10-year patterns of recurrence from early breast cancer: Analyses of the AZURE (BIG 01/04) study of adjuvant zoledronic acid.

INTRODUCTION: The AZURE clinical trial recruited 3360 patients with early breast cancer at moderate/high risk of recurrence, randomised 1:1 to receive standard adjuvant therapy +/-zoledronic acid (ZOL). The study offers an excellent opportunity to record patterns of relapse in the modern treatment era and the impact of adjuvant bisphosphonate treatment, not only in bone but also in local and other distant sites. METHODS: Patients were regularly reviewed during the 5-year treatment phase and then annually to 10 years (total median follow-up: 117 months) or until recurrence in bone or death. Local and distant recurrences were recorded and classified according to recurrence sites (with further breakdown of anatomical site for recurrence in bone), expression of estrogen receptor (ER) in the primary tumour and patient’s menopausal status at diagnosis. Data were adjudicated with trial site where necessary to remove errors or ambiguities. RESULTS: At 10 years, 1191 patients had developed disease recurrence (ZOL arm: 577 relapsed pts out of 1681; Control arm: 614 relapsed patients out of 1678). A summary of the data according to ER and menopausal status is listed below: • In control group, patterns of relapse were similar in postmenopausal (PM: at least 5 years since menopause) and non-postmenopausal (NPM) patients, with distant recurrence prevailing over loco-regional one; • In the same arm, bone recurrence (± extra-skeletal relapse) was the most frequent in ER positive ladies; • As expected, visceral relapse was more common in the ER negative group; • In ZOL group, reduction in bone metastases (especially “bone only”, as compared with bone and other sites at first recurrence) was greater in NPM patients (ER positive: 8.1% vs 11.1% of controls; ER negative: 3.6% vs 6.5% of controls); • ZOL increased extra-skeletal disease free survival (DFS) events in PM patients; • Little effect of ZOL was found on bone metastases occurring after a non-skeletal first relapse. Conclusions: These analyses yield valuable new data on the patterns of local and distant recurrence occurring over 10 years of post-diagnosis follow-up in patients with early breast cancer.