Abstract

Nuclear morphometry was performed on the diagnostic biopsy in 65 cases of non-mucinous ovarian carcinoma (FIGO stage IIB–IV) and its prognostic value regarding patient survival after the second-look operation was compared to that of morphology and clinical observations. In a univariate Cox survival analysis four morphometric factors were found to be significant predictors of survival (the standard deviations (SD) of the nuclear area, perimeter, largest perpendicular axis, and largest axis). Age, the size of residual tumor after the primary operation, and a combined variable describing the status at the second-look operation and also the result of tumor reduction were significant clinical variables. None of the morphologic variables proved to be significant. In the multivariate Cox analysis the SD of the largest perpendicular nuclear axis gave independent prognostic information together with either the size of residual tumor after the primary laparotomy (P = 0.00004) or the second-look variable (P < 0.00001). When the SD of the largest perpendicular nuclear axis and the second-look variables were included in the model the size of residual tumor after the primary operation added no further prognostic information. We conclude that nuclear morphometry is a simple, easily implemented and cheap quantitative method which gives objective and valuable prognostic information regarding survival in advanced ovarian cancer.