Abstract
PURPOSE: Trimodality therapy (TMT) is a bladder-sparing treatment approach for patients with muscle-invasive bladder cancer (MIBC). Although plasma circulating tumor DNA (ctDNA) is correlated with treatment response and outcomes following cystectomy for MIBC, the association of plasma ctDNA with clinical outcomes in patients treated with TMT is poorly characterized. EXPERIMENTAL DESIGN: Patients with MIBC who received TMT at Dana-Farber/Brigham and Women's Cancer Center or Massachusetts General Hospital were consented to a research protocol and underwent ctDNA evaluation at baseline and at regular intervals following TMT using the commercially available Signatera assay. All patients also underwent routine post-TMT surveillance including cross-sectional imaging and cystoscopic evaluation. The association between ctDNA status and clinical disease status was assessed. RESULTS: Eighty-four patients had at least one ctDNA result. Fifty-nine patients had a ctDNA test performed prior to chemoradiotherapy (CRT), and 19 (32%) had detectable ctDNA. Forty-six patients had paired pre- and post-CRT ctDNA results available: nearly all (29/31, 94%) patients with undetectable ctDNA prior to CRT also had undetectable ctDNA at the first post-CRT assessment, whereas 11 of 15 (73%) patients with detectable ctDNA prior to CRT converted to undetectable ctDNA following CRT. Sixteen (19%) patients developed metastatic disease, and detectable ctDNA following CRT was strongly correlated with metastatic recurrence (P < 0.0001). Conversely, no patients with an isolated muscle-invasive (n = 5) or non-muscle invasive (n = 7) bladder recurrence had a post-TMT test demonstrating detectable ctDNA prior to recurrence. CONCLUSIONS: ctDNA detectability is strongly correlated with metastatic but not local disease recurrence in patients with MIBC following TMT.