Association between polyarthritis and thrombocytopenia and increased prevalence of vectorborne pathogens in Californian dogs

TICKBORNE diseases are regionally common problems in dogs, with almost undetectable to fatal clinical manifestations, including fever, weakness, icterus, epistaxis and lameness (May and others 1990, Greig and others 1996, Harrus and others 1997, Pappalardo and others 1997, Greene and Breitschwerdt 1998, Foley and others 2001). In California, important tickborne diseases include Rocky Mountain spotted fever (RMSF), caused by Rickettsia rickettsii, canine monocytic ehrlichiosis, caused by Ehrlichia canis, canine granulocytic anaplasmosis (GA), caused by Anaplasma phagocytophilum (formerly Ehrlichia equi), and borreliosis, caused by Borrelia burgdorferi sensu stricto. Bartonellosis, caused by Bartonella vinsonii berkhoffii, has an unconfirmed arthropod vector but is suspected to be transmitted by ticks (Pappalardo and others 1997, Breitschwerdt and others 1998, Kordick and others 1999b, Chang and others 2001, MacDonald and others 2004). Tickborne diseases are a significant risk in California, with the northern coast mountain range ranked by the Centers for Disease Control and Prevention (CDC) as a ‘moderate risk’ for Lyme disease in human beings (Brown and Lane 1992, CDC 2001). In particular, GA, transmitted by Ixodes pacificus ticks, is a very important threat to dogs, horses and human beings (Madigan and Gribble 1987, Gewirtz and others 1996, Foley and others 1999, 2001, 2004a). Signs of the disease can include pyrexia, lethargy, myalgia, nausea, arthralgia, ataxia, thrombocytopenia and neutropenia (Greig and others 1996, Foley and others 2001). E canis is transmitted primarily by Rhipicephalus sanguineus, a tick that occurs sporadically among dogs in California. Infected dogs may have thrombocytopenia, vomiting, oculonasal discharge, peripheral oedema, arthritis, glomerulonephritis, ataxia, dyspnoea and splenomegaly (Neer 1998). B vinsonii berkhoffii causes endocarditis, myocarditis, granulomatous lymphadenitis, immune-mediated polyarthritis, thrombocytopenia and granulomatous rhinitis in dogs (Pappalardo and others 2000, MacDonald and others 2004, Henn and others 2005). RMSF, transmitted in California by Dermacentor variabilis and Dermacentor andersoni, can induce thrombocytopenia, septic neutrophilic vasculitis, coagulation defects, fever, oedema in various sites, ocular and nasal discharge, and scleral injection, petechiae, epistaxis and melaena, lymphadenopathy, neurological abnormalities and limb necrosis (Greene and Breitschwerdt 1988). Possibly the two most important and common sequelae of tickborne infection in many dogs are polyarthritis and thrombocytopenia. However, both of these problems can also occur as a result of many other diseases. The mechanism for polyarthritis in tickborne disease may include immune-complex deposition, which can occur with ehrlichiosis, RMSF, ana plasmosis and borreliosis; borreliosis is also associated with primary infectious synovitis (Greene 1998). Thrombocytopenia may occur due to ineffective haematopoiesis, immune-mediated platelet destruction, infectious or immune-mediated vasculitis, and sequestration or margination of platelets in infection, with E canis, RMSF, bartonellosis and possibly anaplasmosis (Dumler and Bakken 1995, Wong and Thomas 1998, Harrus and others 1999). At the University of California Davis Veterinary Medical Teaching Hospital, the two most common indications for testing with a ‘tick panel’ are thrombocytopenia and polyarthritis. The aim of this study was to determine whether the prevalence of tickborne pathogens would be increased in dogs with polyarthritis and/or thrombocytopenia. A case-control design was used to evaluate this hypothesis. A secondary goal of the study was to describe regional and demographic risks for serological and PCR-positive results. A panel testing for all suspected canine tickborne diseases was administered to 110 dogs seen at the hospital with signs of thrombocytopenia, polyarthritis, or both, and to 110 control dogs randomly chosen from dogs seen at the hospital on the same day as the case dog. The location of the dog’s residence, based on the owner’s address, was grouped as follows: coast range mountains (a set of mountain ranges from 304 to 1700 m, with a climate varying from temperate rainforest to Mediterranean, extending from the coast to 50 km inland, from the Oregon border southwards to near the Los Angeles area), the hot dry Central Valley (extending north to south between the coast ranges and the Sierra Nevada mountains), the Sierra Nevada mountains (along the eastern state boundary), and out-ofstate regions (including two dogs from Nevada and one from Oregon). Differences in haematological values were assessed by a t test and location by a chi-squared test. P values for odds ratios were obtained by Fisher’s exact test. For all statistical tests P≤0·05. Cases Controls