Ovarian carcinoma: a multivariate analysis of prognostic factors

Abstract

Five hundred and fifty six women with invasive epithelial ovarian carcinoma were assessed for postoperative treatment between 1966 and 1976. The records of this group were reviewed retrospectively. Sixteen characteristics of the patient and tumor were analyzed for prognostic significance by univariate and multivariate techniques. Tumor grade, the presence of residual disease, and patient performance status are identified as stage-specific independent prognostic factors. These independent factors define patient subsets with good, intermediate, and poor prognosis within each stage. The results show that more effective treatment strategies are required for patients in poor prognosis subsets. Estimates of tumor grade, extent of residual disease, and performance status should be included in reports of treatment outcome.