Fifteen-year prognostic discriminants for invasive breast carcinoma

BACKGROUND This report updated previous findings, at the 15-year mark, of the National Surgical Breast and Bowel Project (NSAPB) Protocol B-06 with respect to the treatment of invasive breast carcinoma and the effects of pathologic features and the effects of some clinical features on its natural history. METHODS Thirty-one pathologic and 6 clinical features that were observed in a pathologic subset of 1039 evaluable patients were assessed as to their value in predicting survival, in predicting ipsilateral breast tumor recurrence (IBTR), and in predicting the necessity for local breast irradiation after lumpectomy. The patients had been randomly assigned to treatment by lumpectomy without local irradiation or to treatment by lumpectomy with local irradiation of the breast. A traditional and another statistical method were used for this purpose. RESULTS Multivariate analyses revealed that the presence of IBTR, race, histologic tumor type, nodal status, nuclear grade, and blood vessel invasion affected survival independently. Treatment, patient age, nuclear grade, presence of intraductal carcinoma, and a lymphocytic tumor infiltrate were features that predicted IBTR by multivariate analyses. Irradiation reduced IBTR from 36% to 12% in the analyzed cohort. A test of interaction failed to reveal any pathologic or clinical feature that might have allowed for the omission of local irradiation of the breast after lumpectomy. CONCLUSIONS In addition to the influence of pathologic and clinical features on patient survival and IBTR, the site, histopathologic features, and time of occurrence of the latter allowed for insights into some important biologic considerations concerning invasive breast carcinoma. Cancer 2001;91:1679–87. © 2001 American Cancer Society.