Serosurvey of AntiBabesia Antibodies in Stray Dogs and American Pit Bull Terriers and American Staffordshire Terriers From North Carolina
Stray dogs (n=359) and kennel dogs (n=149) from North Carolina were tested for evidence of antiBabesia antibodies. AntiBabesia antibodies were detected in 21/359 and 22/149 of the stray and kennel dogs, respectively. A total of 57 dogs from both groups were tested for babesiasis by light microscopy and polymerase chain reaction (PCR). Babesia deoxyribonucleic acid (DNA) was detected in 3/28 of the stray dogs and 14/29 of the kennel dogs. When Babesia DNA was detected by PCR, the species-specific PCR results differed from the Babesia species antibody titer results in 6/17 of the PCR-positive dogs. There was no association between antiBabesia antibodies and the presence of ticks. There are currently Babesia gibsoni epizootics affecting American pit bull terrier kennels.
Introduction
Babesiosis is the disease caused by intra-erythrocytic protozoan parasites of the genus Babesia. Babesiosis is typically characterized by hemolytic anemia, thrombocytopenia, fever, and splenomegaly. Babesiasis refers to the presence of circulating Babesia organisms and does not infer whether or not the organisms are causing clinical disease. Historically, canine babesiosis in the United States (US) has been attributed to either Babesia canis (B. canis) or Babesia gibsoni (B. gibsoni), with the vast majority of cases attributed to B. canis vogeli. Since Conrad, et al., reported the first disease outbreak of canine babesiosis caused by a small Babesia in 1991, there have been an increasing number of reports of dogs in the US infected with small Babesia spp.1–6 Babesia gibsoni was presumed to be the only small Babesia spp. to infect dogs; however, recent studies demonstrate that at least three genetically distinct, small Babesia-like organisms or piroplasms naturally infect dogs.478 Based upon evolving data, the nomenclature of these small Babesia-like organisms is likely to undergo revision. In this report, the authors will refer to the Asian genotype as B. gibsoni (GenBank Accession Nos. AF271081, AF271082, AF205636, AF175300, and AF175301),347 the North American genotype as the “western piroplasm” (AF158702 and L13729),79 and the European/Spanish genotype as Theileria annae (T. annae) (AF188001).810 Both B. gibsoni and the western piroplasm have been identified in dogs from North America, while T. annae has only been reported in Europe.810 To date, the western piroplasm has only been reported in dogs from California.111 It is important for the clinician to identify which species and genotype are the cause of babesiosis, since the prognosis for each may be different. Babesia canis vogeli is considered to be an intermediately virulent species for which it is presumed that antibabesial therapy will eliminate the infection. Babesia gibsoni is considered to be a virulent species. Although several antibabesial treatments reduce mortality and clinical signs, none have been able to effectively eliminate infection in dogs.12 However, in one report, the majority of B. gibsoni-infected dogs displayed no outward signs of clinical illness, despite hematological abnormalities such as thrombocytopenia and anemia.6 The western piroplasm is also pathogenic.111 The responsiveness of the western piroplasm to antibabesial therapy has not been well characterized. Theileria annae-infected dogs were often anemic and thrombocytopenic.10 To the authors’ knowledge, the responsiveness to antibabesial therapy for T. annae has not been reported.
The majority of dogs that have been diagnosed with B. gibsoni in the US have been American pit bull terriers (APBT) and American Staffordshire terriers (AST), and they have come from kennels housing multiple dogs.2–6 In California, there was no apparent breed predilection for the western piroplasm.111 This study was undertaken to determine the prevalence of antiBabesia antibodies in three APBT/AST kennels where B. gibsoni had been previously diagnosed, and to detect antiBabesia antibodies and babesiasis in stray dogs in North Carolina (NC).
Materials and Methods
Sample Collection
Samples were collected between May and December of 1999. The age, breed, sex, and the presence or absence of external parasites at the time of collection were recorded for each dog. Each dog’s age was estimated as < or > 6 months based on the presence or absence of deciduous teeth. Collected ticks were stored in 70% ethanol and subsequently identified using a standard key.13
Three regions in NC were sampled for stray dogs. Sera and ethylenediaminetetraacetic acid (EDTA) anticoagulated whole blood were collected from stray dogs housed in animal shelters. A total of 359 stray dogs were sampled; 168 were from eastern NC (sampled between July and September 1999), 140 were from central NC (sampled between May and October 1999), and 51 were from western NC (sampled in July 1999). Every available dog was sampled from each location, with the assumption that sampled dogs were representative of the stray dog populations in the respective areas. This number of subjects would allow a 95% chance of detecting at least one infected/exposed dog if the prevalence was 2% to 5.7%.14
Three APBT/AST kennels located in central NC were selected for study after a single case of B. gibsoni had been previously diagnosed in each kennel. Sera and EDTA anticoagulated whole blood were collected from all kennel dogs. Kennel I housed 59 dogs (sampled in May 1999); kennel II housed 43 dogs (sampled in June 1999); and kennel III housed 47 dogs (sampled in December 1999).
Indirect Fluorescent Antibody Testing
Indirect fluorescent antibody (IFA) testsa were performed as previously described using anticoagulated whole blood obtained from a splenectomized dog that had been infected with either B. canis, B. gibsoni, or the western piroplasm as the antigen sources.15 For each antigen, reciprocal antibody titers ≥64 were considered seroreactive. Since IFA can be associated with a one dilution error, dogs with reciprocal antibody titers of 32 were included for subsequent microscopic examination and PCR testing to minimize potential false-negative IFA results. To examine the specificity of the IFA, all three antigen slides were tested using sera from dogs that had antibodies against Toxoplasma gondii (reactive to a reciprocal antibody titer of 620), Leishmania infantum (reactive to a reciprocal antibody titer of 2,048), Ehrlichia canis (reactive to a reciprocal antibody titer of 8,192), Rickettsia rickettsii (reactive to a reciprocal antibody titer of 2,048), and Bartonella vinsonii (reactive to a reciprocal antibody titer of 2,048).
Microscopic Observation
A total of 57 dogs (including 43 seroreactive dogs with reciprocal antiBabesia antibody titers ≥64, 10 dogs with reciprocal antibody titers of 32, and four nonseroreactive anemic dogs) were tested for babesiasis by microscopic observation. The four nonseroreactive dogs tested were kennel dogs without detectable antiBabesia antibodies that were obviously anemic, with red blood cell volumes <25% after erythrocyte sedimentation. Stainedb thin blood smears were examined by an experienced microscopist at 1,000× for either 30 minutes per animal or until piroplasms were clearly identified. The microscopist was blinded to the polymerase chain reaction (PCR) results at the time of the microscopic examination.
Polymerase Chain Reaction Testing
A total of 57 dogs (including 43 seroreactors, 10 dogs with reciprocal antibody titers of 32, and four nonseroreactive anemic dogs) were tested for babesiasis by PCR amplification as previously described.16 Total deoxyribonucleic acid (DNA) was extracted from 200 μL of EDTA anticoagulated whole blood using a commercially available kit according to the manufacturer’s instructions.c In order to prevent PCR amplicon contamination, sample preparation, reaction setup, PCR reaction, and amplicon detection were all performed in separate areas. Positive and negative controls were used in all processing steps, including the DNA extraction.
Statistical Analysis
An odds ratio (OR), when appropriate, or a Fisher’s exact test, or both, were used to examine potential associations between the presence of antiBabesia antibodies to the following: age, breed, sex, presence of ticks, and kennel/geographic location.17 The stray dogs and kennel dogs were evaluated separately, with the presence of antiBabesia antibodies defined as the dependent variable and either age, breed, sex, presence of ticks, or geographic location as the independent variable. A kappa statistic was used to determine the level of agreement between microscopic observation and PCR tests.
Results
A total of 508 dogs were sampled. Of the stray dogs, 270/359 (75.2%) were of mixed breed, and 89/359 (24.8%) appeared to be purebred. The APBT/AST were the most common purebred stray dogs, accounting for 12/359 (3.3%). Of the stray dogs, 161/359 (44.8%) were male, and 198/359 (55.2%) were female. Approximately 20% (71/359) of the stray dogs were estimated to be <6 months of age. Of the kennel dogs, 138/149 (92.6%) were APBT/AST, 7/149 (4.7%) were American bulldogs, and 4/149 (2.7%) were walker hounds. Of the kennel dogs, 68/159 (42.8%) were male, and 91/159 (57.2%) were female. Approximately 12% (18/149) of the kennel dogs were estimated to be <6 months of age. The kennel owners reported that none of the dogs had traveled outside of the US.
The percentages of stray dogs parasitized by ticks or fleas at the time of sample collection were 27%, 32%, and 65% from western NC, eastern NC, and central NC, respectively. Ticks were available for identification from 86% of the parasitized western NC dogs, and all were adult Dermacentor variabilis (D. variabilis). Ticks were available for identification from 41% of the parasitized eastern NC dogs. These ticks were identified as adult Rhipicephalus sanguineus (R. sanguineus) (52%), nymphal R. sanguineus (2.5%), adult D. variabilis (27.3%), or adult Amblyomma americanum (A. americanum) (4.5%), with some dogs parasitized by more than one species (13.7%). Ticks were available for identification from 23% of the parasitized central NC dogs. These ticks were identified as adult R. sanguineus (54%), nymphal R. sanguineus (4%), adult D. variabilis (20%), or adult A. americanum (12%), with some dogs parasitized by more than one species (10%). Ticks or fleas were not identified on any kennel dogs, nor did the owners report the historical presence of tick attachment on these dogs. However, it was reported that in all of the kennels there was a seasonal problem with biting flies. The owners reported no use of external parasite control.
Indirect Fluorescent Antibody Testing
Twenty-one seroreactive stray dogs with reciprocal antibody titers ≥64 were identified [Table 1]. Sixteen of the seroreactive stray dogs were mixed-breed dogs; two were foxhounds; and one each was a beagle, pointer, and Labrador retriever. Of the seroreactive stray dogs with reciprocal antibody titers ≥64, four were estimated to be <6 months of age, 17 were >6 months of age, eight were male, 13 were female, and five were parasitized by ticks at the time of sampling. Of the five tick-infested seroreactors, adult male and female R. sanguineus were found on two dogs, adult male and female D. variabilis were found on two dogs, and one dog was parasitized by adult male and female D. variabilis and A. americanum. Seven stray dogs had reciprocal antibody titers of 32. Five of these dogs had reciprocal titers of 32 against B. canis; one dog had a reciprocal titer of 32 against B. gibsoni; and one dog had a reciprocal titer of 32 against the western piroplasm.
Twenty-two seroreactive kennel dogs with reciprocal antibody titers ≥64 were identified [Table 2]. The prevalence of dogs with antiBabesia antibodies in kennels I, II, and III were 6.8%, 25.6%, and 14.9%, respectively. Of the seroreactive kennel dogs with reciprocal antibody titers ≥64, two were estimated to be <6 months of age; 20 were estimated to be >6 months of age; eight were male; 14 were female; and none were parasitized by ticks at the time of sampling. All of the seroreactive kennel dogs were APBT/AST. Three kennel dogs had reciprocal antibody titers of 32. Two dogs had titers of 32 against B. gibsoni, and one dog had a titer of 32 against the western piroplasm.
Control canine serum samples that were reactive to Toxoplasma gondii, Leishmania infantum, Ehrlichia canis, Rickettsia ricketsii, or Bartonella vinsonii antigens did not cross-react with any of the three Babesia antigens.
Microscopic Observation
Small piroplasms were identified by microscopic observation in 1/28 stray dogs [Table 1] and in 13/29 of the kennel dogs [Table 2]. No large piroplasms were observed on blood smears.
Polymerase Chain Reaction Testing
A total of 57 dogs (including 43 seroreactors, 10 dogs with reciprocal antibody titers of 32, and four nonseroreactive anemic dogs) were tested by PCR. Three stray dogs were PCR-positive [Table 1]. Discordance between the serological test results and the PCR test results was found in one PCR-positive dog [Table 1; case no. 15] that was only seroreactive against B. canis antigens but was PCR-positive for B. gibsoni. None of the PCR-positive stray dogs were APBT/AST. Fourteen kennel dogs, all APBT/AST, were PCR-positive for B. gibsoni [Table 2]. In five PCR-positive kennel dogs, there was discordance between the serological and PCR test results. Two of these dogs [Table 2; case nos. 12, 14] were only seroreactive against the western piroplasm, and one dog [Table 2; case no. 7] was only seroreactive against B. canis. The other two PCR-positive kennel dogs [Table 2; case nos. 3, 4] with discordant IFA test results were anemic and had no detectable antiBabesia antibodies.
Statistical Analysis
There were no associations between the presence of antiBabesia antibodies and age, breed, sex, presence of ticks, or kennel/geographic location within either sample population. There was a strong agreement (Kappa, 0.9) between the microscopic observation and PCR tests. In all cases where there was disagreement between the microscopic results and PCR, the PCR was positive and microscopic testing was negative.
Discussion
This study is unique as it presents the first serological and molecular data for canine Babesia in the US that includes simultaneous consideration of B. canis, B. gibsoni, and the western piroplasm. The results of this study should further alert practitioners in the US to consider B. gibsoni as a diagnosis in dogs with signs of babesiosis. Scientific reports have documented B. gibsoni infections in dogs from Alabama, Indiana, Oklahoma, and North Carolina.2–6 In addition, the Vector-Borne Disease Testing Laboratory at the North Carolina State University College of Veterinary Medicine has detected B. gibsoni DNA by PCR in canine blood samples from Arizona, California, Florida, Georgia, Kentucky, Maryland, Michigan, Minnesota, Mississippi, Missouri, New Jersey, New York, Ohio, Pennsylvania, South Carolina, Tennessee, Texas, Washington, Wisconsin, and Virginia.d Based on the results of this study and others, when a B. gibsoni-infected dog is identified from a kennel, further investigation is indicated to identify other infected dogs housed on the same property.6
It is important to note that this study demonstrates that there can be discordance between the microscopic, serological, and molecular test results for canine babesiosis. AntiBabesia seroreactivity was not always detectable in infected dogs, and when antiBabesia antibodies were detectable, the highest antibody titer against a given species did not always agree with the PCR detection of DNA in Babesia species. This unpredictable seroreactivity and cross reactivity are likely due to antigenic variation between species, strains, and isolates. Since most IFA tests use whole organism preparations, they usually only consist of a single isolate. Therefore, each test is likely to “favor” isolates that share common or similar immunodominant antigens. It is possible that some B. gibsoni strains actually share more common immunodominant antigens with B. canis or the western piroplasm than they do with each other. It is also possible that the seronegative, PCR-positive dogs were acutely infected and had not yet developed detectable antiBabesia antibody titers. Lastly, coinfection with more than one type of piroplasm cannot be ruled out in the PCR-positive dogs with discordant serological testing results. The significance of seroreactivity against Babesia antigens in conjunction with a negative PCR and microscopic examination result is not known. One possible explanation includes parasitemias below the limit of detection for microscopic examination and PCR, similar to what has been described in bovine babesiosis.18 In order to minimize possible false-positive serological test results in an epidemiological study, other serological surveys have used higher reciprocal antibody titers as their cutoff values: 320 for B. canis and 1280 for the western piroplasm,19 compared to the cutoff value of 64 used in this study. Since the authors’ goal was simply to detect any cases in the stray dog and kennel populations, preferably with confirmation by PCR, microscopic observation, or both, they were willing to accept the possibility of some false-positive seroreactors. The use of antibody titers for the diagnosis of infectious diseases only provides indirect evidence of infection, and the results can be difficult to interpret due to cross-reactive antibodies, antigenic variation between isolates/strains, and interlaboratory variation in testing techniques and antigens. The use of cutoff titers for clinical cases can also lead to an incorrect diagnosis, since detectable antiBabesia antibodies are not always present in dogs that are confirmed to be infected with Babesia spp. by PCR, microscopic observation, or both.
Interestingly, western piroplasm DNA was not detected in any dog in this study. This organism may not be endemic to North Carolina or dogs may clear the infection spontaneously. The seroreactivity to the western piroplasm that was detected in this study may represent cross-reactivity to other Babesia genotypes or other nonBabesia organisms, or it may represent a level of parasitemia that may be below the limit of PCR detection. While there were two B. gibsoni PCR/microscopic observation-positive dogs that were only seroreactive to the western piroplasm, at least the same number of dogs were PCR/microscopic observation-positive and without any detectable antiBabesia antibodies. Since PCR/microscopic observation testing was not performed on all nonseroreactive dogs, statistical analysis cannot be performed to determine whether or not the antiwestern piroplasm antibodies identified in this study were actually associated with detectable Babesia DNA by PCR or were simply false-positive results. Since sera from some dogs from Australia, a country where it was presumed that the western piroplasm is not endemic, had reciprocal antibody titers against western piroplasm antigens as high as 160, Yamane, et al., suggested using a reciprocal antibody titer of 320 as a cutoff for dogs with clinical signs of babesiosis, and a reciprocal antibody titer of 1,280 for epidemiological studies when using the western piroplasm as an antigen source.19
The route of B. gibsoni transmission in the US has not been identified. In other endemic areas, tick transmission is believed to be the primary route of infection.12 Based on the absence of ticks at the time of sampling and the reported historical lack of tick infestation, it does not appear that tick transmission is the primary route of transmission in these kennels. Since detailed histories of the stray dogs infected with or exposed to Babesia were not available, it was impossible to determine how they were infected. Since babesiosis can be chronic, the ticks found on the dogs at the time of sample collection are not necessarily the same ticks that transmitted the infection to the dogs. Other than ticks, the potential routes of transmission for B. gibsoni include vertical transmission, direct transmission, and mechanical transmission. Vertical transmission seems plausible.20 The authors have also diagnosed B. gibsoni in puppies as young as 10 days of age, a time interval that is shorter than the prepatent period following tick transmission.e Direct transmission via blood transfer during fighting contact also seems plausible. The authors have diagnosed at least four cases of B. gibsoni infection in nonAPBT/AST dogs after they have been in an accidental fight with a carrier animal,f and other dogs have been diagnosed with B. gibsoni after being attacked by APBT/AST dogs.5 Possible mechanical vectors to consider include biting flies and needles used to vaccinate multiple animals. There is a growing body of evidence that R. sanguineus is a potential vector for B. gibsoni;21–24 however, definitive transmission studies have yet to be performed. Future studies that examine the transmission of B. gibsoni among dogs in the US are warranted.
Conclusion
Both B. canis vogeli and B. gibsoni appear to be endemic to North Carolina. There are B. gibsoni epizootics currently affecting APBT and AST kennels. This study also demonstrates that no single diagnostic test will confirm all cases of canine babesiosis. Therefore, if canine babesiosis is suspected, the authors recommend that serology, microscopic observation, and PCR be performed. Serology against both B. canis and B. gibsoni should be performed, and if serology against the western piroplasm is performed, the guidelines of Yamane, et al., should be considered. Caution is warranted in the use of antibody titers for Babesia species identification, since the highest antibody titer did not always correlate with the microscopic and PCR testing results. Lastly, PCR testing should screen for all known genotypes.
The Vector-Borne Disease Diagnostic Laboratory, College of Veterinary Medicine, North Carolina State University
Leukostat; Fisher Scientific, Suwanee, GA
Qiagen; Valencia, CA
Birkenheuer AJ, Levy MG, Breitschwerdt EB. Unpublished observations. The Vector-Borne Disease Diagnostic Laboratory, College of Veterinary Medicine, North Carolina State University
Birkenheuer AJ, Breitschwerdt EB. Unpublished observations.
Birkenheuer AJ, Greene CE, Breitschwerdt EB. Unpublished observations.
Acknowledgment
The authors thank Dr. Michael Lappin for kindly providing the canine Toxoplasma gondii reactive serum.
Contributor Notes
