Six-year results of the Eastern Cooperative Oncology Group trial of observation versus CMFP versus CMFPT in postmenopausal patients with node-positive breast cancer.

Abstract
The Eastern Cooperative Oncology Group (ECOG) trial of adjuvant cyclophosphamide, methotrexate, fluorouracil, and prednisone (CMFP) or CMFP plus tamoxifen (CMFPT) for 1 year compared with observation alone in 265 postmenopausal patients with node-positive breast cancer is reported with 74 months median follow-up. Overall relapse-free survival tended to favor CMFPT (P = .08), but no survival differences existed between any treatment group. The addition of tamoxifen to CMFP led to slightly (but not significantly) better relapse-free status in all subgroups analyzed. Subgroup analysis based on stratification variables showed significant benefit from CMFP (+/- T) only in estrogen receptor (ER)-negative patients with respect to disease-free status (P = .0003), but not survival (P = .54). Relapse-free status was actually worse for CMFP-treated patients with ER-positive tumors, but not significantly so (P = .15). By multivariate analysis other significant risk factors for relapse-free status were primary tumor size, number of nodes pathologically involved, and the number of nodes examined. ER status was prognostic only for the observation group with the benefit from chemotherapy on ER-negative patients obliterating this difference in treated patients. Survival was affected by the number of involved nodes, tumor size, presence of tumor necrosis, and patient obesity. Analysis of toxicity showed elevation of liver enzymes during the first year to be more common in the observation group compared with those patients receiving adjuvant treatment and to be associated with early recurrence. Toxicity from adjuvant treatment persisted beyond termination of therapy in 53% of patients, but was usually mild and self-limited. We conclude CMFPT offers relapse-free survival benefit in ER-negative patients, but the value of chemotherapy in ER-positive postmenopausal, node-positive patients must be questioned.