Real-time PCR for detection of Strongyloides stercoralis in human stool samples from Côte d'Ivoire: diagnostic accuracy, inter-laboratory comparison and patterns of hookworm co-infection.

Human infections with the helminth species Strongyloides stercoralis encompass a wide clinical spectrum, ranging from asymptomatic carriage to life-threatening disease. The diagnosis of S. stercoralis is cumbersome and the sensitivity of conventional stool microscopy is low. New molecular tools have been developed to increase sensitivity. We compared the diagnostic accuracy of real-time PCR with microscopy for the detection of S. stercoralis and hookworm in human stool samples, and investigated the inter-laboratory agreement of S. stercoralis-specific real-time PCR in two European laboratories. Stool specimens from 256 randomly selected individuals in rural Côte d'Ivoire were examined using three microscopic techniques (i.e. Kato-Katz, Koga agar plate (KAP) and Baermann (BM)). Additionally, ethanol-fixed stool aliquots were subjected to molecular diagnosis. The prevalence of S. stercoralis and hookworm infection was 21.9% and 52.0%, respectively, whilst co-infections were detected in 35 (13.7%) participants. The diagnostic agreement between real-time PCR and microscopy was excellent when both KAP and BM tested positive for S. stercoralis, but was considerably lower when only one microscopic technique was positive. The sensitivity of KAP, BM and real-time PCR for detection of S. stercoralis as compared to a combination of all diagnostic techniques was 21.4%, 37.5% and 76.8%, respectively. The inter-laboratory agreement of S. stercoralis-specific PCR was substantial (κ=0.63, p<0.001). We conclude that a combination of real-time PCR and stool microscopy shows high accuracy for S. stercoralis diagnosis. Besides high sensitivity, PCR may also enhance specificity by reducing microscopic misdiagnosis of morphologically similar helminth larvae (i.e. hookworm and S. stercoralis) in settings where both helminth species co-exist.