20P - Deciphering tumor microenvironment heterogeneity between primary tumor and metastases in high grade serous ovarian cancer (HGSC) to predict response to immunotherapy (IT): analyses from the NeoPembrOv/GINECO phase II trial

Background

Biomarkers of response to IT are lacking in HGSC and PD-L1 expression failed to predict benefit in the randomized NeoPembrOv trial assessing peri-operative chemotherapy with or without pembrolizumab. We hypothesize that heterogeneity of samples site may impact the biomarker evaluation.

Methods

Biopsy of intra-abdominal metastases (M) or primary tubo-ovaries (TO) were obtained before any treatment. PD-L1 expression (Ventana SP263) (N = 85), multiplex immunofluorescence (mIF) (N = 64) and RNAseq data (N = 57) were assessed. Differentially expressed genes (DEG) and gene set enrichment analysis (GSEA) were obtained with DESeq2. Survival was analyzed with Cox models.

Results

No major difference of immune infiltration was seen between TO and M samples (mIF data). TO were enriched in tumor cells and TPS ≥ 1 samples (Table). In M, CPS ≥1 was predictive of better PFS in the pembrolizumab arm as compared to chemotherapy alone arm (HR interaction = 0.26 CI 95% 0.07-0.93 P = 0.034). No significant benefit was seen with TPS and IC, nor in TO. Intra-tumoral CD8 T cell (iCD8) density was higher in the IC, TPS and CPS ≥ 1 subgroups and was correlated with PD-L1 expression in M but not in TO (Table). Higher iCD8 T cells density (P = 0.03) and an enrichment of immune hallmarks (Interferon alpha and gamma, allograft rejection, inflammatory response, P < 0.05) was seen in CPS ≥ 1 M vs CPS ≥ 1 TO. On the other hand, CPS ≥ 1 TO samples were enriched in proliferative pathways (MYC and E2F target, G2M checkpoint, oxidative phosphorylation, P < 0.05) compared to CPS ≥ 1 M. Table: 20P

Conclusions

CPS ≥ 1 in M, but not in primitive tumor, is associated with pembrolizumab benefit, consistent with the increase of iCD8 T cells in PD-L1 positive M. This indicates that the site of tumor sampling influence prediction of potential efficacy of PDL1 inhibitors in HGSC.

Clinical trial identification

NCT03275506.

Legal entity responsible for the study

The authors.

Funding

Has not received any funding.

Disclosure

M. Leheurteur: Financial Interests, Personal, Advisory Board: MSD. F. Selle: Financial Interests, Personal, Invited Speaker: AstraZeneca, MSD, GSK/Tesaro, Eisai; Financial Interests, Personal, Advisory Board: AstraZeneca, MSD, GSK/Tesaro. C. Sotiriou: Financial Interests, Institutional, Advisory Board: Astellas, Vertex, Seattle Genetics, Amgen, INC, Merck & Co; Financial Interests, Personal, Advisory Board: Cepheid, Puma; Financial Interests, Personal, Invited Speaker: Eisai, Prime oncology, Teva; Financial Interests, Institutional, Other, Travel: Roche; Financial Interests, Institutional, Other, Internal speaker: Genentech; Financial Interests, Personal, Other, Regional speaker: Pfizer; Financial Interests, Institutional, Invited Speaker: Exact Sciences. I.L. Ray-Coquard: Financial Interests, Personal, Advisory Board: Roche, GSK, AstraZeneca, Mersana, Deciphera, Amgen, Oxnea, Merck Sereno, Agenus, Novartis, Macrogenics, Clovis, EQRX, Adaptimmune, Eisai, SUTRO, BMS, Adaptimmune, Daiichi Sankyo; Financial Interests, Institutional, Other, COLIBRI translational research: BMS; Financial Interests, Institutional, Advisory Board, translational research NEOPREMBROV trial: MSD; Non-Financial Interests, Personal, Principal Investigator: PAOLA1; Non-Financial Interests, Personal, Other, President: GINECO. O. Le Saux: Financial Interests, Personal, Advisory Board: Novartis, MSD, GSK; Financial Interests, Personal, Invited Speaker: Lilly, AstraZeneca, Clovis; Financial Interests, Institutional, Trial Chair: Novartis, Hospira-Pfizer foundation, Astellas. All other authors have declared no conflicts of interest.