Abstract 15793: Joint Prediction of Hard Ascvd by Pooled Cohorts Equations Risk and Presence of Breast Arterial Calcification: The Minerva Study

Abstract

Introduction: Prior studies have documented independent associations of breast arterial calcification (BAC) detected on mammographic screening with coronary artery disease or prospective cardiovascular disease outcomes. This study aims to ascertain the additional predictive ability for hard atherosclerotic cardiovascular disease (ASCVD) imparted by presence of BAC on top of the 10-year AHA/ACC Pooled Cohort Equations (PCE) Risk estimator. Hypothesis: BAC presence on mammograms will help further risk stratify women for risk of ASCVD. Methods: We used data from MINERVA, a multiethnic cohort of women aged 60 to 79 years at baseline (2012-15), free of symptomatic CVD, and recruited after mammography screening at Kaiser Permanente of Northern California. The sample available for analyses with complete data on BAC and PCE was 5,046. A BAC continuous mass score (mg) was obtained from digital mammograms using a validated densitometry method. BAC presence was BAC score > 0 mg. Follow-up for hospitalization and death was through 5/31/20 (mean, 6.2 years). We defined hard ASCVD as acute myocardial infarction or stroke or CVD death (n=122 events). Results: The mean (SD) age at baseline was 66 (4) years; 47 percent were non-white. Overall, 26.4 percent had a BAC score > 0 mg, 33 percent had low (< 5%), 36 percent intermediate (between 5 and 10%) and 31 percent high (> 10%) PCE risk. Using Cox models, there was a graded association by level of PCE risk and, within each level of PCE risk, women with BAC presence were at higher risk for ASCVD than women without BAC. In models adjusting for age, race/ethnicity, education level, BMI, statin use, menopausal hormone therapy, breast feeding and parity, being in the intermediate PCE risk group (relative to being at low PCE risk and having no BAC) was associated with 3.1 (95% CI, 1.5 - 6.5) increased hazard of ASCVD when no BAC was present and with 3.8 (95% CI, 1.7 - 8.7) increased hazard of ASCVD when BAC was present. Corresponding hazard ratios for women in the high PCE risks were 4.3 (95% CI, 1.9-9.6) and 7.5 (95% CI, 3.1-17.8), respectively. Conclusions: Our findings support additional predictive utility of BAC for ASCVD. In women at intermediate or high risk for ASCVD, BAC presence may help guide primary prevention.