Clustering of Ovarian High-Grade Serous Carcinoma Based on Estrogen-Induced Gene Expression Predicts Poor Survival
M Schlumbrecht, D Urbauer, R Broaddus. M.D. Anderson Cancer Center, Houston, TX
Background: Assessment of estrogen receptor (ER) expression has yielded mixed results as a prognostic indicator in epithelial ovarian carcinoma. In breast and endometrial cancers, the use of panels of estrogen-induced genes has improved prognostic capability over the use of ER alone. For both breast and endometrial cancer, over-expression of estrogen-induced genes is associated with better prognosis. We hypothesized that estrogen-induced gene expression can predict outcome in ovarian carcinoma and differentiate between tumors of varying estrogen sensitivities.
Design: qRT-PCR was used to quantify the expression of six genes known to be induced by estrogen in the female reproductive tract (EIG121, sFRP1, sFRP4, RALDH2, PR, and IGF-1) and ER in 83 patients with advanced stage, high-grade serous carcinoma of the ovary or peritoneum who subsequently received adjuvant treatment with platinum and taxane agents. Clinical data was collected by retrospective chart review. Unsupervised cluster analyses in multiple permutations were used to categorize patients as high or low estrogen-induced gene expressors. Statistical analyses were performed using Fisher's exact test, Cox proportional hazards models, and the Kaplan-Meier method.
Results: Median follow-up time was 38.7 months (range 1-68 months). In a multivariate Cox model, overall survival was predicted by optimal debulking (HR 0.37 [CI 0.15-0.95], p=0.038), sFRP1 expression (HR 1.04 [CI 1.00-1.07], p=0.0274), and EIG121 expression (HR 1.19 [CI 1.06-1.33], p=0.0026). EIG121 expression was also associated with recurrence-free survival (HR 1.15 [CI 1.03-1.24], p=0.011). A cluster defined by EIG121 and ER best segregated tumors into groups of high and low estrogen-induced gene expressors. High expressors demonstrated significantly shorter overall survival compared to low expressors, even when adjusting for race, body mass index, and residual disease at debulking (HR 2.73, p=0.037). Nonwhite women also represented a greater proportion of high expressors compared to low expressors (p=0.044).
Conclusions: In sharp contrast to breast and endometrial cancers, high estrogen-related gene expression negatively predicts overall survival in patients with high-grade serous ovarian carcinoma. An estrogen-induced gene biomarker panel may have utility both as an indicator of prognosis and as a guide to management with estrogen antagonists.
Category: Gynecologic & Obstetrics
Tuesday, March 23, 2010 11:45 AM
Platform Session: Section D, Tuesday Morning