Comparison of classic and quantitative prognostic factors in hormone receptor-positive and hormone receptor-negative female breast cancer.

The prognostic variables of 281 women with breast carcinoma (followed up for more than 8 years) were studied using Cox's analysis. Clinical and histologic features, nuclear morphometric variables, and mitotic indices were analyzed separately in progesterone receptor-negative (PR-) and -positive tumors (PR+). In PR- tumors, axillary lymph node status (p = 0.0025) and tumor size (p = 0.03) were predictors of survival in the univariate analysis. Tumor size (p < 0.0001), axillary lymph node status (p = 0.0006), the volume-corrected mitotic index (M/V index) (p = 0.0023), and the mitotic activity index (MAI) (p = 0.0067) were found to be related to survival according to univariate analysis of PR+ tumors. In PR- tumors, axillary lymph node status (p = 0.002), year of treatment (p = 0.017), and circumscription of the tumor margin (p = 0.02) had independent predictive value. In PR+ tumors, tumor size (p < 0.001), the MAI (p = 0.001), and axillary lymph node status (p = 0.04) predicted survival independently in Cox's analysis. In PR- tumors, histologic type (p = 0.008) was an independent predictor of recurrence-free survival, whereas in PR+ tumors, the M/V index (p < 0.001), tumor size (p = 0.007), and the standard deviation of the nuclear perimeter (p = 0.026) were independently related to recurrence-free survival. The results indicated that mitotic indices and nuclear morphometric variables are of limited value in predicting patient survival in breast carcinomas that are hormone receptor negative. Thus, a separate analysis is advocated for hormone receptor-positive and -negative tumors when the predictive value of quantitative measurements and histologic variables is tested in patients with breast cancer.