Abstract 3286: Breast cancer risk factors and survival by tumor subtypes: A pooled analysis from the breast cancer association consortium studies

Abstract

Abstract Background: Breast cancer (BC) is a heterogeneous disease with differing risk factors, prognosis and response to treatment strategies. Lifestyle and personal factors have been shown to influence BC prognosis. It is unknown whether lifestyle factors are differentially associated with prognosis according to BC subtypes. Methods: Analyses were based on 102,823 BC patients from 58 studies participating in the Breast Cancer Association Consortium. Cox regression models were used to assess associations between prediagnosis lifestyle and reproductive BC risk factors—including age at menarche, age at first full-term pregnancy, parity, breastfeeding, time since last full-term birth, BMI, oral contraceptive (OC) use, use of estrogen and progesterone therapy (EPT) and cigarette smoking- and 10-year overall survival (OS) and BC survival in all patients and by tumor subtypes. We used two tumor classifications: by estrogen receptor (ER) status; and by invasive subtypes based on ER, progesterone receptor, the human epidermal growth factor receptor 2 (HER2), and grade. Multiple imputation of risk factors and clinico-pathological variables was performed simultaneously. All analyses were stratified by study and adjusted for age at diagnosis, tumor size, nodal status, tumor grade (except for the luminal B HER2- subgroup), and systemic treatment. Multiple testing was accounted for using Bonferroni correction. Results: There were a total of 13,953 total deaths and 6,893 BC deaths over 10 years. The most significant associations were observed between OS and the following risk factors: parous vs nulliparous (HR (95%CI): 0.84 (0.78,0.89), P=1.4E-07); OC use vs never use (0.82 (0.78,0.87), P=1.2E-11); EPT use vs never use (0.64 (0.58,0.70), P=1.5E-19); BMI > 30 kg/m2 vs 18.5-25.0 kg/m2 in postmenopausal women (1.25 (1.17,1.34), P=1.9E-10 and 1.34 (1.18, 1.53), P=4.6E-06 in premenopausal women); cigarette smoking vs nonsmokers (1.37 (1.28,1.47), P=2.1E-19); time since last full-term birth < 5 years vs ≥ 10 years (2.09 (1.78, 2.44), P=1.3E-19). Associations with BC survival were similar to those for OS but weaker. We observed no differential associations by tumor subtype. The strongest evidence of subtype differential effects was for time since last full-term birth < 5 years vs ≥ 10 years for luminal A (3.00 (2.25,3.98)) vs luminal B-HER2- (1.69 (1.27,2.24)), HER2+ enriched (1.40 (0.96,2.05)), and triple negative (1.49 (1.15,1.93)), respectively. Multivariable models are being constructed to account for the interplay between risk factors. Conclusion: Our results suggest that BC risk factors influence survival after diagnosis. The effect appears to be stronger when OS is considered as endpoint, which may be due to power or associations of BC risk factors with other diseases. The data did not support a strong differential effect of BC risk factors on prognosis by tumor subtype. Citation Format: Anna Morra, Audrey Y. Jung, Sabine Behrens, Rose Yang, Heather Eliassen, Michelle Holmes, Montserrat Garcia-Closas, Marjanka K. Schmidt, Jenny Chang-Claude, on behalf of the Breast Cancer Association Consortium. Breast cancer risk factors and survival by tumor subtypes: A pooled analysis from the breast cancer association consortium studies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3286.