Serial circulating tumour DNA monitoring in first-line advanced NSCLC (COPE): a randomized, two-arm non-comparative phase II trial
Candice Francheska B. Tambaoan, Letuan Phan, Sophie Cousin, Sylvestre Le Moulec, Thomas Grellety, Sophie Schneider, Marielle Sabatini, Laura Leroy, Mathilde Cabart, Yechan Laizet, Isabelle Soubeyran, Melissa Alame, Laura Blouin, Fabienne Gerson, Emmanuel Khalifa, Russell Madison, Giuseppe Di Caro, Amaya Gasco, Merrida Childress, Noemie Huchet, Marina Pulido, Amandine Crombé, Antoine ItalianoAbstract
Purpose: COPE evaluated serial circulating tumor DNA (ctDNA) profiling with molecular tumor board (MTB) review during first-line therapy for advanced non–small cell lung cancer (NSCLC). Experimental Design: COPE was a multicenter, randomized (2:1), noncomparative phase II trial. Arm A underwent serial plasma profiling with every report reviewed at a weekly virtual MTB; Arm B received standard management. The primary endpoint was 18-month overall survival (OS) in Arm A versus a prespecified 40% historical benchmark; no between-arm comparison was planned. Results: Among 176 patients, 117 were assigned to Arm A and 59 to Arm B. Eighteen-month OS was 53.8% (95% CI, 44.3–62.5) and 65.7% (95% CI, 52.0–76.3), respectively. Baseline plasma profiling provided contributory information in 32 of 35 patients with noncontributory or unavailable tissue, increasing genotyping success from 69% to 97%. Paired baseline and early on-treatment samples were available in 90 patients (median, 21 days; range, 19–64). ctDNA clearance was associated with higher objective response rate (83% vs. 52%), longer progression-free survival (22.9 vs. 5.4 months; P<0.001), and longer OS (not reached vs. 17.1 months; P=0.004) than persistent detection. Conclusions: Serial ctDNA profiling with weekly MTB review was operationally feasible and complemented tissue genotyping. Clinical benefit was not demonstrated; early ctDNA kinetics require validation in interventional trials.