Adjuvant nivolumab versus placebo for high-risk muscle-invasive urothelial carcinoma: 5-year efficacy and ctDNA results from CheckMate 274.

BACKGROUND: Despite surgery, recurrence rates remain high in muscle-invasive urothelial carcinoma (MIUC). The phase III randomized, double-blind, multicenter CheckMate 274 trial comparing adjuvant nivolumab versus placebo in patients with high-risk MIUC after radical surgery demonstrated that adjuvant nivolumab improved disease-free survival (DFS) versus placebo. We report results with 5-year minimum follow-up including exploratory analyses of circulating tumor DNA (ctDNA). PATIENTS AND METHODS: Eligible patients had radical surgery (R0, with negative surgical margins) within 120 days before being randomized, with or without neoadjuvant cisplatin-based chemotherapy. Overall, 709 patients with MIUC at high risk of recurrence after surgery were randomly assigned 1:1 to nivolumab 240 mg or placebo every 2 weeks for ≤1 year. The primary endpoint was DFS. Overall survival (OS) was a key secondary endpoint. In a post hoc exploratory analysis, baseline ctDNA was measured using the Signatera assay. RESULTS: Median DFS was 21.9 months [95% confidence interval (CI) 18.8-36.9 months] with nivolumab versus 11.0 months (95% CI 8.3-16.6 months) with placebo in all randomized patients [hazard ratio (HR) 0.74, 95% CI 0.61-0.90], and 55.5 months (95% CI 25.8-66.5 months) versus 8.4 months (95% CI 5.6-20.0 months) in patients with tumor programmed death-ligand 1 expression ≥1% (HR 0.58, 95% CI 0.42-0.79). In all randomized patients, median OS was 75.0 months (95% CI 56.7 months-not estimable) with nivolumab versus 50.1 months (95% CI 38.0-72.1 months) with placebo (HR 0.83, 95% CI 0.67-1.02). Baseline ctDNA after radical surgery was detected in 54 (40.6%) of 133 assessable patients. Median DFS was 52.1 months (95% CI 19.4 months-not estimable) versus 5.0 months (95% CI 2.8-6.5 months) in patients with undetectable versus detectable ctDNA at baseline (HR 0.30, 95% CI 0.18-0.48). In patients included in the ctDNA analysis who had detectable ctDNA, the median DFS HR for nivolumab-treated patients versus placebo-treated patients was 0.35 (95% CI 0.18-0.66), while in patients with undetectable ctDNA, the HR for nivolumab-treated patients versus placebo-treated patients was 0.99 (95% CI 0.51-1.93). There were no new safety signals. CONCLUSIONS: Five-year results confirm durable DFS benefit and consistent safety profile for adjuvant nivolumab in high-risk MIUC. Post hoc exploratory ctDNA assessment identified patients at highest risk of recurrence and may inform tailored adjuvant therapy strategies, pending further validation.