Abstract P4-03-17: Parity, Use of Statins and Low-dose Aspirin, and Breast Cancer Risk – data from two large cohort studies
Julie A. Schmidt,Agnès Fournier,Manon Cairat,Aurélie Mailhac,Henrik Sørensen,Marc Gunter,Deirdre Cronin-Fenton
DOI: https://doi.org/10.1158/1538-7445.sabcs22-p4-03-17
IF: 11.2
2023-03-01
Cancer Research
Abstract:Abstract Background: Statins and low-dose aspirin have been associated with a reduced breast cancer (BC) incidence, but results are inconsistent. Based on emerging evidence that parity, a protective factor for breast cancer, and these drugs modulate immunity, we hypothesized that the association between drug use and breast cancer risk may differ by parity. Objectives: To assess the associations of statin and low-dose aspirin use with BC incidence according to parity in French and Danish cohorts. Methods: We conducted two cohort studies, using data from the French E3N study and a Danish nationwide population-based cohort, respectively. From E3N, 51,482 women, mean age 65.5 years, were enrolled in 2005 and followed until 2014. Data on parity (here full-term pregnancies), drug use, and incident BC were acquired from questionnaires, a drug reimbursement database, pathology verified self-reports and the national cause-of-death registry, respectively. From nationwide health registries, we included all Danish women free of BC and aged 45 years in 2000-2005 (n=198,575), with follow-up to 2014. Use of the exposure drugs were defined as at least two reimbursements/filled prescriptions. In both cohorts, multivariable-adjusted Cox regression was used to compute hazard ratios (HR) for drug exposure, treated as time-varying lagged variables, and BC risk stratified by parity (0, 1, 2, 3+). Results: In E3N and Denmark, 1,878 and 5,436 incident BC cases occurred in a mean follow-up of 8.5 and 11.5 years, respectively. At the end of follow-up, 35% and 19% of E3N and 16% and 8% of the Danish cohort had been exposed to statins and low-dose aspirin, respectively. In E3N, effect modification was observed between parity and statins, but not low-dose aspirin. For statins, the HRs for use vs no use were 1.29 (0.97-1.72), 1.27 (0.99-1.65), 1.08 (0.91-1.27), and 0.76 (0.61-0.95) among women with 0, 1, 2, and 3+ full-term pregnancies, respectively (p-het=0.005). The corresponding estimates for low-dose aspirin were: 1.17 (0.80-1.71), 0.81 (0.54-1.20), 1.13 (0.90-1.43), and 1.22 (0.93-1.61; p-het=0.6). In contrast, the Danish data did not suggest any effect modification. HRs for use vs no use of statins were 0.82 (0.60-1.13), 1.05 (0.80-1.37), 0.91 (0.75-1.10), and 1.09 (0.83-1.44) among women with parity of 0, 1, 2, and 3+, respectively (p-het=0.3). For low-dose aspirin HRs were: 1.09 (0.75-1.59), 0.79 (0.54-1.15), 1.04 (0.82-1.32), and 0.99 (0.71-1.40; p-het=0.8). Conclusions: We observed effect modification by parity for the association between statins and breast cancer risk in the French but not the Danish cohort. Whether the age difference between the cohorts explains the inconsistent results should be explored. Citation Format: Julie A. Schmidt, Agnès Fournier, Manon Cairat, Aurélie Mailhac, Henrik Sørensen, Marc Gunter, Deirdre Cronin-Fenton. Parity, Use of Statins and Low-dose Aspirin, and Breast Cancer Risk – data from two large cohort studies [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P4-03-17.
oncology