Detection of Tropheryma whipplei in stool samples by one commercial and two in-house real-time PCR assays

Objective Tropheryma whipplei, the causative agent of Whipple's disease, can also be identified in stool samples of humans without systemic disease. It is much more frequently detected in human stool samples in tropical environments than in industrialized countries. PCR-screening has been applied for point prevalence studies and environmental assessments in tropical settings, but results depend on the applied assay. We compared one commercial qPCR kit with two well-described in-house assays for detection of T. whipplei from stool. Methods Residual materials from nucleic acid extractions of stool samples from two groups with presumably different prevalences and increased likelihood of being colonized or infected by T. whipplei were tested. One group comprised 300 samples from study participants from western Africa (group 1); the second group was of 300 returnees from tropical deployments (group 2). Each sample was assessed with all three qPCR assays. Cycle threshold (C-t) values were descriptively compared. Results Based solely on mathematical modeling, the three PCR assays showed considerably different detection rates of T. whipplei DNA in stool samples (kappa 0.67 (95% confidence interval [0.60, 0.73])). Considering the calculated test characteristics, prevalence of 28.3% for group 1 and 5.0% for group 2 was estimated. Discordant test results were associated with later C-t values. The study did not validate the assays for the detection of T. whipplei in Whipple's disease and for diagnostic purposes since clinical specificity and sensitivity were not investigated. Conclusions In spite of the observed diagnostic uncertainty, PCR-based screening approaches can be used for epidemiological purposes and environmental samples to define the source and reservoir in resource-limited tropical settings if prevalence is calculated using diagnostic accuracy-adjusted methods.