Overall survival with adjuvant zoledronic acid in patients with premenopausal breast cancer with complete endocrine blockade: Long-term results from ABCSG-12.

Abstract

520 Background: Adjuvant zoledronic acid (ZOL) significantly improved disease-free survival (DFS) in premenopausal patients (pts) with endocrine-responsive early breast cancer in ABCSG-12 (Gnant et al. NEJM 2009). The ZOL anticancer effect was confirmed in other prospective trials, such as ZO-FAST and the postmenopausal (> 5 yrs) cohort of AZURE. Long-term follow-up of ABCSG-12 (> 6 yr) now allows pre-planned subgroup analyses that more precisely identify the ZOL benefit, and interactions with pt and tumor characteristics. Methods: Premenopausal pts with early breast cancer (N = 1,803) were randomized to ovarian function suppression with goserelin (3.6 mg q28d) and tamoxifen (TAM; 20 mg/d) or anastrozole (ANA; 1 mg/d) ± ZOL (4 mg q6mo) for 3 yr. Endpoints included DFS and overall survival (OS), both analyzed using log-rank test and Cox models. Preplanned subgroup analyses included subpopulation treatment effect pattern plot (STEPP). Results: At 68 mo median follow-up, ZOL significantly reduced the risk of DFS events by 32% vs no ZOL (HR = 0.68; P = .009); in addition, there was a strong trend toward reduced risk of death (HR = 0.66; P = .08) with ZOL vs no ZOL. Multivariate analyses showed no interaction between ZOL benefit and ANA/TAM or any classic tumor parameter (eg, T, N, grade, ER). However, a strong interaction between ZOL and pt age was demonstrated. Pts > 40 yr of age (n = 1,390) derived significant DFS benefits from ZOL (HR = 0.58; 0.40-0.83; P = .003), whereas there was no apparent benefit in pts < 40 yr of age. ZOL reduced the risk of death by 42% in pts > 40 yr of age (P = .05), with presumed complete ovarian blockade. In several age subgroups > 40 yr, HRs < 0.5 were observed for both DFS and OS. Conclusions: ABCSG-12 long-term follow-up at 68 mo suggests that anticancer benefits of adjuvant ZOL mainly occur in pts with a low-estrogen environment. As in the postmenopausal (> 5 yrs) cohort of AZURE, ZOL leads to highly significant DFS and OS benefits. This suggests that both estrogen deprivation and reduction of bone-turnover–derived growth factors in the bone marrow microenvironment are needed for sufficient suppression of dormant micrometastases.