Associations between residual disease and survival in epithelial ovarian cancer by histologic type.

Abstract

Surgical cytoreduction has been postulated to affect survival by increasing the efficacy of chemotherapy in ovarian cancer. We hypothesized that women with high-grade serous ovarian cancer, which usually responds to chemotherapy, would derive greater benefit from complete cytoreduction than those with histologic subtypes that are less responsive to chemotherapy, such as mucinous and clear cell carcinoma. We conducted a retrospective cohort study of patients who underwent primary cytoreductive surgery and adjuvant chemotherapy for stage IIIC or IV epithelial ovarian cancer from 2011 to 2013 using data from the National Cancer Database. We constructed multivariable models to quantify the magnitude of associations between residual disease status (no residual disease, ≤1cm, or >1cm) and all-cause mortality by histologic type among women with clear cell, mucinous, and high-grade serous ovarian cancer. Because 26% of the sample had unknown residual disease status, we used multiple imputations in the primary analysis. We identified 6,013 women with stage IIIC and IV high-grade serous, 307 with clear cell, and 140 with mucinous histology. The association between residual disease status and mortality hazard did not differ significantly among histologic subtypes of ovarian cancer (p for interaction=0.32). In covariate adjusted models, compared to suboptimal cytoreduction, cytoreduction to no gross disease was associated with a hazard reduction of 42% in high-grade serous carcinoma (hazard ratio [HR]=0.58, 95% confidence interval [CI]=0.49-0.68), 61% in clear cell carcinoma (HR=0.39, 95% CI=0.22-0.69), and 54% in mucinous carcinoma (HR=0.46, 95% CI=0.22-0.99). We found no evidence that surgical cytoreduction was of greater prognostic importance in high-grade serous carcinomas than in histologies that are less responsive to chemotherapy.