Abstract B18: miRNA landscape analysis of stage I EOC, identifies miR-199a-5p associated to poor prognosis in grade 3 subgroup

Abstract

Abstract Introduction. Within stage I epithelial ovarian cancer (EOC), the current clinic-pathological parameters, like tumor grade, fail to accurately stratify patient prognosis and it is therefore crucial for optimal treatment that the biological properties of stage I EOCs are further elucidated. We have previously demonstrated miR-200c as a predictor of survival, and a biomarker of relapse (Marchini et al. Lancet Oncology, 2011), suggesting that miRNA profile could be a useful tool to dissect molecular networks in stage I EOC. The aim of the current study is to identify a miRNA signature for each tumor grade, that integrated with clinical variables would be used to improve stage I patients stratification. Experimental procedures. A cohort of 219 snap frozen tumor biopsies, with median follow up of seven years, was gathered together from three independent Italian tumor tissue collections. miRNA landscape was generated with commercially available arrays (Agilent, Palo Alto CA) and analysis performed as recently published (Calura et al., CCR 2013). Signature validation was performed by qRT-PCR using commercially available primers and reagents (Qiagen, Milano, Italy). Results. The entire cohort of patients was stratified by sub-stage, grade and relapse into a training set (n= 151), used for miRNA landscape generation, and a validation set (n= 68) used for qRT-PCR validation. “Resampling score” (RS) strategy (Calura et al., CCR 2013) reported that the largest number of miRNAs found differentially expressed is between grade three and grade one (n= 72), while the comparison between grade one versus borderline tumors showed the lowest number (n= 14). Signature validation in both training and validation set by qRT-PCR of the top seven selected miRNAs with highest RS, confirmed hsa-miR-376c, hsa-miR-377 and hsa-miR-214 as down-regulated in grade three compared to the other grades; hsa-miR-96 expression increases directly from grade one to grade three, while hsa-miR-199a-5p was down-regulated in grade three compared to borderline tumors. No differences were observed for hsa-miR-183 and hsa-miR-29c. miRNA expression profile was correlated to clinical variables in both univariate and multivariate model and only miR-199a-5p, resulted associated to PFS in multivariate Cox proportional hazard model. Conclusions. In the present study we observed that, regardless of tumor histological subtype (Calura et al. CCR 2013), known morphological differences across tumor grades mirror molecular differences in term of miRNA expression profile. To optimize patients stratification and thus improving clinical management of stage I EOC, we are now drawing new miRNA-based networks (i.e. miRNA-gene expression integration) for each tumor grade that will be correlated with known clinical parameters. Citation Format: Enrica Calura, Robert Fruscio, Lara Paracchini, Eliana Bignotti, Paolo Martini, Antonella Ravaggi, Mariacristina Di Marino, Gabriele Sales, Luca Beltrame, Federica Dell'Orto, Romina Baldo, Sergio Pecorelli, Enrico Sartori, laura Zanotti, Dionyssios Katsaros, Germana Tognon, Maurizio D'Incalci, Chiara Romualdi, Sergio Marchini. miRNA landscape analysis of stage I EOC, identifies miR-199a-5p associated to poor prognosis in grade 3 subgroup. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research: From Concept to Clinic; Sep 18-21, 2013; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2013;19(19 Suppl):Abstract nr B18.