Abstract P3-07-05: Bisphosphonate use after primary breast cancer and risk of contralateral breast cancer using pharmacy data

Abstract

Abstract Background: It is not clear if bisphosphonates (BP) are associated with improved breast cancer prognosis in women with early breast cancer. Clinical trials have reported mixed results, yet BPs may be most beneficial in patients with low estrogen levels. However, BPs and risk of second (contralateral) breast cancer has been minimally studied. We examined the association of oral BP use (ever/never and duration) on risk of contralateral breast cancer (CBC) in 17,224 women with early stage breast cancer treated with tamoxifen. Materials and Methods: A cohort was assembled of women diagnosed with their first primary breast cancer (Stage 0, I, II) from 1996 to 2007 on tamoxifen and followed through 31 December 2009 at Kaiser Permanente Northern and Southern California. Demographic, tumor, pharmacy, and cancer treatment information was extracted from electronic medical records and SEER-affiliated cancer registries at each site. Second (contralateral) tumors were identified from the cancer registries. Detailed information on oral BP use (before and after initial breast cancer diagnosis) was obtained from pharmacy databases. A record of >90 days supply was considered the minimum exposure. Initiation and duration of post-diagnosis use was categorized as 1) non-use (≤90 days supply) and use (>90 days supply) and 2) non-use (≤90 days supply), 91 days–<1 year supply, 1-<2 years supply, and ≥2 years supply. Delayed entry Cox regression models were used to estimate hazard ratios (HR) and 95% confidence intervals (CI) with BP use modeled as a time-varying covariate. Women were followed through CBC, health plan disenrollment, death, or end of study, whichever occurred first. We adjusted for age, stage and year of diagnosis; race/ethnicity; income; comorbidity; tumor factors; primary and adjuvant cancer treatments; previous BP use, and study site. Results: Over a mean (SD) follow-up of 6.4 (6.3) years, 586 women (3.4%) were diagnosed with CBC after their initial breast cancer diagnosis. Around 19.2% (n = 3303) of the cohort used a BP post-initial breast cancer diagnosis (>93% alendronate). To minimize confounding by indication, all models were adjusted for prior use of BPs before breast cancer diagnosis (n = 1047, 6.1%) and excluded women with prior 5-year history of osteopenia, osteoporosis, and fractures (n = 1808, 10.5%). Compared to non-users, ever use (>90 days supply) was associated with a modestly lower CBC risk (HR = 0.80; 95% CI: 0.61, 1.16). However, stratified analyses by age at breast cancer diagnosis (<50 years vs. ≥50 years) suggested lower risk among older women (HR = 0.80; 95% CI: 0.57, 1.12) compared with younger women (HR = 1.00; 95% CI: 0.39, 2.52). Increasing duration of use was not associated with CBC risk compared to non-use and possibly reflective of some residual confounding by indication: <1 year (HR = 0.56; 95% CI: 0.30, 1.02), 1-<2 years (HR = 0.99; 95% CI: 0.56, 1.72), ≥2 years (HR = 0.91; 95% CI: 0.58, 1.42). Discussion: While we found modestly lower CBC risk with BP use after diagnosis of primary breast cancer, the protective effects were perhaps confined to older women compared with younger women. Next steps include adding BMD to further explore confounding by indication and examining BP use and breast cancer recurrence and mortality. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P3-07-05.