In North America, the only endemic focus for Angiostrongylus vasorum (French heartworm) was historically thought to occur in the southeastern part of the island of Newfoundland. However, reports of A. vasorum infection in wild canids in West Virginia, USA, and Nova Scotia, Canada, suggest the introduction of the parasite to mainland North America. We screened for A. vasorum in coyotes from across southern Ontario. Additionally, we evaluated the performance of ELISAs for detection of circulating A. vasorum antigen (Ag-ELISA) and antibodies against A. vasorum (Ab-ELISA) designed for use in sera or blood of foxes for use with coyotes in this region. Autopsies were performed on 397 coyotes, and lung tissue extract prepared from each carcass was tested via both ELISAs. The sensitivity and specificity for both tests were estimated in the absence of a gold standard using a 2-test single population Bayesian model; sensitivity and specificity priors were based on the performance of the assays in foxes in Switzerland. Eight coyotes tested positive for A. vasorum antigen; no animal was antibody positive. The estimated sensitivity and specificity of the Ag-ELISA were 90.8% (95% credible interval [CrI]: 83.8-95.6%) and 95.5% (95% CrI: 93.4-97.2%), respectively. For the Ab-ELISA, the estimated sensitivity and specificity were 41.9% (95% CrI: 32.1-51.9%) and 98.0% (95% CrI: 96.3-99.0%), respectively. Based on these findings and negative postmortem data for the same animals, there is insufficient evidence to suggest the presence of A. vasorum in southern Ontario coyotes.
Keywords: Angiostrongylus vasorum; Bayesian analysis; ELISA; coyotes; sensitivity; southern Ontario; specificity.