Positive DLS485 1:256 Negative DLS1447 1:256 Negative DLS1137 1:128 Negative DLS1496 1:128 Negative DLS78 1:64 Negative DLS307 1:64 Equivocal DLS337 1:64 Negative DLS377 1:64 Negative DLS 1373 1:64 Positive DLS1573 1:64 Negative DLS 1724 1:64 Negative DLS1 170 1:32 Negative DLS1244 1:32 Positive DLS1269 1:32 Negative DLS1368 1:32 Positive DLS1551 1:32 Negative DLS1610 1:32 Negative DLS203 -a Negative DLS286 - Positive DLS289 - Negative DLS345 - Negative DLS355 - Negative DLS1273 - Negative DLS 1393 - Negative DLS1416 - Equivocal DLS1600 - Equivocal DLS1651 - Negative a = IFA titer < 1:32 Reference dog sera (one positive and one negative) were added to duplicate wells of all microtiter plates. Test results were expressed as the ratio of the optical density (OD) of the test serum to the OD of the positive reference serum. An OD 2 2 standard deviations above the mean OD ratio of 100 normal dogs was considered a positive test result (30). Immunofluorescent antibody assay The IFA test - was performed on all ELISA-positive sera by the method of Wilkinson (30) using a goat antidog fluorescein isothiocyanate-labelled anti-igg conjugate (Cappel). Sera were tested at starting dilutions of 1:32 with subsequent twofold dilutions until endpoint titers were obtained. Western blot The western blot was performed as previously described (31) and using a peroxidase conjugated goat antidog IgG conjugate (Cappel). Western blot results were interpreted using the criteria of Karlsson et al (32). An immunoblot was considered positive if it contained a band corresponding to the 41 kilodalton (kda) flagellum protein and a band corresponding to a designated low molecular weight protein <41 kda. A serum was considered equivocal if the blot showed a 41 kda band and negative if it showed less than four bands. Immunodot canine Borrelia test This assay (Immunodot Lyme Test, General Biometrics, Inc., San Diego, California, USA) was used to test diagnostic submissions that had IFA titers to B. burgdorferi. Each assay strip included a positive antibody control; a negative antibody control; B. burgdorferi (strain B3 1), a partially purified sonicate of whole organism blended with a purified 39 kda recombinant protein (P39) that is specific for B. burgdorferi; a partially purified sonicate of whole organism; P39 recombinant protein alone; and flagellin extracted and purified from B. burgdorferi (strain B3 1). The test was run following kit instructions. Sera were considered reactive for B. burgdorferi antibodies if reactions were obtained with whole organism + P39, whole organism alone, and P39 recombinant antigen alone. Sera were considered reactive for Borrelia and related species if no reaction was observed to P39 alone, but reactions were observed in various combinations to whole organism + P39, whole organism alone, and flagellin antigen alone. Results Serological survey Sixty-five (5.9%) of the 1,095 sera tested were positive by ELISA. All sixty-five reactors were tested by IFA. Forty-three sera were nonreactive (IFA titers <1:32), whereas the remaining 22 sera had the following titers: 1:32 six, 1:64 eight, 1:128 two, 1:256 four, 1:512 one, and 1:1024 one. Western blot analyses were undertaken on 21 of the 22 canine sera that were ELISA-positive with IFA titers 2 1:32 and on 10 sera that were ELISA-positive, but IFA-negative. Eight western blot positive and three equivocal reactors were obtained (Table 1). Four of the six dogs that were positive on ELISA and had IFA titers 2 1:256 were positive on western blot, whereas only four of the 25 ELISA reactors with IFA titers < 1: 128 were positive on western blot. The geographical locations and clinical histories of the eight western blot reactors are presented in Table 2. Three dogs, DLS 6, DLS 384, and DLS 1368, had been to known Lyme disease endemic areas (Long Point, Ontario; the New England states; and central Pennsylvania, respectively), where they likely became infected. Three of the remaining five dogs, DLS 1373, DLS 456, and DLS 1244, resided in Mildmay, Clifford, and Harriston, respectively, which are situated within approximately 20 miles of each other (43055 'N, 8000'W); the Kitchener dog, DLS 464, was approximately 35 miles from Harriston (43025'N, 80 30'W). None of these four dogs had travelled to known Lyme disease endemic areas in the United States or Canada, although dog DLS 456 had been to Chicago for breeding purposes (Table 2). Finally, dog DLS 286 resided in North Bay (46 20'N, 79025'W) and had lived in this part of northern Ontario for its entire life. Clinical histories of the three dogs that had been to known Lyme disease endemic areas revealed that one dog, DLS 6, had no documented signs, one dog had salivary adenocarcinoma, while the third dog, DLS 384, was described as having had Lyme disease in Massachusetts, but no signs were related. None of the 545

Table 2. Geographical locations and clinical histories of Borrelia burgdorferl ELISA-positive, western blot reactive dogs from the serological survey of dogs in Ontario Dog Immunofluorescent Geographil number antibody titer location alinical history DLS286 <1:32 North Bay, Lethargic, inappetant, no travel mildly polydypsic DLS1244 1:32 Harriston, Chronic uveitis, corneal no travel ulcer associated with ectopic cilia DLS1368 1:32 Port Elgin Salivary adenocarcinoma and Mississaugaa DLS1373 1:64 Mildmay Pregnant DLS6 1:256 Port Doverb Asymptomatic DLS456 1:512 Cliffordc Entropion, no systemic disease DLS464 1:512 Kitchener Seizures, dental plaque, and Muskoka increased lung sounds DLS384 1:1024 Wasaga Beachd Had Lyme disease in Massachusetts adog had visited central Pennsylvania, but otherwise had not travelled outside Ontario bdog visited Long Point, Ontario, which is a known Lyme disease endemic area. Dog was treated for tick infestation CDog had travelled to Chicago for breeding purposes ddog was in known Lyme disease endemic area in New England, United States Table 3. Serological reactors to Borrella burgdorferi for diagnostic submissions of Ontario dogs from August 1987 to February 1992 Dog number IFA titer Immunodot result DL9 1:1024 Antibodies to Borrelia burgdorferi DL18/91 1:512 Antibodies to Borrelia burgdorferi DL29 1:256 Antibodies to Borrelia and related speciesb DL102 1:256 Antibodies to Borrelia burgdorferi DL20/91 1:256 Antibodies to Borrelia burgdorferi DL29/91 1:256 Antibodies to Borrelia burgdorferf DL13 1:128 Not tested DL86 1:128 Negative DL25 1:64 Not tested DL28 1:64 Not tested DL216 1:64 Negative DL217 1:64 Negative DL72 1:32 Negative DL78 1:32 Negative DL94 1:32 Negative DL106 1:32 Negative DLII1I :32 Negative DL197 1:32 Negative DL214 1:32 Negative DLI I/91 1:32 Negative DL I1/92 1:32 Negative areacted to whole organism and to P39 recombinant protein considered specific for B. burgdorferi breacted to whole organism but not to P39 recombinant protein. Therefore, this might represent a cross-reaction with a related spirochete CA convalescent serum from this dog taken three months postantibiotic treatment had an IFA titer of 1:32 and gave an immunodot result indicative of antibodies to Borrelia and related species remaining five dogs had signs that could be clearly linked to Lyme disease, although the general signs of lethargy and inappetent behavior (DLS 286) and clinical manifestations involving the eyes (DLS 1244) have been reported for Lyme disease in people (33). 546 Diagnostic submissions Sera from 223 dogs obtained through diagnostic submissions were examined using IFA as the initial screening test. Twenty-one dogs exhibited IFA reactivity with the following titers: 1:32 nine, 1:64 four, 1:128

Table 4. Travel and clinical histories of Borrelia burgdorferi immunodot reactive dogs with significant IFA titers from diagnostic submissions in Ontario Dog number Date submitted Travel Clinical history DL9 May 30, 1988 New Yorka Suppurative arthritis left tarsocrural joint compatible with Lyme disease; small bowel diarrhea; moist dermatitis DL29 July 11, 1988 New Jerseyb No clinical abnormality DL102 Sept. 26, 1989 Wisconsin Lethargic, anorexic, sore in hindquarters DL18/91 May 29, 1991 United Statesc No clinical abnormality DL20/91 May 29, 1991 United Statesc No clinical abnormality DL29/91 July 11, 1991 Long Point Lethargy, swelling of right carpus, joint pain athis dog was diagnosed as having Lyme disease in New York, 45 days previously bthis dog was a New Jersey resident temporarily in Canada for training cthese dogs were reported to have been in known Lyme disease endemic areas in the United States. No direct follow-up possible two, 1:256 four, 1:512 one, and 1: 1024 - one. Sera from 18 of these dogs were further tested by the immunodot canine Borrelia test (Table 3). An excellent correlation was observed between the IFA titers and immunodot reactivity. All dogs with IFA titers c 1:128 were nonreactive by immunodot. Sera from five of the six dogs with IFA titers > 1:256 demonstrated a pattern indicating antibodies to B. burgdorferi, while the sixth showed antibdies to Borrelia and related species. Travel and clinical histories of the six immunodot reactive dogs are presented in Table 4. All dogs were likely infected in areas where Lyme disease activity is currently recognized. Discussion Several serosurveys of dogs have been undertaken in the United States in which sera with IFA titers 1:64 have been considered positive. Using this criterion, the following prevalence rates have been found in dogs in known Lyme disease endemic areas: 26Gb in Mendocino and Humboldt counties, California (22), 28.67o and 66.5% in Connecticut (16,18), 42.67o in Monmouth county, New Jersey (23), 76.3% in the lower Hudson valley, New York (18), and 537o in Wisconsin (25). By contrast, a prevalence of seroreactors more comparable with this serosurvey was reported by Greene et al (24) in North Carolina, where only 3.6% and 2.7% of sera taken in 1983 and 1985, respectively, were positive. At that time, only a few human cases of Lyme disease had been reported in North Carolina. It was recognized at the outset of this study that some false serological positives would likely be encountered. Previous experience of testing over 2,000 human sera with this in-house ELISA has indicated that the assay has a high degree of sensitivity but a false positive rate of about 5.0Vo (H. Artsob, unpublished observations). False positive seroreactions to B. burgdorferi in people have been observed due to other spirochetes and in patients with autoimmune disease, infectious mononucleosis, Rocky Mountain spotted fever, and other illnesses (34). Given the facts that a positive serological test can be expected to have a low predictive value in areas where Lyme disease does not occur (35) and that the only conclusive evidence to date for the occurrence of B. burgdorferi in Ontario is at Long Point (11), further serological documentation of ELISA survey reactors was done. Western immunoblotting was undertaken since the western blot is generally considered to be the diagnostic standard for confirming B. burgdorferi infections of dogs (36). Only eight western blot confirmed and three equivocal reactors were demonstrated. Western blots were not performed on one ELISA-positive, IFApositive (titer 1:64) dog and on thirty-three ELISApositive, IFA-negative dogs due to the unavailability of adequate serum. Thus, a few confirmed reactors may have been missed. However, since only one of the ten dogs that were ELISA-positive but IFA-negative and further tested was western blot positive, it is likely that only three or four additional western blot reactors would have been found in the serosurvey. These results likely do not represent the true prevalence of infection of dogs in Ontario with B. burgdorferi, due to the problem of false positive serology, and the nonrandom method of selection of serum for testing. Nevertheless, by the criterion of a positive reaction having to be confirmed by western blot as evidence of infection, it is clear that the exposure rate of the dogs from Ontario that were tested is extremely low: three of the eight western blot confirmed reactors had been to known Lyme endemic areas, leaving only five reactors that might have been infected in previously undocumented areas for B. burgdorferi activity in Ontario. The findings from the diagnostic submissions supported the results of the serosurvey in that little evidence was obtained for widespread exposure to B. burgdorferi. At no time during the four and onehalf year diagnostic surveillance period was a case of Lyme disease discovered that had been contracted in other than a currently known endemic area in Canada or the United States. The asymptomatic (DLS 6, Table 2) and symptomatic (DL 29/91, Table 4) infec- 547

tions contracted at Long Point are not surprising, since B. burgdorferi is known at this site (11) and eight symptomatic human infections were reported from here between 1984 and 1989 (27). It does indicate, however, that careful monitoring of dogs for seroreactivity to B. burgdorferi may be useful for determining in which areas people are at risk for infection. The limited Lyme disease activity surmised from this study is consistent with data from Barker et al (11). They documented the epidemiologically relevant deer tick, L dammini, at Long Point National Wildlife Area and Long Point Provincial Park from 1987 to 1989, but found no evidence of I. dammini at 23 other Ontario localities. Although investigation of canine sera to date does not provide evidence that Lyme disease is widespread in Ontario, it is useful to maintain diagnostic vigilance by testing symptomatic dogs, as well as humans, for Lyme disease. This will help to document any spread of B. burgdorferi to what are currently believed to be nonendemic areas. The current data suggest that sporadic exposure to B. burgdorferi is occurring at a sufficiently low incidence in most areas of the province to make widespread field studies seeking B. burgdorferi activity relatively unproductive. Acknowledgments We thank Dr. J.F. Anderson, Connecticut Agricultural Experiment Station, New Haven, Connecticut, for kindly supplying Borrelia burgdorferi antibody positive and negative dog sera for this study; Ms. Sanja Huibner for technical assistance; Mr. Ev Grift of the Ontario Veterinary College clinical pathology laboratory and Dr. Brent Hoff of the OMAF VLS clinical pathology laboratory, their staffs, and Mr. Gareth Jones for their cooperation in collecting sera and John McCleary for tracing patient records. References 1. Steere AC, Malawista SE, Snydman DR, et al. Lyme arthritis: an epidemic of oligoarticular arthritis in children and adults in three Connecticut communities. Arthritis Rheum 1977; 20: 7-17. 2. Tsai TS, Baily RE, Moore PS. National surveillance of Lyme disease 1987-1988. Conn Med 1989; 53: 324-326. 3. Stanek G, Pletschette M, Flamm H, et al. European Lyme borreliosis. Ann NY Acad Sci 1988; 539: 274-282. 4. Chengxu A, Yuxin W, Yungguo Z, et al. Clinical manifestations and epidemiological characteristics of Lyme disease in Hailin County, Heilongjiang Province, China. Ann NY Acad Sci 1988; 539: 302-313. 5. Kawabata M, Baba, S, Iguchi K, Yamaguti N, Russell H. Lyme disease in Japan and its possible incriminated tick vector, Ixodes persulcatus. J Infect Dis 1987; 156: 854. 6. Stewart A, Glass J, Patel A, Watt G, Cripps A, Clancy R. Lyme arthritis in the Hunter Valley. Med J Aust 1982; 1: 139. 7. Russell PK, Woody JN. Tropical medicine studied overseas. US Med 1990; 26: 11-12. 8. Schafrank SN, Kurban AK, Martone G. Lyme disease acquired in southeast Africa. Arch Dermatol 1990; 126: 685-286. 9. Need J, Escamilla J. Lyme disease in South America? J Infect Dis 1991; 163: 681-682. 10. MacKenzie SG. Update - Lyme disease in Canada. Can Dis Wkly Rep 1990; 16: 141-142. 11. Barker IK, Surgeoner GA, Artsob H, et al. Distribution of the Lyme disease vector, Ixodes dammini (Acari:Ixodidae) and isolation of Borrelia burgdorferi in Ontario, Canada. J Med Entomol 1992; 29: 1011-1022. 12. Artsob H, Garvie M, Cawthorn RJ, et a!. Isolation of the Lyme disease spirochete, Borrelia burgdorferi, from Ixodes dammini (Acari: Ixodidae) collected on Prince Edward Island, Canada. J Med Entomol 1992; 29: 1063-1066. cvi 13. Costero A. Identification of the Lyme disease vector in Canada. Can Dis Wkly Rep 1990; 16: 142-147. 14. Galloway TD, Christie JE, Sekla L, Stackiw W. Current status of the Lyme borrelliosis vector, Ixodes dammini, in Manitoba. Can Dis Wkly Rep 1991; 17: 259-260. 15. Lissman BA, Bosler EM, Camay H, Ormiston BG, Benach JL. Spirochete-associated arthritis (Lyme disease) in a dog. J Am Vet Med Assoc 1984; 185: 219-220. 16. Magnarelli LA, Anderson JF, Kaufmann AF, Lieberman LL, Whitney GD. Borreliosis in dogs from southern Connecticut. J Am Vet Med Assoc 1985; 186: 955-959. 17. Kornblatt AN, Urband PH, Steere AC. Arthritis caused by Borrelia burgdorferi in dogs. J Am Vet Med Assoc 1985; 186: 960-964. 18. Magnarelli LA, Anderson JF, Schreier AB, Ficke CM. Clinical and serologic studies of canine borreliosis. J Am Vet Med Assoc 1987; 191: 1089-1094. 19. Levy SA, Duray PH. Complete heart block in a dog seropositive for Borrelia burgdorferi. Similarily to human Lyme carditis. J Vet Intern Med 1988; 2: 138-144. 20. Grauer GF, Burgess EC, Cooley AJ, Hagee JH. Renal lesions associated with Borrelia burgdorferi infection in a dog. J Am Vet Med Assoc 1988; 193: 237-239. 21. Roush JK, Manley PA, Dueland RT. Rheumatoid arthritis subsequent to Borrelia burgdorferi infection in two dogs. J Am Vet Med Assoc 1989; 195: 951-953. 22. Teitler J, Madigan J, Derock E, Pedersen N, Carpenter T, Franti C. Prevalence of Borrelia burgdorferi antibodies in dogs in northern California. Risk factors and zoonotic implications. Ann NY Acad Sci 1988; 539: 500-503. 23. Schulze TL, Bosler EM, Shisler JK, Ware IC, Lakat MF, Parkin WE. Prevalence of canine Lyme disease from an endemic area as determined by serosurvey. Zentralbl Bakteriol [A] 1986; 263: 427-434. 24. Greene RT, Levine JF, Breitschwerdt EB, Berkhoff HA. Antibodies to Borrelia burgdorferi in dogs in North Carolina. Am J Vet Res 1988; 49: 473-476. 25. Burgess EC. Natural exposure of Wisconsin dogs to the Lyme disease spirochete (Borrelia burgdorferi). Lab Anim Sci 1986; 36: 288-290. 26. Liu YM, Bennett CE, White JE, Waitkins SA. Lyme disease in the United Kingdom. An ELISA study of dog sera with antigen comparison between the American B31 strain and a British IWG strain. Ann NY Acad Sci 1988; 539: 465-467. 27. Le Ber C, Njoo H. Indigenous Lyme disease in Ontario. Public Health Epidemiol Rep Ont 1990; 1: 54-59. 28. Laxer RM, Artsob H. Seroconversion to Borrelia burgdorferi in a patient with Juvenile Arthritis in Ontario, Canada. J Rheumatol 1988; 15: 1580-1582. 29. Barbour AG. Isolation and cultivation of Lyme disease spirochetes. In: Steere AC, Malawista SE, Craft JE, Fischer DK, Garcia-Blanco M, eds. Lyme Disease. First International Symposium. Yale J Biol Med 1984; 57: 521-525. 30. Wilkinson HW. Immunodiagnostic tests for Lyme disease. In: Steere AC, Malawista SE, Craft JE, Fischer DK, Garcia-Blanco M, eds. Lyme Disease. First International Symposium. Yale J Biol Med 1984; 57: 567-572. 31. Fister RD, Weymouth LA, McLaughlin JC, Ryan RW, Tilton RC. Comparative evaluation of three products for the detection of Borrelia burgdorferi antibody in human serum. J Clin Microbiol 1989; 27: 2834-2837. 32. Karlsson M, Mollegard I, Stiernstedt G, Wretlind B. Comparison of western blot and enzyme-linked immunosorbent assay for diagnosis of Lyme Borreliosis. Eur J Clin Microbiol Infect Dis 1989; 8: 871-877. 33. Steere AC, Bartenhagen NH, Craft JE, et al. Clinical manifestations of Lyme disease. Zentralbl Bakteriol [A] 1986; 263: 201-205. 34. Magnarelli LA. Serologic diagnosis of Lyme disease. Ann NY Acad Sci 1988; 539: 154-161. 35. Sorveillo FJ, Nahlen B. Lyme disease. Letter to the Editor. N Engl J Med 1990; 322: 474-475. 36. Lindenmayer J, Weber M, Bryant T, Marquez E, Onderdonk A. Comparison of indirect immunofluorescent antibody assay, enzyme-linked immunosorbent assay and western immunoblot for the diagnosis of Lyme disease in dogs. J Clin Microbiol 1990; 28: 92-96. 548