Testing Multiple Tissues Improves Performance of Diagnostic Assays for Chronic Wasting Disease in Elk (Cervus...

Effective wildlife disease management and research rely on trustworthy disease diagnostics. For chronic wasting disease (CWD), immunohistochemistry (IHC) is recognized as the gold standard, and ELISA is a reliable high-throughput screening method validated for use on obex and retropharyngeal lymph node (RPLN) tissues. However, diagnostic performance of the assays when used on tissues from elk (Cervus canadensis) varies, particularly under varying CWD population prevalence. We used hierarchical Bayesian modeling to predict sensitivity and specificity of ELISA on obex and RPLN and IHC on obex, RPLN, and rectoanal mucosa-associated lymphoid tissue (RAMALT) from elk. We detected CWD prion in at least one tissue from 46/283 individuals tested postmortem between 2016 and 2020. Specificity of ELISA was nearly perfect, providing strong confidence in CWD-detected results. Predicted mean test sensitivity was greater for ELISA than for IHC and slightly greater for RPLN tissue than for obex. The most sensitive individual test was ELISA RPLN, whereas the most reliable CWD diagnosis was achieved by parallel testing of both the obex and RPLN by ELISA, with serial confirmation by IHC. Although RAMALT can be used for antemortem testing, its low sensitivity makes negative results unreliable for proving an animal is disease free. When positive and negative predictive values of these methods were evaluated, we found the reliability of diagnostic results varied widely within the observed range of disease prevalence, and poor reliability was exaggerated by low diagnostic performance. Test-negative results are more likely to be true early in disease invasion, whereas the risk of a false-negative result grows with increasing disease prevalence in the sampled population. Wildlife managers should consider both the objectives and the stage of disease invasion when interpreting diagnostic test results.