Ehrlichiosis: New Developments

Edward Breitschwerdt United States

Rickettsial organisms in general, and Ehrlichia species in particular, are difficult to isolate from animals or human patients using conventional tissue culture systems.(1) Therefore, substantiation of a clinical diagnosis of ehrlichial infection has been limited to visualization of intracellular rickettsiae (individual organisms or morulae) on stained blood smears, or to the use of serologic testing for E. canis antibodies. Obviously, both of these modalities can have substantial limitations for the diagnosis of a chronic infectious disease. Until the recent identification of monocytic (2) and neutrophilic (3,4) human ehrlichiosis in Japan, Sweden, and the United States, Ehrlichia species were considered solely animal pathogens, for which limited interest or funding were available for research studies. These and other historical factors resulted in substantial limitations in our understanding of infections caused by members of the genus Ehrlichia. In dogs, ehrlichial infection is caused by Ehrlichia canis, Ehrlichia chaffeensis, Ehrlichia ewingii, Ehrlichia equi, Ehrlichia platys, and an uncharacterized Ehrlichia species, genetically similar to Ehrlichia risticii.

Of comparative medical interest, cats, dogs, humans, and other domestic and wild animal species can all be infected with the same Ehrlichia sp. For example, E. chaffeensis has been shown to infect dogs, goats, deer, and humans. In general, our knowledge of tick-borne diseases in cats is substantially less than our knowledge of the comparable disease in canine or human patients. Recent molecular evidence indicates that cats can be infected with E. phagocytophilia and E. canis. The infrequent diagnosis of ehrlichiosis in cats may be related to a number of factors including a general under-recognition of tick-borne diseases in cats, decreased pathogenicity of tick-borne pathogens in cats as compared to other species, or the more rapid removal of ticks from cats resulting in decreased opportunity for disease transmission. Most tick-transmitted pathogens require a 24 to 48 hour period of attachment to the host before there can be successful transmission of infectious organisms. Fastidious grooming may result in the early removal of most ticks.

The advent of applied molecular biologic techniques to the study of vector-borne disease problems continues to substantially clarify the role of established agents in the pathogenesis of previously undocumented disease sequelae. In many respects, the immunopathogenic consequences of tick-borne diseases such as ehrlichiosis are similar among infected animal and human patients. The experimental characterization of the immunopathologic response of animals to various tick-transmitted organisms has provided important insights as to the pathogenic consequences that would be expected in humans. Conversely, observations in human patients have contributed to the recognition of an increased spectrum of disease manifestations in animals.

Based on current understanding, E. canis is the most important cause of ehrlichial infection in dogs worldwide. However, future studies may confirm that granulocytic ehrlichiosis, caused by members of the E. phagocytophilia genogroup (E. equi, E. phagocytophilia, [HGE] Human Granulocytic Ehrlichiae) are capable of infecting cats, dogs, horses, cattle, and humans in endemic regions of the Northern Hemisphere. When we review the spectrum of disease manifestations attributed to E. canis infection, there is considerable variation in the type, duration, and severity of historical, physical, and clinicopathologic abnormalities reported in naturally infected dogs. Whether the substantial variation in disease manifestations or clinicopathologic abnormalities reflect strain differences in pathogenicity, variability in the immunologic response of the host to the rickettsia, co-infection with other tick-transmitted pathogens, or other unknown factors remains unclear. Recent evidence suggests that each of these factors may play a role in selected E. canis cases.

The �typical� acute clinical presentation of E. canis infection is associated with nonspecific findings, including anorexia, fever, and lymphadenopathy. Chronically infected dogs may appear healthy or may develop chronic weight loss, peripheral edema, or bleeding abnormalities, including prolonged bleeding from venipuncture sites, petechiae, retinal hemorrhage, or epistaxis. Factors responsible for the induction of disease signs in chronically infected dogs remain uncharacterized. Since both experimentally and naturally-infected dogs can remain healthy for several years following exposure, documentation of seroreactivity to E. canis antigen in a sick dog does not confirm that the dog�s illness is related to E. canis infection. For example, an older dog residing in an E. canis endemic region might present with fever, lymphadenopathy, thrombocytopenia, and hyperglobulinemia. Despite chronic infection and seroreactivity to E. canis antigen, all of the dog�s disease manifestations could be attributed to lymphosarcoma, or other neoplastic or infectious diseases. In this instance, lack of response to anti-rickettsial drugs would be expected due to concurrent neoplastic disease. An additional important problem confronting clinicians relates to valid documentation of therapeutic elimination of E. canis infection. There is scanty information in the veterinary literature regarding long-term post-treatment follow-up of dogs with ehrlichiosis. More recent studies, including those from our laboratory, indicate that many naturally and experimentally infected dogs clear E. canis infection following two weeks of doxycycline (5 mg/kg PO q12h) therapy. Despite rapid clinical improvement, some dogs remain hematologically abnormal for several months following a two-week course of doxycycline. It remains unclear as to whether persistence of high post-treatment antibody titers and/or the persistence of PCR amplicons in some dogs correlates with persistence of ehrlichial infection. A recent experimental study documented treatment failure in three of five dogs treated with 10 mg/kg of doxycycline for seven days.(5) Generally, a duration of therapy longer than seven days is recommended; therefore, the relevance of this experimental observation to treatment outcome in natural infection awaits additional study. In selected cases of �refractory� ehrlichiosis, we have been unable to detect E. canis DNA by polymerase chain reaction analysis, isolate rickettsiae in tissue culture, or transmit infection to naive dogs through blood transfusion. These findings suggest that persistence of E. canis seroreactivity in conjunction with clinical or hematologic abnormalities following anti-rickettsial drug therapy may not always be associated with persistence of the organism. We have also isolated other potential pathogens, particularly Bartonella vinsonii, from �refractory� cases that may have contributed to the persistence of disease manifestations.

In 1994, Kakoma and colleagues reported serologic evidence of over 100 cases of atypical canine ehrlichiosis with three fatalities.(7) These dogs were not seroreactive to E. canis antigen by immunofluorescence antibody (IFA), but were reactive to E. risticii antigen at generally low antibody titers. Ehrlichia risticii is the cause of Potomac horse fever. Serum samples were derived from California, Texas, Arizona, Illinois, Washington, Florida, and Michigan. Isolates obtained from three dogs were morphologically and genetically (based on partial 16S rRNA gene sequence) indistinguishable from E. risticii. Previous experimental infection of dogs and cats with E. risticii failed to induce clinical abnormalities, and successful experimental transmission of a dog isolate was not reported. Clinical abnormalities were described for six dogs with atypical ehrlichiosis, including lethargy, vague signs of abdominal discomfort with intermittent vomiting, persistent bleeding or petechial hemorrhages, polyarthritis, dependent edema, and posterior paralysis. Hematologic abnormalities were variably present and included: anemia, thrombocytopenia, prolongation of prothrombin time, and hypercalcemia. Therapy with tetracycline also appears to be variably efficacious for treatment of this organism, with substantial improvement in some dogs and a complete lack of response in other dogs. The author proposes that pending additional research, the canine isolate be referred to as E. risticii subsp. atypicalis. The definitive pathophysiologic characterization of this rickettsiae as a cause of disease in dogs awaits additional study.

Base upon isolation from patients, E. chaffeensis causes monocytic ehrlichiosis and E. phagocytophilia and E. ewingii cause granulocytic ehrlichiosis in people. However, the zoonotic role of dogs as a reservoir for human infection has not been clearly established for any Ehrlichia species. Dogs are the major reservoir for E. canis; however, it is probable that rodents and other small mammals serve as the major reservoir for most Ehrlichia species, with dogs playing only a minor role in the maintenance of the organism in a given geographic location. Recently, the detection of E. chaffeensis DNA by PXE amplification provided the first documentation for natural infection of dogs residing in animal shelters or in a kennel in southeastern Virginia. Subsequently, we documented E. chaffeensis infection in dogs that was clinically and serologically indistinguishable from E. canis or E. ewingii infection. Treatment with doxycycline resulted in therapeutic elimination of E. canis; however, based upon species-specific PCR amplification, E. chaffeensis DNA could be detected in all three dogs for up to one year following treatment, potentially due to frequent re-exposure to E. chaffeensis-infected Amblyomma americanum ticks. The clinical or zoonotic implications of this observation await additional clarification.

Recently, human infection has been associated with a granulocytic Ehrlichia species in United States and Europe. There have been several human fatalities, particularly in the elderly. Experimentally, isolates obtained from human patients will infect dogs, cats, sheep, and horses that appear to be naturally infected with the same organism in regions where human disease has been reported. Genetically, analysis of eubacterial 16S ribosomal DNA derived from human isolates is indistinguishable from Ehrlichia equi or Ehrlichia phagocytophilia. In 1982, Madwell and Gribble transmitted granulocytic ehrlichiae morulae via blood transfusion from a naturally infected dog to two horses, and from these horses to both dogs and cats. In Sweden, granulocytic Ehrlichia isolates derived from cats, dogs, and horses have 100% DNA sequence homology with canine, human, tick, and selected equine isolates from the United States. Granulocytic human ehrlichiosis has been reported in Minnesota, Wisconsin, and Massachusetts, but most probably has a much broader distribution within the United States. Experimentally, Ixodes scapularis ticks are competent vectors for transmission of granulocytic ehrlichiae, and a 10% prevalence of infection occurs in adult I. scapularis from Wisconsin and Nantucket, Massachusetts. Tick transmission to a broad host range, including cats, dogs, horses, sheep, goats, rodents, non-human primates, and humans is probable. In naturally-infected dogs, fever, lethargy, thrombocytopenia, elevated serum alkaline phosphatase and amylase activities, and hypoalbuminemia are the more frequently identified abnormalities.(7) Diagnosis requires visualization of granulocytic morulae or seroreactivity to E. equi antigen. Treatment with tetracycline elicits a rapid clinical response. Current data indicates that this granulocytic rickettsia is the cause of an important emerging zoonotic disease.

Several recent reports emphasize the potential for simultaneous infection with multiple tick-transmitted pathogens. For example, I. scapularis ticks can induce human infection with B. burgdorferi, E. equi (HGE), and Babesia microti. In dogs, we have observed simultaneous infection with E. canis or E. chaffeensis, Bartonella vinsonii, Babesia canis, and Rickettsia rickettsii. Infection with both E. ewingii and E. canis has been reported in naturally infected dogs from Oklahoma. When considered in conjunction with the isolation of an organism genetically identical to E. canis from a veterinarian from Venezuela, the recent documentation of E. chaffeensis infection in dogs and E. phagocytophilia infection in multiple animal species, the species specificity of various Ehrlichia species should perhaps be re-examined.

Perhaps now more than any other time in history, we have the opportunity to contribute to knowledge about the role of tick-borne diseases on animal and human health. Historically, much emphasis has been directed at diseases in which obvious, and frequently severe, clinical manifestations are apparent. Although these diseases will continue to be of critical importance, the role of infectious organisms in the genera Babesia, Bartonella, or Ehrlichia, as a cause of chronic, insidious, debilitating disease, deserves increased consideration by the biomedical community. Evolving evidence indicates that ticks continue to pose a serious health hazard for companion animals and people.

REFERENCES are AVAILABLE ON REQUEST