Background: Overexpression of the erbB-2 protein by breast cancer cells has been suggested to be a predictor of response to doxorubicin. A retrospective study was designed to test this hypothesis. Methods: In National Surgical Adjuvant Breast and Bowel Project protocol B-ll, patients with axillary lymph node-positive, hormone receptor-negative breast cancer were randomly assigned to receive either L-phenylalanine mustard plus 5-fluorouracil (PF) or a combination of L-phenylalanine mustard, 5-fluorouracil, and doxorubicin (PAF), Tumor cell expression of erbB-2 was determined by immunohistochemistry for 638 of 682 eligible patients. Statistical analyses were performed to test for interaction between treatment and erbB-2 status (positive versus negative) with respect to disease-free survival (DFS), survival, recurrence-free survival (RFS), and distant disease-free survival (DDFS), Reported P values are two-sided. Results: Overexpression of erbB-2 (i.e., positive immunohistochemical staining) was observed in 239 (37.5%) of the 638 tumors studied. Overexpression was associated with tumor size (P = .02), lack of estrogen receptors (P = .008), and the number of positive lymph nodes (P = .0001), After a mean time on study of 13.5 years, the clinical benefit from doxorubicin (PAF versus PF) was statistically significant for patients with erbB-2-positive tumors-DFS: relative risk of failure (RR) = 0.60 (95% confidence interval [CI] = 0.44-0.83), P = .001; survival: RR = 0.66 (95% CI = 0.47-0.92), P = .01; RFS: RR = 0.58 (95% CI = 0.42-0.82), P = .002; DDFS: RR = 0.61 (95% CI = 0.44-0.85), P = .003, However, it was not significant for patients with erbB-2-negative tumors-DFS: RR = 0.96 (95% CI = 0.75-1.23), P = .74; survival: RR = 0.90 (95% CI = 0.69-1.19), P = .47; RFS: RR = 0.88 (95% CI = 0.67-1.16), P = .37; DDFS: RR = 1.03 (95% CI = 0.79-1.35), P = .84, Interaction between doxorubicin treatment and erbB-2 overexpression was statistically significant for DFS (P = .02) and DDFS (P = .02) but not for survival (P = .15) or RFS (P = .06), Conclusions: These data support the hypothesis of a preferential benefit from doxorubicin in patients with erbB-2-positive breast cancer.
