Abstract

PURPOSE There is significant interest in identifying complete responders to neoadjuvant chemotherapy (NAC) before radical cystectomy (RC) to potentially avoid removal of a pathologically benign bladder. However, clinical restaging after NAC is highly inaccurate. The objective of this study was to develop a next-generation sequencing–based molecular assay using urine to enhance clinical staging of patients with bladder cancer. METHODS Urine samples from 20 and 44 patients with bladder cancer undergoing RC were prospectively collected for retrospective analysis for molecular correlate analysis from two clinical trials, respectively. The first cohort was used to benchmark the assay, and the second was used to determine the performance characteristics of the test as it correlates to responder status as measured by pathologic examination. RESULTS First, to benchmark the assay, known mutations identified in the tissue (M T ) of patients from the Accelerated Methotrexate, Vinblastine, Doxorubicin, Cisplatin trial (ClinicalTrials.gov identifier: NCT01611662 , n = 16) and a cohort from University of California-San Francisco (n = 4) were cross referenced against mutation profiles from urine (M U ). We then determined the correlation between M U persistence and residual disease in pre-RC urine samples from a second prospective clinical trial (The pT0 trial; ClinicalTrials.gov identifier: NCT02968732 ). Residual M U status correlated strongly with residual disease status (pT0 trial; n = 44; P = .0092) when M U from urine supernatant and urine pellet were assessed separately and analyzed in tandem. The sensitivity, specificity, PPV, and NPV were 91%, 50%, 86%, and 63% respectively, with an overall accuracy of 82% for this second cohort. CONCLUSION M U are representative of M T and thus can be used to enhance clinical staging of urothelial carcinoma. Urine biopsy may be used as a reliable tool that can be further developed to identify complete response to NAC in anticipation of safe RC avoidance.