BRAF and MEK inhibitors increase PD-1-positive melanoma cells leading to a potential lymphocyte-independent synergism with anti–PD-1 antibody

Purpose: BRAF and MEK inhibitors (BRAF/MEKi) favor melanoma-infiltrating lymphocytes, providing the rationale for current combinatorial trials with anti–PD-1 antibody. A portion of melanoma cells may express PD-1, and anti–PD-1 antibody could have a direct antitumor effect. Here, we explore whether BRAF/MEKi modulate rates of PD-1^þ melanoma cells, supporting an additional—lymphocyte-independent—basis for their therapeutic combination with anti–PD-1 antibody. Experimental Design: With data mining and flow cytometry, we assessed PD-1, PD-L1/2 expression on melanoma cell lines (CCLE, N ¼ 61; validation cell lines, N ¼ 7) and melanoma tumors (TCGA, N ¼ 214). We explored in vitro how BRAF/ MEKi affect rates of PD-1^þ, PD-L1/2^þ melanoma cells, and characterized the proliferative and putative stemness features of PD-1^þ melanoma cells. We tested the functional lymphocyte-independent effect of anti–PD-1 antibody alone and in combination with BRAF/MEKi in vitro and in an in vivo immunodeficient murine model. Results: PD-1 is consistently expressed on a small subset of melanoma cells, but PD-1^þ cells increase to relevant rates during BRAF/MEKi treatment [7.3% (5.6–14.2) vs. 1.5% (0.7–3.2), P ¼ 0.0156; N ¼ 7], together with PD-L2^þ melanoma cells [8.5% (0.0–63.0) vs. 1.5% (0.2–43.3), P ¼ 0.0312; N ¼ 7]. PD-1^þ cells proliferate less than PD-1 cells (avg. 65% less; t ¼ 7 days) and are preferentially endowed with stemness features. In vivo, the direct anti-melanoma activity of PD-1 blockage as monotherapy was negligible, but its association with BRAF/MEKi significantly delayed the development of drug resistance and tumor relapse. Conclusions: BRAF/MEKi increase the rates of PD-1^þ melanoma cells that may sustain tumor relapse, providing a lymphocyte-independent rationale to explore combinatory strategies with anti–PD-1 antibody.