PATIENTS AND METHODS
We conducted a randomized, multicenter study of intravenous cyclophosphamide 500 mg/m2 plus fluorouracil 500 mg/m2 combined with either mitoxantrone (Novantrone, Lederle Cyanamid Canada Ltd, Willowdale, Ontario) 10 mg/m2 (CNF) or doxorubicin (Adriamycin, Adria Laboratories of Canada Ltd, Mississauga, Ontario) 50 mg/m2 (CAF) every 3 weeks in advanced breast cancer.
The response rate in 249 randomized patients was 36% with CNF (44 of 121) and 48% with CAF (62 of 128) (P = .054), with complete remissions in 10 patients (8.3%) on CNF and in 13 (10.2%) on CAF. If only fully assessable patients are considered, the response rate was 48% (44 of 91) with CNF and 60% (62 of 103) with CAF (P = .098). At time of analysis, all except 10 patients (one CNF and nine CAF) had died. The median survival time with CAF was longer than with CNF (15.2 v 10.9 months; P = .003), and time to progression was also longer with CAF (5.3 v 3.2 months; P < .03). Survival differences remained significant (P = .006) if patients who failed to meet all eligibility criteria were excluded. Favorable prognostic factors for survival in a Cox regression model included good performance status (P < .0001); less than two organ systems involved by tumor (P < .0001); no involvement of lung, liver, or brain (P < .003); involvement of bone or bone marrow (P < .009), prior surgery for breast cancer (P < .006); being premenopausal (P < .03); > or = 3 years from diagnosis until randomization on this study (P < .03); and treatment with CAF (P < .03). Alopecia > or = grade 3 was reported in 55% of patients with CAF and 12% of patients with CNF (P < .001), while other > or = grade 3 toxicities did not differ significantly. Priestman-Baum quality-of-life assessment was comparable on the two study arms.
In patients with advanced breast cancer, CAF was associated with longer survival than was CNF, with an increase in alopecia, but not in other toxicities.