Effect of MAF amplification on treatment outcomes with adjuvant zoledronic acid in early breast cancer: a secondary analysis of the international, open-label, randomised, controlled, phase 3 AZURE (BIG 01/04) trial

Background: Adjuvant use of bisphosphonates can reduce the incidence of bone metastases in early breast cancer. Recurrence and survival seem to be improved only in postmenopausal patients, but the underlying mechanisms remain unclear. We investigated whether MAF amplification (a biomarker for bone metastasis) in primary tumours could predict the treatment outcomes of adjuvant zoledronic acid. Methods: The study population included patients enrolled in the international, open-label, randomised, controlled, phase 3 AZURE trial at eligible UK sites who had stage II or III breast cancer and who gave consent for use of their primary tumour samples. Patients were randomly assigned (1:1) to receive standard adjuvant systemic therapy alone (control group) or with zoledronic acid every 3–4 weeks for six doses, then every 3–6 months until the end of 5 years. Minimisation took into account the number of involved axillary lymph nodes, clinical tumour stage, oestrogen-receptor status, type and timing of systemic therapy, menopausal status, statin use, and treating centre. The primary endpoint was disease-free survival; the secondary endpoint, invasive-disease-free survival, was the primary disease endpoint for the analysis in this report. MAF amplification was assessed by fluorescence in-situ hybridisation of two cores of breast tumour tissue in a microarray, done in a central laboratory by technicians unaware of treatment assignment. We used multivariate analyses to assess disease outcomes by intention to treat. We also assessed interactions between MAF-positive status and menopausal status on efficacy of zoledronic acid. The AZURE trial is registered with the International Standard Randomised Controlled Trial Registry, number ISRCTN79831382. Findings: 1739 AZURE patients contributed primary tumour samples, of whom 865 (50%) had two assessable cores (445 in the control groups and 420 in the zoledronic acid group). 184 (21%) tumours were MAF positive (85 in the control groups and 99 in the zoledronic acid group) and the remaining tumours were MAF negative. At a median follow-up of 84·6 months (IQR 72·0–95·8), MAF status was not prognostic for invasive-disease-free survival in the control group (MAF-positive vs MAF-negative: hazard ratio [HR] 0·92, 95% CI 0·59–1·41), but was in the zoledronic acid group (0·52, 0·36–0·75). In patients with MAF-negative tumours, zoledronic acid was associated with higher invasive-disease-free survival than was control treatment (HR 0·74, 95% CI 0·56–0·98), but not in patients who had MAF-positive tumours. Additionally, among 121 patients not postmenopausal at randomisation with MAF-positive tumours, zoledronic acid was associated with lower invasive-disease-free survival (HR 2·47, 95% CI 1·23–4·97) and overall survival (2·27, 95% CI 1·04–4·93) than control treatment. Interpretation: MAF status can predict likelihood of benefit from adjuvant zoledronic acid and merits further investigation as a potential companion diagnostic