Molecular clustering based on ERα and EIG121 predicts survival in high-grade serous carcinoma of the ovary/peritoneum

Assessment of estrogen receptor (ER) expression by immunohistochemistry has yielded inconsistent results as a prognostic indicator in ovarian carcinoma. In breast and endometrial carcinomas, panels of estrogen-induced genes have shown improved prognostic capability over the use of ER immunohistochemistry alone. For both breast and endometrial cancers, overexpression of estrogen-induced genes is associated with better prognosis. We hypothesized that analysis of a panel of estrogen-induced genes can predict the outcome in ovarian carcinoma and potentially differentiate between tumors of varying hormonal responsiveness. From a cohort of 219 women undergoing ovarian cancer surgery from 2004 to 2007, 83 patients were selected for inclusion. All patients had advanced stage ovarian/primary peritoneal high-grade serous carcinoma and underwent primary surgical debulking, followed by adjuvant treatment with platinum and taxane agents. The expression of ERα and six genes known to be induced by estrogen in the female reproductive tract (namely EIG121, sFRP1, sFRP4, RALDH2, PR, and IGF-1) was measured using quantitative RT-PCR. Unsupervised cluster analyses were used to categorize patients as high or low gene expressors. Gene expression results were then compared with those for ER immunohistochemistry. Clusters were compared using χ2analyses, and Cox proportional hazards models were used to evaluate survival outcomes. The median follow-up time was 38.7 months (range: 1–68). A cluster defined by EIG121 and ERα segregated tumors into distinct groups of high and low gene expressors. Shorter overall survival (OS) was associated with high gene expression (HR 2.84 (1.11–7.30), P=0.03), even after adjustment for other covariates. No difference in ER immunohistochemistry expression was noted between gene clusters. In contrast to other hormonally driven cancers, high expression of ERα and the estrogen-induced gene EIG121 predicts shorter OS in patients with high-grade serous ovarian carcinoma. Such a biomarker panel may potentially be used to guide management with estrogen antagonists in this patient population.