Longitudinal utDNA dynamics predict recurrence risk in upper tract urothelial carcinoma: A prospective cohort study.

BACKGROUND AND

OBJECTIVE

Current surveillance of bladder recurrence for upper tract urothelial carcinoma (UTUC) is limited by invasiveness and sensitivity. There is an urgent need for noninvasive detection tool. We previously developed utLIFE, a urine tumor DNA model combining mutation profiling and fragmentomics. This study validates its utility for surveillance prediction.

METHODS

This prospective cohort included 47 UTUC patients undergoing radical nephroureterectomy (RNU). Serial urine samples were collected: pre-RNU (n = 47), post-RNU (n = 47), and at 3 to 6 months after RNU (n = 32). Kaplan-Meier curve (KM) analysis was used to compare recurrence-free survival (RFS), which was radiologically diagnosed with bladder cancer via CT.

RESULTS

The sensitivity of utLIFE at 3 to 6 months for predicting RFS were 83.3%, surpassing corresponding urine cytology and cystoscopy (16.7%). Notably, utLIFE positivity at 3 to 6 months demonstrated strong predictive value for RFS (P = 0.001). Dynamic utLIFE between post-RNU and 3 to 6 months, proved particularly informative, with both conversion from negative to positive status (P = 0.044) and persistent positivity (P = 0.005) serving as significant predictors of RFS. In patients receiving adjuvant therapy, utLIFE measurements effectively tracked treatment response. utLIFE preceded radiographic recurrence detection by a median of 5.67 months (P = 0.008), highlighting its potential for early identification.

CONCLUSIONS

This study establishes utLIFE as a noninvasive and promising monitoring tool for early recurrence detection.