Results of this study demonstrate that concurrent or sequential infection with A. platys and E. canis can impact the hematological changes induced by these pathogens and can also alter the anticipated host immune response that would be induced following exposure to only one organism. Simultaneous infection with E. canis and A. platys in dogs resulted in a more pronounced anemia and thrombocytopenia, when compared to the sole infection with either pathogen. Both sequential and simultaneous infections with A. platys and E. canis produced an enhanced immune response to A. platys when compared to infection with A. platys alone. Also, co-infection with E. canis and A. platys appeared to result in a more persistent A. platys infection than was observed in those dogs that were infected only with A. platys. While the dogs in this study were infected experimentally, there is substantial evidence to support natural exposure to and infection with multiple tick-borne pathogens in dogs [5–8, 10, 11, 23–27]. Under natural conditions, tick transmission potentially influences the course of infection and clinical manifestations, and is therefore a limitation of experimental infection studies. It is likely that co-infection or sequential infections contribute to some of the "atypical" manifestations that have been historically and clinically attributed to single pathogen infections.
In this study, the hematologic effects of infection with only A. platys or only E. canis were similar to those previously reported [28–31]. The cyclic nature of the thrombocytopenia reported in A. platys infected dogs was not clearly demonstrated in this study due to the comparatively low frequency (e.g. twice weekly) in which platelet concentration was measured, and due to the effect of averaging platelet concentrations from multiple dogs per study group at a point in time. When compared directly, the initial decrease in platelet concentrations (~day 10 PI) occurred more rapidly in dogs infected with only A. platys, as compared to dogs infected with only E. canis. This suggests that each of these organisms may induce pathophysiologically different mechanisms that contributed to the thrombocytopenia documented in these dogs. However, the more rapid onset of thrombocytopenia in A. platys infected dogs may reflect a difference in either the strain, dosage or the specific isolate of the organisms used in this study. Nevertheless, compared to E. canis, which induces thrombocytopenia in association with the development of anti-platelet antibodies, A. platys directly infects platelets and may have a more immediate effect on the platelet circulating half-life [32–34].
An unexpected alteration in the pattern of seroconversion occurred in dogs that were initially infected with A. platys and later challenged with E. canis. Following E. canis challenge infection, there was a dramatic increase in anti-Anaplasma antibodies; even for one dog in which A. platys serum antibodies were no longer detectable at the time of E. canis infection. In addition, there was no molecular evidence (PCR positivity) that A. platys organisms were present in the circulation of these dogs at the time this increase in Anaplasma serum antibodies occurred. The sensitivity and specificity of the ELISA for antibodies to Anaplasma and Ehrlichia species has also been shown to be high, reducing the likelihood of cross-reacting or false positive results . This finding suggests that infection with E. canis, and potentially other pathogenic organisms, can induce an immunogenic effect that results in an increased anamnestic response to previously recognized antigens, in addition to a specific humoral immune response to E. canis. This result is potentially consistent with previous findings which demonstrated that acute E. canis infections do not result in immunosuppression in the dog .
This study is the first to report the long-term serologic and PCR results for dogs experimentally infected with A. platys. Previous A. platys experimental studies reported on the acute phase of infection; with dogs being monitored for a maximum of 75 days PI independent of treatment [29, 30]. Immunological clearance of A. platys was supported in this study by the progression from PCR positive to PCR negative blood analyses by day 160 PI in all dogs infected only with A. platys. All dogs appeared to have cleared their infection prior to antibiotic treatment. These findings support prior clinical impressions that most A. platys strains in the United States are considered to cause minimal clinical disease, despite concurrent documentation of thrombocytopenia . However, isolates from other parts of the world are reported to induce a more severe disease in dogs [5, 30]. This study was limited to the strains of A. platys and E. canis available for establishing the experimental infection and the results of single, sequential and simultaneous infections may differ depending upon the strain encountered in nature. Likewise, all inoculums were prepared stored and administered in an identical manner however, undetermined variability in the infectious dose administered to each dog could have influenced the results. As dogs co-infected with A. platys and E. canis in this study developed a more persistent infection in conjunction with more severe thrombocytopenia and anemia, clinicians investigating natural infection due to A. platys should consider the potential influence of other known or unknown tick-borne pathogens.
In this study, similar to several previous studies utilizing experimental infections, doxycycline was found to be an efficacious therapy for E. canis infection when administered for four weeks [17, 38]. Those dogs treated at 99 (Group A→E) and 211 (Groups E, E→A, and A+E) days post-E. canis infection were E. canis PCR negative within 7 days of beginning doxycycline therapy. For those E. canis infected dogs that did not receive doxycycline, E. canis DNA could be found through the last time point tested by PCR (day 420 of the study) with blood and lymph node samples being more reliable sources for testing than bone marrow, similar to a previous studies of naturally infected dogs [7, 27].
Dexamethasone-induced immunosuppression resulted in a marked increase in the platelet counts for all dogs, which was more pronounced in chronically thrombocytopenic dogs infected with E. canis. Multiple mechanisms have been proposed for the thrombocytopenia associated with E. canis infections including increased platelet consumption, splenic sequestration and immune-mediated mechanisms associated with increased platelet destruction [1, 33, 39]. If immunosuppression was able to inhibit the immune-mediated destruction and the removal of platelets by tissue macrophages, the rapid rise and fall in platelet counts before and after corticosteroid administration may reflect an ongoing hyperplastic bone marrow response, which could potentially lead to hypoplastic bone marrow (exhaustion) in the chronic phase of canine ehrlichiosis [40, 41]. The use of immunosuppressive corticosteroids for treatment of in immune-mediated thrombocytopenia must be considered carefully when a dog is knowingly or unknowingly infected with a pathogen. Severe immunosuppression in dogs with chronic, undiagnosed infections could contribute to highly variable clinical outcomes, including death.