Vaccine-induced immune responses, HLA genotyping, and molecular profiling in patients with metastatic ovarian cancer (the PESCO trial).

Abstract

5600 Background: PESCO, a phase 1/2 trial of Maveropepimut-S (MVP-S) in combination with Pembrolizumab and metronomic cyclophosphamide, showed safety and efficacy signals in metastatic ovarian cancer (OC). MVP-S uses the lipid-based DPX delivery platform to stimulate T cell immune response to survivin epitopes restricted by 5 HLA class I haplotypes (A1, A2, A3, A24, B7). Survivin is highly expressed in OC. We report exploratory correlative analysis including immune responses, HLA typing and molecular profiling. Methods: The study included a phase 1 dose escalation and expansion cohorts: A (platinum-sensitive high-grade serous [HGS]), B (platinum-resistant HGS), and C (rare OC histologies). Survivin-specific T-cell responses were assessed by flow cytometric detection of IFNγ following ex vivo PBMC re-stimulation with pooled survivin epitopes at on-treatment timepoints. A positive immune response was predefined as >0.1% IFNγ⁺ CD8⁺ T cells. HLA genotyping was mapped to MVP-S–restricted alleles. Molecular profiling was performed using next-generation sequencing (NGS) and/or whole-genome transcriptomic sequencing (WGTS). Results: Across 44 patients (pts), 37 (84%) carried ≥1 vaccine-relevant HLA allele and 23 (52%) carried ≥2. The most frequent alleles were A24 (16/44; 36%), A02 (15/44; 34%), A03 (12/44; 27%), B07 (11/44; 25%), and A01 (8/44; 18%). Survivin-specific immune responses were detected in 15/24 (62%) immune-evaluable pts and were strongly enriched among pts with clinical benefit: CR 1/1 (100%), PR 7/7 (100%), SD 6/10 (60%), versus PD 1/6 (16%); overall, 14/15 (93%) immune responders had clinical benefit. The longest survivin-specific response persisted for 195 weeks in a pt with MMR deficient tumor who had CR for 3 years. Exploratory HLA-outcome analysis suggested enrichment of A24 in pts with clinical benefit but absent in PD. NGS was performed in 44 pts and WGTS in 36. NGS identified at least one SNV/indel in 43/44 (97.7%) pts. In Phase 1 and Cohorts A/B (n=34, 68% HGS), TP53 was dominant (76.5%, 86% in HGS), followed by MYC (17.6%), NF1 (14.7%), and BRCA1/2 (each 8.8%). In Cohort C (n=10, 8 clear cell), ARID1A (70%) and PIK3CA (60%) were frequent. In platinum resistant pts (n=24), alterations included TP53 (67%), BRCA1/2 (21%), CCNE1(17%). Tumors with cell cycle/replication stress alterations were enriched among pts with PD (29% vs 8% in non-PD; P = 0.099) and immune non-responders (37.5% vs 6.7% in IFN responders; P = 0.103). Tumor mutation burden was low in all 26 pts tested. Conclusions: Survivin-specific immune responses were strongly enriched among pts with clinical benefit, supporting survivin as an active target in OC. HLA typing demonstrated broad coverage of MVP-S epitopes across pts. Molecular subtype may have influenced immune responses in this study. Our findings support biomarker-integrated studies to optimize immunotherapy combinations in OC. Clinical trial information: NCT03029403 .