Background:

The aim of this study was to assess the efficacy of neoadjuvant anastrozole and fulvestrant treatment of large operable or locally advanced hormone-receptor-positive breast cancer not eligible for initial breast-conserving surgery, and to identify genomic changes occurring after treatment.

Methods:

One hundred and twenty post-menopausal patients were randomised to receive 1 mg anastrozole (61 patients) or 500 mg fulvestrant (59 patients) for 6 months. Genomic DNA copy number profiles were generated for a subgroup of 20 patients before and after treatment.

Results:

A total of 108 patients were evaluable for efficacy and 118 for toxicity. The objective response rate determined by clinical palpation was 58.9%(95% CI= 45.0–71.9) in the anastrozole arm and 53.8%(95% CI= 39.5–67.8) in the fulvestrant arm. The breast-conserving surgery rate was 58.9%(95% CI= 45.0–71.9) in the anastrozole arm and 50.0%(95% CI= 35.8–64.2) in the fulvestrant arm. Pathological responses> 50% occurred in 24 patients (42.9%) in the anastrozole arm and 13 (25.0%) in the fulvestrant arm. The Ki-67 score fell after treatment but there was no significant difference between the reduction in the two arms (anastrozole 16.7%(95% CI= 13.3–21.0) before, 3.2%(95% CI= 1.9–5.5) after, n= 43; fulvestrant 17.1%(95% CI= 13.1–22.5) before, 3.2%(95% CI= 1.8–5.7) after, n= 38) or between the reduction in Ki-67 in clinical responders and non-responders. Genomic analysis appeared to show a reduction of clonal diversity following treatment with selection of some clones with simpler copy number profiles.

Conclusions:

Both anastrozole and fulvestrant were effective and well-tolerated, enabling breast-conserving surgery in over 50% of patients. Clonal changes consistent with clonal selection by the treatment were seen in a subgroup of patients.