ABSTRACT
This large randomized trial compared disease-free survival (DFS) and overall survival (OS) in tamoxifen non-responsive node-positive breast cancer patients randomized to AC 4 (60 mg/m2
and 600 mg/m2 d1 q21d), CMF 6 (C 100 mg/m2 p.o. d1-14, M 40 mg/m2 d1,8 i.v., F 600 mg/m2 d1,8 i.v. q28d), or AC 4 followed by a 6-month break and then CMF 3 (C 750 mg/m2
i.v. d1, M 40 mg/m2 d1,8 i.v., F 600 mg/m2 d1,8 i.v. q28d). Patients were stratified for the number of positive nodes, level of progesterone receptor (PgR) positivity, and type of surgery. At 26 months median follow-up, 2194 eligible patients were analysed: DFS (AC 62%, CMF 63%, AC/CMF 68% p 0.5), distant-DFS, and OS (AC 90.6%, CMF 90%, AC/CMF 90% p 0.8) were similar for all three groups. AC was tolerated better in that it was of shorter overall duration and had less hematologic toxicity, nausea, diarrhea, weight gain, and hemorrhagic cystitis; however, this group experienced more alopecia and vomiting. In the two AC groups, dose reductions were not allowed: delays were instituted if the white blood cell (WBC) or platelet counts were too low to deliver the next cycle at the scheduled time. In the CMF-only arm, after a delay to allow hematologic recovery, subsequent doses were reduced to 75% if the WBC was between 2500 and 3500 or platelets were 75,000-100,000; 100% were given if WBC was 3500 and platelets were 100,000. The median delivered doses of C, M, and F were slightly lower in this group compared with A and C in the first 2 groups: 30-35% of patients in the CMF group did not receive at least 80% of drugs at the planned dose intensity, compared with only 11-15% in the two AC groups (for A and C).
