EP1116 Integrated pathway analysis identifies a 3-gene signature predicting platinum response and outcome of high grade serous ovarian carcinoma patients

Introduction/Background High-grade serous ovarian carcinoma (HGSOC), the most common and deadly epithelial ovarian cancer histotype is characterized by a heterogeneous genomic landscape. After surgery, all patients are treated with adjuvant platinum (Pt)-based chemotherapy to which the response is heterogeneous, ranging from cases sensitive (Platinum-sensitive, Pt-s cases), to intrinsically resistant patients, who relapse within 6 months from the end of therapy (Pt-resistant, Pt-r). The aim of the study was to investigate the mechanisms characterizing the biology of primary resistance to upfront Pt-based chemotherapy through an integrated pathway analysis. Methodology Gene and miRNA microarray experiments were carried out on 36 Pt-s and 41 Pt-r tumor samples. Gene and miRNA expression have been integrated using Micrographite algorithm. Expression levels of selected predictive and prognostic genes were validated on a total of 242 HGSOC samples by qRT-PCR and on the Curated Ovarian Database (including 838 HGSOC). Results Expression profiles of 131 mRNAs and five miRNAs, belonging to five different and functionally-related molecular pathways, discriminate Pt-s and Pt-r cases. We selected 23 elements of the networks for orthogonal validation and 19 coding genes confirmed as differentially expressed between Pt-r and Pt-s cases. Among them, 16 elements also showed independent prognostic value in terms of PFS and OS in multivariate analysis. The prognostic impact of this signature was in silico validated on the Curated Ovarian Database, resulting in a 3-gene signature as independent prognostic biomarker of survival for HGSOC. Conclusion In the current study, we used an integrated pathway analysis on miRNA and gene expression signatures to capture the key pathways shaping the different biology of Pt-r and Pt-s tumors, validating our results using two independent cohorts of HGSOC biopsies. Finally, we investigated the association of the validated signature with survival across multiple HGSOC databases. This strategy identified a three-gene signature with predictive and prognostic impact in HGSOC. Disclosure Nothing to disclose