Shedding new light on Cyclospora: how the use of ultraviolet fluorescence microscopy can improve diagnosis of cyclosporiasis.

Cyclosporiasis has historically been underdiagnosed due to a lack of clinical recognition and challenges in laboratory detection. Microscopic detection of Cyclospora requires either the modified Kinyoun's acid-fast or safranin stains, which are not part of the standard ova-and-parasite examination. Alternatively, the use of fluorescence microscopy for the detection of Cyclospora due to the organism's autofluorescent properties has been an available tool in the field of clinical parasitology. However, it is unknown how ultraviolet (UV) fluorescence performs compared to permanent stains in routine diagnostic laboratory settings. In this study, we present a direct head-to-head comparison of UV fluorescence against modified acid-fast (MAF) stains using 50 (35 Cyclospora positive, 5 Cystoisospora positive, and 10 Cystoisospora negative) blinded, concentrated stool specimens. Five technically competent medical laboratory technologists of varying years of experience in clinical parasitology at a large reference laboratory independently read each of the MAF slides and UV wet mounts. Overall, the performance of UV fluorescence was similar to MAF slides with a higher agreement of 88.8% (95% CI 86.3%-91.3%) compared to 85.2% (95% CI 84.2%-86.2%), respectively. A higher false-negative rate was associated with MAF slides (MAF: 11.6%, UV: 6.8%). However, UV fluorescence was found to have a higher proportion of false positives (UF: 3.2%, MAF: 2%) and misidentified coccidia. Discrepant results regardless of methods were more frequently observed with technologists who have less experience in parasite morphology. This study is the first to document evidence to support the increased sensitivity and utility of UV fluorescence to improve laboratory diagnosis of cyclosporiasis.

Importance: This study is important as there is a dearth of studies in our field of clinical parasitology that investigate and establish performance characteristics of classic and newer, non-molecular methods. While these studies may not seem as heavy hitting as some new technologies described in our related disciplines, for our field, such studies are long overdue and critically lacking.