Association of HIPEC response in ovarian cancer with PI3K/RAS/Notch gene signatures: A whole transcriptomic analysis of U.S. and French HIPEC treated ovarian cancer patients.

Abstract

5569 Background: Hyperthermic intraperitoneal chemotherapy (HIPEC) is associated with improved overall survival in Stage III epithelial ovarian cancer (EOC) patients. We set out to evaluate the gene signatures associated with HIPEC response in EOC patients. Methods: Ninety-one EOC patients who underwent HIPEC with pre-operative tumor samples at City of Hope (51) and CHU Lyon (40) were identified between 2014 and 2022. RNA isolation was performed from formalin-fixed paraffin-embedded samples, followed by Whole-transcriptome library construction. Following exclusion of non-high grade serous (HGS) samples, and quality control steps, twenty-four samples were excluded. Progression-free survival (PFS) was used to define HIPEC response. Cut-off PFS values were used to distinguish good vs poor responders in primary EOC patients (18 months, based on KGOG, CARCINO-HIPEC trials), and recurrent EOC patients (12 months, based on MSK, CHIPOR HIPEC trials). Differential Gene Expression Analysis comparing good and poor HIPEC responders identified significantly changed genes. Pathway analysis was conducted using gene set enrichment analysis (GSEA) against Hallmark. Results: A total of sixty HGS tumor samples with available survival data were analyzed. 63.3% were primary EOC, 36.7% recurrent EOC. Germline BRCA mutations affected 21.7% of patients. With a median follow up of 31.9 months, median PFS was 29.3 (95%CI: 15.3, 63.5) months in primary EOC patients and 26.0 (95%CI: 14.7, 37.1) months in recurrent patients. Median OS was not reached in either group. 60.0% had a recurrence. Thirty-eight patients were identified as good responders, with a median PFS of 37.1 mos. (95%CI: 26.4, NR); 18 patients were identified as poor responders, with median PFS of 11.4 months (95%CI: 7.5, 14.2). Differential gene expression analysis between good and poor responders revealed 29 significantly upregulated 35 downregulated genes in HIPEC responders. Top upregulated genes in HIPEC responders include MAPK signaling pathway genes (RIB2, ETV5, CAPN8, IGFR1), in addition to CCND1 and CEACAM1. In HIPEC responders, the top-ranking gene sets in the transcriptional signature included Notch, KRAS, and Wnt/beta-catenin signaling pathways. In poor HIPEC responders, the DNA damage repair associated pathways E2F targets and G2M checkpoint, were activated. Similar transcriptomic pathway signatures were observed in Non-recurrent versus Recurrent HIPEC patients: Non-recurrent tumors were enriched with Notch signaling, while Recurrent tumors were enriched with E2F target and G2M checkpoint pathways. Conclusions: Good HIPEC response is characterized by transcriptional signatures consistent with Type I EOC characteristics of PI3K/RAS/Notch signaling. Recurrence after HIPEC in HGS ovarian cancer is higher in patients with E2F/G2M transcriptional signatures.