Abstract S6-4: Vitamin D, but not bone turnover markers, predict relapse in women with early breast cancer: an AZURE translational study.

Abstract

Abstract Background: The AZURE trial evaluated the addition of zoledronic acid (ZOL) to standard adjuvant therapy on relapse rates and survival in pts with stage II/III breast cancer. While the overall analysis reported no benefit, a pre-planned subgroup analysis showed significant benefits in postmenopausal (PM) women. Here we report whether baseline measurements of 25-hydroxyvitamin D (25-OHD) and the serum bone turnover markers N-terminal propeptide type I procollagen (P1NP) and beta C-terminal telopeptide type I collagen (β-CTX) can identify pts at high risk of relapse and/or modify the treatment effects of ZOL. Methods: 872 AZURE pts consented for serum to be collected and stored for translational research projects (ZOL group n=431; control group n=441). Serum P1NP, β-CTX and 25-OHD levels were measured in 867, 863 and 856 pts respectively. These markers were analysed as categorical variables with determined cut-off points to predict bone as 1st site of distant recurrence (DR) or any distant relapse (P1NP (ng/ml), >70 versus ≤70; βCTX (ng/ml), >0.299 versus ≤0.299; 25-OHD (ng/ml), ≤30 versus >30). Additionally, the markers were assessed for an interaction with ZOL. Analyses were adjusted for systemic therapy, lymph node and ER status. As in previous analyses of AZURE, the PM cohort was defined as ≥5 years since last menses; all other patients are included in a “not postmenopausal” (not-PM) group. Results: The baseline disease characteristics of the translational cohort were similar to those of the total AZURE population and showed a similar benefit of ZOL among the PM patients: hazard ratio (HR) for DR = 0.61 (95% CI 0.35, 1.07; p-value for interaction with not-PM women = 0.0173). At a median follow-up of 53.3 months, 150 pts have had a 1st DR event; 47 had bone-only relapse, 19 bone plus other distant site and 84 developed non-bone DR. Mean value of 25-OHD (ng/ml) was 18.30 (range, 3.16–54.82; SD=9.19) with 61.0% of pts <20, 26.8% 20–30 and 10.3% >30. 25-OHD levels >30 ng/ml were significantly associated with lower risk for bone relapse (HR 0.11, 95% CI 0.02, 0.76; p-value 0.0257). A similar trend was seen for any site of DR (HR 0.56, 95% CI 0.31, 1.01; p-value = 0.0519). Additionally, 25-OHD levels were prognostic for bone relapse when analysed as a continuous value (HR 0.97, 95% CI 0.94, 1.00; p-value = 0.0474). Finally, 25-OHD levels >30 ng/ml predicted for benefit of treatment with ZOL in PM women with regards to DR (HR 0.09, 95% CI 0.01, 0.82; interaction p-value 0.0747) but the trend was not seen in not-PM women. Mean values for P1NP and βCTX (range; SD) were 59.1 ng/ml (<5–229.6; 27.0) and 0.259 (0.027–1.07; 0.153) respectively. Neither of these markers reached statistical significance as either prognostic or predictive factors. Conclusions: High pre-treatment serum levels of 25-OHD are associated with lower risk for bone relapse with a trend for any DR. Amongst PM women, high levels can additionally predict improved outcomes for treatment with ZOL. P1NP and βCTX failed to significantly discriminate breast cancer pts who may be at high risk for DR. Our results suggest that baseline bone turnover is not directly associated with the benefits of ZOL amongst PM women. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr S6-4.