Methods We did a meta-analysis of three clinical trials-the Austrian Breast and Colorectal Cancer Study Group (ABCSG 8), Arimidex-Nolvadex (ARNO 95), and the Italian Tamoxifen Anastrozole (ITA) studies-in which postmenopausal women with histologically confirmed, hormone-sensitive early-stage breast cancer were randomised to 1 mg/day anastrozole (n=2009) after 2-3 years of tamoxifen treatment or to continued 20 or 30 mg/day tamoxifen (n=1997). We analysed the data with a stratified Cox proportional hazards model with the covariates of age, tumour size, nodal status, grade, surgery, and chemotherapy.

Findings Patients who switched to anastrozole had fewer disease recurrences (92 vs 159) and deaths (66 vs 90) than did those who remained on tamoxifen, resulting in significant improvements in disease-free survival (hazard ratio 0 59 [95% CI 0.48-0.74]; p<0.0001), event-free survival (0.55 [0.42-0.71]; p<0.0001), distant recurrence-free survival (0 61 [0.45-0.83]; p=0.002), and overall survival (0 .71 [0.52-0.98]; p=0.04).

Interpretation Our results show that the clinical benefits in terms of event-free survival seen in individual trials for those patients who switched to anastrozole translate into a benefit in overall survival. These findings confirm that clinicians should consider switching postmenopausal women who have taken adjuvant tamoxifen for 2-3 years to anastrozole.