PD-L1 expression as a biomarker for nivolumab (NIVO) plus ipilimumab (IPI) and NIVO alone in advanced melanoma (MEL): A pooled analysis

Abstract

Background: NIVO + IPI and NIVO showed superior clinical activity vs IPI in a phase 3 trial of MEL patients (pts), irrespective of PD-L1 tumor expression. Among pts with high PD-L1 expression (≥5%), median progression-free survival (mPFS) was similar between NIVO + IPI and NIVO, but overall response rate (ORR) was higher with NIVO + IPI. We describe PD-L1 as a biomarker for NIVO + IPI and NIVO efficacy across phase 2 (CheckMate 069) and phase 3 (CheckMate 066 and 067) trials. Methods: Treatment-naïve pts (N = 832) with MEL received NIVO 1 mg/kg + IPI 3 mg/kg Q3W × 4 or NIVO 3 mg/kg Q2W, followed by NIVO 3 mg/kg Q2W until progression or unacceptable toxicity. Tumor tissue from primary or metastatic sites, obtained at screening, was assessed for PD-L1 expression using a validated Dako immunohistochemistry assay. Minimum pt follow-up was 18 months (mos). Survival data remain immature. Results: The proportion of pts with PD-L1 expression ≥5% was 26% (92/358) for NIVO + IPI and 29% (139/474) for NIVO. Pt characteristics were similar between PD-L1 subgroups, although fewer pts had LDH > ULN in the PD-L1 ≥5% subgroup. Among pts with PD-L1 expression ≥5%, mPFS of NIVO + IPI was not reached (NR) and was 22.0 mos for NIVO alone (hazard ratio [HR]: 0.99, 95% CI: 0.66─1.46). For pts with low to no PD-L1 (<5%), mPFS was 11.1 mos for NIVO + IPI and 4.9 mos for NIVO (HR: 0.70, 95% CI: 0.57─0.87). ORR was higher with NIVO + IPI vs NIVO in pts with ≥5% (68.5% vs 59.0%) and <5% (54.9% vs 39.7%) PD-L1 expression. Median duration of response was NR in both PD-L1 subgroups for NIVO + IPI, and 20.8 and 22.3 mos in NIVO ≥5% and <5% PD-L1 subgroups, respectively. The frequency and types of treatment-related grade 3-4 adverse events were consistent with earlier reports (NIVO + IPI: 56.5%, NIVO: 18.2%) and did not differ by PD-L1 expression. Conclusions: While pts with ≥5% PD-L1 tumor expression have better efficacy outcomes, those with <5% PD-L1 expression still benefit from NIVO + IPI or NIVO. Among pts with high PD-L1, mPFS of NIVO + IPI and NIVO were similar, but the ORR of NIVO + IPI was numerically higher across PD-L1 subgroups. As OS data have not yet matured, caution is advised when applying these results to assess the relative benefit of NIVO + IPI vs NIVO.