Primary analysis of a phase 2, open-label, multicenter trial of talimogene laherparepvec (T-VEC) plus pembrolizumab (pembro) for the treatment (Tx) of patients (pts) with advanced melanoma (MEL) who progressed on prior anti–PD-1 therapy: MASTERKEY-115.

Abstract

9518 Background: Despite advances in anti–PD-1–based Tx for MEL, an unmet need remains for immunotherapy failure in advanced metastatic or unresectable MEL. Also, there is a growing population of pts who received adjuvant anti–PD-1 and recurred; yet trials to address this population are lacking. Combination Tx with T-VEC, an oncolytic viral intratumoral Tx designed to produce GM-CSF, and pembro, an anti–PD-1 agent, may overcome immunotherapy failure. We report the MASTERKEY-115 primary analysis on the efficacy and safety of T-VEC + pembro in pts with advanced MEL who had progressive disease (PD) on prior anti–PD-1. Methods: This open-label, single-arm, multicenter, phase 2 study (NCT04068181) enrolled pts at 26 international sites (Jan 2020‒Feb 2021). Cohorts 1 and 2, primary and acquired resistance, respectively, received anti–PD-1 in a locally recurrent or metastatic setting and had PD within 12 wks of the last anti–PD-1 dose. Cohorts 3 and 4 included only pts who received adjuvant anti–PD-1 and were disease-free for < 6 mos (Cohort 3) or ≥ 6 mos (Cohort 4) before confirmed PD. Eligible pts had histologically confirmed unresectable or metastatic stage IIIB–IVM1d MEL, measurable and injectable disease, ECOG PS 0/1, and progressed on anti–PD-1 directly before enrollment. T-VEC at standard dosage and pembro 200 mg were given Q3W. The primary endpoint was objective response rate (ORR). Key secondary endpoints were complete response (CR) rate, progression-free survival (PFS), and safety. Tx decisions were per modified immune-related response criteria (irRECIST). Results: 72 pts (median age, 65 y) were enrolled. Of the 71 evaluable pts, 37 (52.1%) had stage IVM1b‒d disease, 30 (42.3%) had confirmed PD-L1–positive tumor (CPS ≥ 1%), 20 (28.2%) had a BRAFV600 mutation, and 21 (29.6%) had LDH > 1ULN. At data cutoff (Aug 2021), 47 (65.3%) pts remained on study. ORR was 0%, 6.7%, 40%, and 46.7% in cohorts 1–4, respectively (table). Any-grade Tx-related adverse events (TRAEs) were reported in 54 (76.1%) pts; the most common were pyrexia (29.6%), fatigue, and influenza-like illness (15.5% each). Grade ≥ 3 TRAEs occurred in 9 (12.7%) pts. Conclusions: T-VEC + pembro showed manageable safety in pts with advanced MEL after anti–PD-1 failure; the promising ORR observed in pts who progressed on prior adjuvant anti–PD-1 warrants further analysis. Clinical trial information: NCT04068181. [Table: see text]