Evaluation of antemortem sampling to estimate chronic wasting disease prevalence in free-ranging mule deer

We conducted it field study to evaluate tonsillar biopsy immunohistochemistry (IHC) as it tool For diagnosing chronic wasting disease (CWD) in live, free-ranging mule deer (Odocoileus hemionus) arid estimating CWD prevalence. Initially, we evaluated arid refined techniques for collecting tonsillar biopsies front mule deer. Using a simple mouth gag arid a 6-mm biopsy, forceps, and taking the biopsy starting at the rostral rim of the tonsillar sinus, we obtained 155/161 (96%) samples that yielded greater than or equal to1 lymphoid follicle. To Compare antemortem arid postmortem survey techniques arid assure biopsy-based estimates would not substantially underestimate "true" prevalence, we examined tonsillar biopsies front 161 free-ranging mule deer from 2 populations where CWD is endemic. We then calculated prevalence (p(b)) and compared this to prevalence (p(h)) estimated from tonsil samples from 161 (feel, hill-vested or culled in spatial arid temporal proximity to our study areas; we considered the latter a close approximation of "true" prevalence. Biopsy-based prevalence estimates exceeded prevalence estimated by tonsillar IHC of samples from harvested or culled deer. Although 95% CIs for p(h) - p(b), included 0 for area-specific estimates, biopsy-based estimates were greater than or equal to3 times higher than harvest-based estimates ill both study areas. Moreover. when data front both study areas were combined, p(h) (= 0.025) was lower than p(b) (= 0.081) and the 95% CI for p(h) - p(b) (-0.104 to -0.007) did not include 0. Observed differences in prevalence most likely reflected spatial or temporal variation in populations (or subpopulations) of deer sampled. Tonsillar biopsy IHC, appears to be reliable for detecting CWD infections in live mule deer and estimating prevalence in affected populations, thereby representing it new tool with potential utility in CWD management, particularly it) areas where harvest-based sampling is infeasible.