UriFind assay and association with whole course management of urothelial carcinoma in a large multicentric, prospective China cohort.

Abstract

e16588 Background: The gold standard to diagnose and monitor urothelial carcinoma is cystoscopy, which is invasive and painful. The currently available non-invasive methods to detect urothelial carcinoma are either of unsatisfying sensitivity or limited specificity in China. Previously, a DNA methylation assay, termed UriFind, was developed, and showed a compelling sensitivity and specificity in a retrospective cohort. This study aimed to further validate its performance in a large multicentric, prospective China cohort. Methods: Patients involved in the whole course of UC management were recruited in this cohort. The cohort contained 931 patients diagnosed with urothelial carcinoma (UC), and 994 controls with benign diseases and 78 healthy participants from 12 hospitals across China. Urine samples were collected before the therapy, and analyzed with quantitative PCR of ONECUT2 and VIM methylation (UriFind), FISH and urinary cytology. Sensitivity, specificity, accuracy, PPV and NPV of these assays were evaluated as compared to the diagnostic gold standard of cystoscopy, histology results or clinical diagnosis. Results: The UC group of the cohort consisted of 683 non-muscle invasive UC (73.4%), 203 muscle invasive UC (21.8%) and 45 unavailable stage (4.8%). The patients of low, high-grade UC and unavailable grade were 244(26.2%), 651(69.9%) and 36 (3.9%), respectively. Overall sensitivity, specificity, accuracy, PPV and NPV of UriFind was 82.3%, 91.2%, 87.3%, 88.1% and 86.7%, respectively. The specificity of healthy participants was 100%. The overall sensitivity and accuracy of UriFind were significantly higher than urine cytology (40% and 55.7%) and FISH (68.2% and 73.6%), while the specificity was similar to cytology (93.9%) and FISH (87.3%). Importantly, UriFind achieved a great improvement of sensitivity in whole course management, including low grade tumor (56.1%), small and single tumor (63.3%), residual tumor after therapy (71.4%) and recurrent tumor (84.3%). Conclusions: The performance of UriFind is well validated with significantly higher sensitivity and accuracy than urine cytology and FISH in UC detection in this cohort with 2003 cases. This assay facilitates whole course management of urothelial carcinoma, which may reduce the burden of cystoscopy and unnecessary second TURBT. Clinical trial information: NCT04314245 .