Colorectal Cancer Metastatic dMMR Immunotherapy (COMMIT) study: A randomized phase III study of atezolizumab (atezo) monotherapy versus mFOLFOX6/bevacizumab/atezo (FFX/bev) in the first-line treatment of patients (pts) with deficient DNA mismatch repair (dMMR) or microsatellite instability-high (MSI-H) metastatic colorectal cancer (mCRC)— NRG-GI004/SWOG-S1610.

This three-arm prospective phase III open-label trial randomized first-line dMMR/MSI-H mCRC pts (1:1:1) to either: FFX/bev, or atezo monotherapy (840mg IV q2wks), or the combination of FFX/bev+atezo. Primary endpoint was progression-free survival (PFS) by intent-to-treat. Because of the KEYNOTE 177 results, COMMIT’s FFX/bev arm was closed (trial amended 6/4/20), leaving two arms: atezo monotherapy v FFX/bev+atezo with 80% power to detect a hazard ratio of 0.6 for PFS, one-sided alpha=0.025. The study was also modified to enroll 120 total pts. Secondary endpoints included safety, objective response rate (ORR), disease control rate (DCR), duration of response, and overall survival.

Pt accrual was suspended on 3/31/25 because of the results from the Checkmate 8HW trial, while a pre-planned interim analysis occurred near the same time. From 11/2017 to 3/2025, a total of 102 pts were enrolled (median age 62 yrs; 48% female gender; 23% BRAF V600E mutation positive): FFX/bev: n=20, atezo: n=41, and FFX/bev+atezo: n=41. The interim analysis had a median follow-up of 3.5 yrs, 45 PFS events, and 65% information for the primary endpoint of PFS, demonstrating the superiority of the FFX/bev+atezo over atezo monotherapy, with a hazard ratio of 0.439 (95% CI 0.23-0.84, p=0.0103), which was below the critical value of 0.0152 based on the O’Brien Flemming monitoring boundary. Median PFS was 30.0 and 4.3 months, ORR was 80.6% v 46%, and DCR at 12 months was 62.9% v 32.4% in the FFX/bev+atezo arm compared to the atezo-only arm, respectively. Grade 3 or higher adverse events (AEs) occurred in 52 pts (atezo: 18; combination arm: 34). There were six total Grade 5 AEs (two unrelated, two unlikely, and two possible), of which one was in the atezo arm (death NOS/progression) and five were in the combination arm (two NOS deaths, one disease progression, one hepatic hemorrhage, and one cardiac arrest). Additional endpoint analyses are ongoing and will be presented.