Randomized dose evaluation of nivolumab + relatlimab (NIVO + RELA) in patients (pts) with advanced melanoma: Results from RELATIVITY-020.

9526 Background: NIVO + RELA is approved at 480 mg NIVO + 160 mg RELA Q4W (480/160) dosing for treatment of advanced melanoma based on RELATIVITY-047. A higher dose of 480 mg NIVO + 480 mg RELA (480/480) is under investigation in other tumor types. Here, we report results from RELATIVITY-020 Part E, which aimed to compare efficacy and safety of first-line (1L) 480/480 vs 480/160 dosing in pts with treatment-naive advanced melanoma and evaluate 480/480 dosing in pts with advanced melanoma who progressed on prior anti–PD-1 therapy (PD-1 refractory [RF]). Methods: RELATIVITY-020 (NCT01968109) is a phase 1/2a, dose-escalation and cohort-expansion, open-label trial. For Part E, pts with treatment-naïve advanced melanoma were randomized 1:1 to 480/480 vs 480/160 A single-arm cohort was enrolled to evaluate 480/480 dosing in pts with PD-1 RF advanced melanoma. Primary endpoints were safety and objective response rate (ORR) per BICR using RECIST v1.1. Secondary endpoints included duration of response (DOR) and PFS. Exploratory analyses included OS and pharmacodynamics. Results: As of clinical cutoff (May 22, 2024, min follow-up 33 mo), ORR was higher with the 1L 480/480 vs the 480/160 dose, while median (m) DOR, mPFS, and mOS were similar across the two arms, with numerically lower mPFS and 24-mo PFS/OS rates with the 480/480 dose (Table). With the 1L 480/480 vs 480/160 doses, 34% vs 36% of pts had grade 3–4 treatment-related AEs (TRAEs), and any grade TRAEs led to treatment discontinuation (d/c) in 29% vs 19% of pts, respectively. There was 1 treatment-related death (immune-mediated lung disease) in the 1L 480/480 arm vs none in the 480/160 arm. Treatment duration was shorter with the 1L 480/480 (m 5.6 mo) vs 480/160 dose (m 8.3 mo). Higher LAG-3 occupancy was observed with 1L 480/480 vs 480/160 dosing; however, there was no difference in Th1-associated cytokine levels. For the RF 480/480 arm (Table), outcomes were similar to published Part D data for RF 480/160 dosing. Conclusions: Although the 480/480 dose yielded higher ORR and LAG-3 occupancy levels than the 480/160 dose, it did not translate into improved survival outcomes or differences in Th1 cytokine levels. The 1L 480/480 dose also led to a higher rate of d/c due to TRAEs than the 480/160 dose. RF 480/480 dosing yielded similar results to RF 480/160 dosing from RELATIVITY-020 Part D. Clinical trial information: NCT01968109 . NIVO + RELA480/480(N = 77) NIVO + RELA480/160(N = 77) NIVO + RELARF 480/480(N = 95) ORR, % (95% CI) 61.0 (49.2–72.0) 48.7 (37.0–60.4) 10.6 (5.2–18.7) mDOR, mo (95% CI) NR (40.3–NR NR (32.3–NR) NR (3.1–NR) mPFS, mo (95% CI) 26.8 (11.1–NR) 33.3 (11.0–NR) 1.8 (1.8–3.4) PFS rate, % (95% CI)12 mo24 mo 61 (49–71)51 (39–62) 61 (49–72)56 (43–67) 19 (12–28)10 (5–19) mOS, mo (95% CI) NR (29.0–NR) NR (40.0–NR) 12.9 (8.2–16.6) OS rate, % (95% CI)12 mo24 mo 83 (73–90)66 (54–75) 82 (71–89)73 (62–82) 51 (40–60)30 (21–40) Grade 3/4 TRAEs, % 34 36 17 Any grade TRAE leading to d/c, % 29 19 9 NR, not reached.