Dynamic Changes in PD-L1 Expression and Immune Infiltrates Early During Treatment Predict Response to PD-1 Blockade in Melanoma.

Abstract

Purpose: Disruption of PD-L1/cytotoxic T-cell PD-1 signaling by immune checkpoint inhibitors improves survival in cancer patients. This study sought to identify changes in tumoral PD-L1 expression and tumor-associated immune cell flux with anti-PD-1 therapies in patients with melanoma, particularly early during treatment, and correlate them with treatment response.Experimental Design: Forty-six tumor biopsies from 23 patients with unresectable AJCC stage III/IV melanoma receiving pembrolizumab/nivolumab were analyzed. Biopsies were collected prior to (PRE, n = 21), within 2 months of commencing treatment (EDT, n = 20) and on disease progression after previous response (PROG, n = 5). Thirteen patients responded (defined as CR, PR, or durable SD by RECIST/irRC criteria), and 10 did not respond.Results: PRE intratumoral and peritumoral PD-1+ T-cell densities were sevenfold (P = 0.006) and fivefold higher (P = 0.011), respectively, in responders compared with nonresponders and correlated with degree of radiologic tumor response (r = -0.729, P = 0.001 and r = -0.725, P = 0.001, respectively). PRE PD-L1 expression on tumor and macrophages was not significantly different between the patient groups, but tumoral PD-L1 and macrophage PD-L1 expression was higher in the EDT of responders versus nonresponders (P = 0.025 and P = 0.033). Responder EDT biopsies (compared with PRE) also showed significant increases in intratumoral CD8+ lymphocytes (P = 0.046) and intratumoral CD68+ macrophages (P = 0.046).Conclusions: Higher PRE PD-1+ T cells in responders suggest active suppression of an engaged immune system that is disinhibited by anti-PD-1 therapies. Furthermore, immunoprofiling of EDT biopsies for increased PD-L1 expression and immune cell infiltration showed greater predictive utility than PRE biopsies and may allow better selection of patients most likely to benefit from anti-PD-1 therapies and warrants further evaluation. Clin Cancer Res; 23(17); 5024-33. ©2017 AACR.