Value of Coronary Artery Calcium Scanning in Association With the Net Benefit of Aspirin in Primary Prevention of Atherosclerotic Cardiovascular Disease

Abstract

Importance: Higher coronary artery calcium (CAC) identifies individuals at increased atherosclerotic cardiovascular disease (ASCVD) risk. Whether it can also identify individuals likely to derive net benefit from aspirin therapy is unclear.

Objective: To examine the association between CAC, bleeding, and ASCVD and explore the net estimated effect of aspirin at different CAC thresholds.

Design, setting, and participants: Prospective population-based cohort study of Dallas Heart Study participants, free from ASCVD and not taking aspirin at baseline. Data were analyzed between February 1, 2020, and July 15, 2020.

Exposures: Coronary artery calcium score in the following categories: 0, 1-99, and 100 or higher.

Main outcomes and measures: Major bleeding and ASCVD events were identified from International Statistical Classification of Diseases and Related Health Problems, Ninth Revision codes. Meta-analysis-derived aspirin effect estimates were applied to observed ASCVD and bleeding rates to model the net effect of aspirin at different CAC thresholds.

Results: A total of 2191 participants (mean [SD], age 44 [9.1] years, 1247 women [57%], and 1039 black individuals [47%]) had 116 major bleeding and 123 ASCVD events over a median follow-up of 12.2 years. Higher CAC categories (CAC 1-99 and ≥100 vs CAC 0) were associated with both ASCVD and bleeding events (hazard ratio [HR], 1.6; 95% CI, 1.1-2.4; HR, 2.6; 95% CI, 1.5-4.3; HR, 4.8; 95% CI, 2.8-8.2; P < .001; HR, 5.3; 95% CI, 3.6-7.9; P < .001), but the association between CAC and bleeding was attenuated after multivariable adjustment. Applying meta-analysis estimates, irrespective of CAC, aspirin use was estimated to result in net harm in individuals at low (<5%) and intermediate (5%-20%) 10-year ASCVD risk and net benefit in those at high (≥20%) ASCVD risk. Among individuals at lower bleeding risk, a CAC score of at least 100 identified individuals who would experience net benefit, but only in those at borderline or higher (≥5%) 10-year ASCVD risk. In individuals at higher bleeding risk, there would be net harm from aspirin irrespective of CAC and ASCVD risk.

Conclusions and relevance: Higher CAC is associated with both ASCVD and bleeding events, with a stronger association with ASCVD. A high CAC score identifies individuals estimated to derive net benefit from primary prevention aspirin therapy from those who would not, but only in the setting of lower bleeding risk and estimated ASCVD risk that is not low.

Figure 1.. Cumulative Incidence of Bleeding Events and Atherosclerotic Cardiovascular Disease (ASCVD) Events Stratified by Coronary Artery Calcium (CAC)

Figure 2.. Estimated Increase in Bleeding Event Rate vs Reduction in Atherosclerotic Cardiovascular Disease (ASCVD) Event Rate With Aspirin Use For Primary Prevention In Dallas Heart Study (DHS) Participants Stratified by ASCVD Risk and Coronary Artery Calcium (CAC)

Estimates were generated by applying an 11% relative risk reduction to 10-year ASCVD event rates and a 43% relative risk increase to 10-year bleeding event rates. The number of individuals in each subgroup were as follows: all (N = 2191), pooled cohort equation (PCE) less than 5% (n = 1063), PCE 5% to 20% (n = 967), and PCE at least 20% (n = 161).

Figure 3.. Estimated Increase in Bleeding Event Rate vs Reduction in Atherosclerotic Cardiovascular Disease (ASCVD) Event Rate With Aspirin Use for Primary Prevention in Dallas Heart Study (DHS) Participants at Lower and Higher Bleeding Risk Stratified by ASCVD Risk and Coronary Artery Calcium (CAC)

Estimates were generated by applying an 11% relative risk reduction to 10-year ASCVD event rates and a 43% relative risk increase to 10-year bleeding event rates. The low bleeding risk group excluded individuals that reported the use of nonsteroidal anti-inflammatory drugs (NSAIDs), corticosteroids, anticoagulants, selective serotonin reuptake inhibitors (SSRIs), statins, protein pump inhibitor (PPI)/antacids, antiplatelet agents, peptic ulcer disease (PUD), cirrhosis, cancer, chronic kidney disease (CKD), thrombocytopenia, and uncontrolled blood pressure. The number of individuals in each subgroup were as follows: lower bleeding risk: all (n = 1430), pooled cohort equation (PCE) less than 5% (n = 1050), PCE 5% to 20% (n = 294), PCE at least 20% (n = 25); higher bleeding risk: all (n = 761), pooled cohort equation (PCE) less than 5% (n = 497), PCE 5% to 20% (n = 214), and PCE at least 20% (n = 36).