Prognostic utility of ctDNA before and after trimodality therapy (TMT) for muscle invasive bladder cancer.

Abstract

4578 Background: Outside of standard patient and tumor characteristics, biomarkers predicting outcomes after TMT for non-metastatic muscle-invasive bladder cancer (MIBC) are lacking. Tumor-informed circulating tumor DNA (ctDNA) has identified MIBC patients at risk of relapse following radical cystectomy. This study assessed the clinical utility of a ctDNA assay in predicting disease progression following TMT. Methods: We retrospectively identified patients (pts) with MIBC (cT2-T4, cN0-N2, cM1) who had TMT between 2019 and 2024 and ctDNA assessment with a personalized, tumor-informed assay (Signatera, Natera, Inc) before and/or after TMT. Standard clinical variables extracted include TNM stage, risk factors at TURBT (carcinoma in situ (CIS) and hydronephrosis) age of diagnosis, primary radiotherapy (RT) dose, chemotherapy regimen/duration, and if elective nodal irradiation (ENI) was received. We analyzed the impact of pre-TMT or post-TMT ctDNA (MTM/mL), as well as changes in ctDNA, on disease-free survival (DFS) using the Kaplan-Meier method. We assessed the impact of ENI on DFS in cN0 pts with positive ctDNA at baseline. Predictors of recurrence were assessed with univariate Cox proportional hazard models. Results: Among 34 pts, median age was 69 (interquartile range (IQR) 62-79), majority were male (91%), had cT2 (73.5%), were cN0 (85%), had no CIS (76%) or hydronephrosis (94%). The median dose of RT was 64 Gy. Ten (29%) patients had detectable ctDNA prior to TMT. Median follow-up was 13.2 months from last RT dose of TMT. Pts with detectable ctDNA after TMT had inferior DFS compared to pts with persistently undetectable (pre/post-TMT) or pts with ctDNA clearance (median survival of 5 months vs 15 and 19 months, respectively, log-rank p = 0.056). Table 1 shows results for predictors of DFS after TMT. Pts with detectable ctDNA post-TMT (n = 4) all developed distant metastatic recurrence. Local recurrences did not present with a detectable ctDNA following TMT. Pts with measurable ctDNA at baseline demonstrated potentially prolonged DFS after ENI (n = 4) compared to the non-ENI cohort (n = 6), but did not meet statistical significance (7.6 months vs not reached, p = 0.2). Conclusions: Persistently detectable ctDNA following TMT correlates with disease-free survival. Larger cohorts are needed to assess role of ENI in pts with detectable ctDNA prior to TMT. These results are hypothesis-generating and should be validated prospectively. Predictors of DFS after TMT. HR (95% CI) P-value Histologic type (Reference (Ref) group: Pure urothelial) Urothelial with minor variant 0.76 (0.21-2.84) 0.69 CIS on TURBT (Ref group: No CIS) CIS 0.61 (0.16-2.28) 0.46 Pre-TMT Hydronephrosis (Ref group: No Hydronephrosis) Hydronephrosis 3.63 (0.77-17.19) 0.1 Response of ctDNA status to TMT (Ref group: Detectable to Detectable) Detectable to Undetectable 0.12 (0.02-1.00) 0.05 Undetectable to Undetectable 0.26 (0.04-1.75) 0.17