LAUSR

OBJECTIVE To provide prognostic information from a large cohort of women with granulosa cell tumor we analyzed the National Cancer Database. METHODS We performed an observational retrospective cohort analysis of 2680 women with ovarian granulosa cell tumor from the 1998-2013 National Cancer Database. Kaplan-Meier and multivariable Cox proportional-hazards survival analyses were performed for the overall cohort and propensity score matched cohorts to examine the association of surgical staging and adjuvant chemotherapy with survival. A random forest was used to determine important prognostic factors in stages II-IV granulosa cell tumor. RESULTS Adjuvant chemotherapy, hormonal therapy, and radiotherapy were not associated with survival. Older age, more comorbidities, prior malignancy, higher stage, poor differentiation, larger tumor size, incomplete surgical staging, and residual disease at a surgical margin were independently associated with increased hazard of death. Among women with stage I disease, each one centimeter increase in tumor size was associated with 4% (2-6%) increased hazard of death (P<0.001). By matched cohort analyses, the hazard ratio (HR) (95% CI) for death associated with incomplete surgical staging was 1.77 (1.30-2.41), P<0.001 among women with stage I disease. Receiving adjuvant chemotherapy was not associated with increased survival among women with stages II-IV disease compared to no adjuvant treatment. CONCLUSION Incomplete surgical staging was associated with increased hazard of death. There was no evidence of increased survival with use of adjuvant chemotherapy. Early and complete surgical resection remains the best evidenced treatment for ovarian granulosa cell tumor.