Endocrine therapy-associated cardiovascular phenotypes in breast cancer survivors: a retrospective analysis using the TriNetX database

Aromatase inhibitors (AIs) and tamoxifen (TAM) are standard adjuvant endocrine therapies for postmenopausal women with hormone receptor–positive breast cancer. Although prior studies suggest a higher overall cardiovascular risk with AIs, the specific cardiovascular phenotypes driving this risk in real-world breast cancer survivors remain incompletely characterized. We conducted a retrospective cohort study using the TriNetX Global Collaborative Network. Women aged ≥ 50 years with stage I–III breast cancer who newly initiated AIs or TAM between 2010 and 2024 were included. A 1:1 propensity score–matched new-user design with a 6-month lag period was applied. Cardiovascular outcomes were grouped a priori into three phenotype-specific composites: atherosclerotic/vascular disease, myocardial dysfunction/electrical instability, and thromboembolic disease. Inflammatory cardiac disorders were evaluated separately as exploratory outcomes. Hazard ratios (HRs) were estimated using Cox proportional hazards models. After matching, each treatment group included 33,974 patients. During up to 5 years of follow-up, AI users had a higher cumulative incidence of composite CVD outcome than TAM users (19.81% vs. 18.38%; log-rank p = 0.0058) and a modestly higher hazard of composite CVD outcome (HR 1.057; 95% CI, 1.016–1.099). Higher hazards were observed for the atherosclerotic/vascular phenotype (HR 1.072; 95% CI, 1.007–1.141), cerebrovascular events (HR 1.100; 95% CI, 1.017–1.190), non-ischemic cardiomyopathy (HR 1.239; 95% CI, 1.034–1.486), and exploratory inflammatory cardiac disorders (HR 1.280; 95% CI, 1.047–1.565). No significant differences were observed for the myocardial dysfunction/electrical instability composite or thromboembolic outcomes. No significant effect modification was identified across prespecified subgroups. AI use was associated with a modestly higher hazard of composite CVD outcome than TAM, with differences most evident for cerebrovascular events and non-ischemic cardiomyopathy. Most other cardiovascular outcomes did not differ significantly. These findings support phenotype-informed cardiovascular risk assessment while warranting cautious interpretation of secondary outcomes.