Determination of diagnostic cycle threshold (Ct) cut-offs for qPCR-based prevalence surveys of soil-transmitted helminth infections

WHO guidelines for control of soil-transmitted helminths (STH) rely on coproscopic methods to assess population prevalence. In low-prevalence and light-intensity STH settings, quantitative PCR (qPCR) has higher sensitivity and specificity for detection. For qPCR to accurately identify transmissible infections of public health significance, it is essential to interpret the qPCR cycle threshold (Ct) results. As part of the DeWorm3 community-based cluster randomized trial on interrupting STH transmission, we conducted population-based surveys using high-throughput qPCR and aimed to establish appropriate Ct cut-offs to detect transmissible infections. Experimental approaches including egg and genome-equivalent spiking experiments were hindered by inefficient fecal DNA extraction despite optimization efforts. The Ct results for 29,980 samples (pre-intervention, cross-sectional surveys) revealed a bimodal distribution for two of the four species tested, N. americanus and A. lumbricoides. The first peak was assumed to represent transmissible infections, and the second peak to represent indeterminate or non-transmissible infections. Using a finite mixture model, we defined true qPCR positivity as any Ct result with a [≥]5% chance of belonging to the first peak. This approach yielded Ct cut-offs of 34.4398 for N. americanus and 28.57587 for A. lumbricoides. For hookworms, sensitivity of qPCR was 96.7%, compared to 73.2% for Kato-Katz and moderate- to heavy-intensity infections (median Ct: 19.1, interquartile range [IQR]: 17.9-19.8) were differentiated from light-intensity infections and Kato-Katz negative samples (25.3, IQR: 22.5-27.9). Our findings demonstrate the feasibility and utility of evidence-based Ct cut-offs to identify transmissible STH infections in large scale surveys, and to categorize infection intensity as programmatically relevant. Author SummaryQuantitative PCR (qPCR) has been used for the detection of soil-transmitted helminths but with limited emphasis on determining cycle threshold (Ct) cut-offs to accurately identify transmissible infections which are of public health significance. When qPCR is used to assess interventions, or, in the future, to potentially make programmatic decisions, it will be crucial to validate positivity criteria to avoid underestimation (false negatives) or overestimation (false positives) of results. As part of the DeWorm3 trial, a community-based cluster randomized trial on interrupting transmission of STH, we developed and applied a validated STH qPCR to test 29,980 samples collected pre-intervention and explored experimental approaches to establish assay-specific Ct cut-offs. The Ct values from these samples showed a bimodal distribution for N. americanus and A. lumbricoides, suggesting the presence of two distinct groups. For this reason, a statistical approach with a finite mixture model was employed to determine Ct cut-offs that differentiated epidemiologically relevant, transmissible, egg-positive STH infections from those that are likely to represent detection of non-transmissible DNA or indeterminate results. While our data were applied in three different country settings, India, Benin and Malawi, it is important to note that no single Ct cut-off may be applicable across all epidemiological scenarios. Our findings demonstrate the feasibility of developing evidence-based Ct cut-offs with high sensitivity to accurately detect transmissible STH infections in large scale surveys.