Performance of Urine Tumor DNA Detection in Patients with High-grade Non-muscle-invasive Bladder Cancer Undergoing Repeat Transurethral Resection.

Abstract

Background and objective: Patients with high-risk non-muscle invasive bladder cancer (NMIBC) experience high recurrence rates and potential understaging, supporting guideline recommendations for repeat transurethral resection (reTUR). However, identifying patients best suited for reTUR remains challenging. Here, we analyze urine tumor DNA (utDNA) for mutations associated with urothelial cancer in a cohort of patients undergoing reTUR.

Methods: Retrospective evaluation of urine samples from 43 patients with pathologically confirmed high-grade NMIBC who were undergoing reTUR. Urine was collected immediately prior to reTUR and evaluated using UroAmp. The primary objective of this study was to compare utDNA disease classification with pathological staging from reTUR. Prognostic (high/low-risk) and minimal residual disease (MRD) (positive/negative) classifications were evaluated.

Key findings and limitations: The stage distribution from reTUR was 32% T0 (no residual disease), 29% Ta ±CIS, 28% T1 ±CIS, and 11% T2. Prognostic risk classification had 97% sensitivity (95% confidence interval [CI] 86-100) and 89% negative predictive value (95% CI 59-99). MRD classification had 80% specificity (95% CI 56-94) and 88% positive predictive value (95% CI 72-97). When long-term recurrence outcomes were considered, positive predictive value increased to 92% (95% CI 78-98) for both classification algorithms. Among T0 samples, utDNA high-risk patients had recurrence-free survival of 36%, compared to 100% for utDNA low-risk patients (p = 0.006).

Conclusions and clinical implications: utDNA testing identifies residual disease prior to reTUR. utDNA can serve as an adjunct following initial transurethral resection, aiding in the assessment of resection completeness, determining the need and timing of repeat resection, and stratifying risk of future recurrence.