Prognosis in canine idiopathic immune-mediated haemolytic anaemia

Abstract

Canine idiopathic immune-mediated haemolytic anaemia (iIMHA) is one of the most frequently occurring immune-mediated diseases in dogs. A gel-based Coombs' test was shown to perform equally well as a classical Coombs' test. Since the gel-based Coombs' test can be commercially produced and is easy and fast to perform, this offers the opportunity to standardize anti-erythrocyte antibody testing. Two retrospective cohorts of dogs with iIMHA were studied; one cohort was treated with prednisolone and azathioprine (n=149), and the other cohort (n=73) was treated with prednisolone only. Dogs of both cohorts presented with severe anaemia, leukocytosis, and a third of the dogs had increased coagulation times and thrombocytopenia. There was no significant difference in survival between the two protocol groups; the one year survival times were 64% (95% CI 54-77%) for the group treated with prednisolone, and 69% for the group treated with prednisolone and azathioprine, respectively. It can be concluded that, contrary to current belief, the addition of azathioprine to a protocol of prednisolone does not decrease mortality. Azathioprine causes bone marrow suppression, however, in 8% of dogs. Contrary to current recommendations for lifelong immunosuppresion, a duration of immunosuppression for 3 months was sufficient to maintain remission of the haemolytic anaemia in both protocol groups. Almost all mortality occurred in the first 2 weeks after diagnosis. Multivariate Cox's proportional hazards analysis showed that main predictors of this mortality were the presence of icterus, increases in plasma urea and creatinine, a left shift, increases in coagulation times, and thrombocytopenia. Since leucocytes have been shown to express tissue factor, it was hypothesized that leucocytes taking part in the inflammatory response in dogs with iIMHA had increased expression of interleukin-8 and tissue factor. Nine commonly used reference genes were examined for their suitability as reference genes in canine whole blood. It was shown that the leukocyte count affected the expression of these reference genes. This emphasized that commonly used reference genes may not be stable under all experimental conditions and the assumption of stable expression under all conditions should be checked for each new experiment. The basal interleukin-8 and tissue factor expression were determined in healthy dogs. These expressions were compared with the expression of dogs with sepsis, dogs with disseminated intravascular coagulation, and dogs with iIMHA. Dogs with systemic illness and dogs that had surgery but did not have sepsis or DIC, were included as controls. Since patient care related issues such as intravenous catheters may change tissue factor and interleukin-8 expressions as well. The dogs with iIHA had a pronounced leukocytosis with left shift and signs of hypercoagulability. The coagulation factor activities were low and platelet parameters suggested platelet activation and high platelet turnover. Tissue factor expression was high in the dogs with iIMHA, but the interleukin-8 expression was low. Since both tissue factor and interleukin-8 result from activation of the NF-kB signaling pathway this suggests that the tissue factor expression is not increased in leucocytes, but possibly in other blood cells. A possible role for thrombocytes should be investigated.