Nivolumab + ipilimumab (NIVO + IPI) vs NIVO in microsatellite instability high/mismatch repair-deficient metastatic colorectal cancer (MSI-H/dMMR mCRC): CheckMate 8HW subgroup analyses including treatment-free status and outcomes in long-term survivors.

3532 Background: In the phase 3 CheckMate 8HW trial, both dual primary endpoints of progression-free survival (PFS) were met for first-line (1L) NIVO + IPI vs chemotherapy (HR 0.21; P < 0.0001) and NIVO + IPI vs NIVO in all lines (HR 0.62; P = 0.0003) in patients (pts) with centrally confirmed (CC) MSI-H/dMMR mCRC. We report explorative subgroup analyses of NIVO + IPI vs NIVO. Methods: The study design was described previously (Andre NEJM 2024). Pts with MSI-H/dMMR mCRC per local testing were enrolled. After randomization, IHC and PCR based tests were used for central confirmation. Results: In all randomized pts, 84% (296/354) in the NIVO + IPI and 81% (286/353) in the NIVO arm had CC MSI-H/dMMR; in 1L pts, 85% (171/202) and 85% (170/201) had CC MSI-H/dMMR, respectively. Median follow-up in all randomized pts was 55.1 (range 24.7–68.5) mo. Pts receiving NIVO + IPI had improved disease control rates vs NIVO in 1L (85% vs 81%, respectively) and all lines (84% vs 76%, respectively). Pts with complete or partial response (CR/PR) or stable disease (SD) on 1L NIVO + IPI had improved PFS vs NIVO. In all lines, pts with CR/PR or SD receiving NIVO + IPI had improved overall survival (OS) vs those on NIVO (Table). Pts receiving NIVO + IPI had improved disease-specific OS vs NIVO (HR 0.47; 95% CI 0.32–0.68). In all lines, pts on NIVO + IPI had improved PFS and OS vs NIVO regardless of the number of sites of metastasis (Table). After treatment, 82% (185/226) of pts on NIVO + IPI were alive and treatment free vs 67% (125/186) on NIVO; 90% (198/220) and 79% (141/178) of pts were alive and systemic treatment free, respectively. In pts surviving ≥4 years, baseline characteristics and incidence of treatment-related adverse events were consistent with the all randomized CC population. Further analyses will be presented. Conclusions: NIVO + IPI demonstrated clinical benefit vs NIVO across subgroups, as shown by numerically increased PFS and OS in pts with CR/PR or SD, and improved PFS and OS regardless of number of metastatic sites. No defining characteristic was identified in long-term survivors vs all randomized CC pts. These results further support NIVO + IPI as a standard of care treatment for MSI-H/dMMR mCRC. Clinical trial information: NCT04008030 . 1L NIVO + IPI(n = 171) NIVO(n = 170) HR (95% CI) mPFS (95% CI), mo CR/PR NR (NE) NR (60.8–NE) 0.82 (0.43–1.54) SD 33.1 (5.8–NE) 9.2 (4.4–22.1) 0.52 (0.24–1.14) All lines (n = 296) (n = 286) mPFS (95% CI), mo 1 metastatic site at BL NR (NE) 60.8 (26.3–NE) 0.51 (0.33–0.77) ≥2 metastatic site at BL 59.2 (38.4–NE) 23.0 (11.8–NE) 0.70 (0.51–0.97) mOS (95% CI), mo CR/PR NR (NE) NR (NE) 0.92 (0.46–1.83) SD NR (35.3–NE) 26.8 (15.7–48.9) 0.43 (0.23–0.80) 1 metastatic site at BL NR (NE) NR (NE) 0.52 (0.31–0.88) ≥2 metastatic site at BL NR (NE) NR (NE) 0.68 (0.46–0.99) BL, baseline; m, median; NE, not estimable; NR, not reached.