9530 Background: An ITC comparing NIVO + RELA and NIVO + IPI, approved dual immunotherapy treatment options for patients (pts) with advanced melanoma, was previously conducted using pt-level data from the pivotal RELATIVITY-047 (RELA-047; NIVO + RELA vs NIVO) and CheckMate 067 (CM-067; NIVO + IPI or NIVO vs IPI) trials (Long JCO 2024). Here we present updated results using 5-year follow-up data from RELA-047. Methods: Inverse probability of treatment weighting was used to adjust for cross-trial imbalances in baseline characteristics. CM-067 follow-up was truncated to best align with the follow-up length in RELA-047. Progression-free survival (PFS) per investigator (INV), confirmed objective response rates (ORRs) per INV, overall survival (OS), and melanoma-specific survival (MSS) were analyzed. Outcomes were also evaluated across key subgroups. The weighted NIVO arms from each trial were compared for validation. Results: After weighting, key baseline characteristics were balanced for NIVO + RELA (n = 339) and NIVO + IPI (n = 297). Outcomes after weighting were similar between NIVO + RELA and NIVO + IPI (hazard ratio HR 95% CI: PFS, 1.10 0.90–1.34; OS, 0.96 0.77–1.20; MSS, 0.88 0.68–1.12; table). Grade 3–4 treatment-related adverse event (TRAE) rates were numerically lower with NIVO + RELA than with NIVO + IPI (23% vs 62%, respectively). Across subgroups, efficacy appeared similar between treatments, although trends favoring NIVO + IPI were observed for ORRs among pts with BRAF mutant disease (59% vs 48%) or serum lactate dehydrogenase > 2× the upper limit of normal (33% vs 12%). Among pts progression-free at 3 years, both NIVO + RELA and NIVO + IPI were associated with sustained benefit, with 5-year OS of 96% and 94% and MSS of 100% and 98%, respectively. Conclusions: Consistent with earlier findings, this updated ITC with a 5-year minimum follow-up from RELA-047 suggests that 1L treatment with NIVO + RELA may have efficacy comparable to NIVO + IPI in pts with advanced melanoma, overall and across most, but not all, subgroups. Results should be interpreted with caution because of differences in study design and the evolving treatment landscape over time. Efficacy outcomes after weighting. NIVO + RELA (n = 339) NIVO + IPI (n = 297) HR/OR (95% CI) a NIVO RELA-047 (n = 338) NIVO CM-067 (n = 288) HR/OR (95% CI) b Median PFS per INV, mo (95% CI) 12.0 (8.2–17.1) 11.2 (8.5–18.1) 1.10 (0.90–1.34) 6.7 (4.6–10.2) 5.7 (3.9–9.1) 0.96 (0.79–1.16) Confirmed ORR per INV, % 48 50 0.92 (0.73–1.15) 40 40 0.99 (0.78–1.26) Median OS, mo (95% CI) 64.2 (38.6–NR) 61.4 (37.1–NR) 0.96 (0.77–1.20) 34.7 (27.3–47.3) 35.7 (26.4–52.7) 1.04 (0.85–1.28) Median MSS, mo (95% CI) NR (NR–NR) NR (61.4–NR) 0.88 (0.68–1.12) 51.2 (34.7–NR) 44.8 (32.3–NR) 1.00 (0.80–1.25) a NIVO + RELA vs NIVO + IPI. b NIVO arms. NR, not reached; odds ratio, OR.
Tawbi et al. (2026) studied this question.