Background: Coronary artery calcification (CAC) is a measure of subclinical atherosclerosis that strongly predicts future clinical coronary artery disease events. Previous genome wide association studies have identified 3 loci associated with CAC. We aimed to use the expanded genomic coverage provided by whole genome sequencing data generated by the National Heart, Lung, and Blood Institute (NHLBI) Trans-Omics for Precision Medicine (TOPMed) program to explore genetic associations with CAC.

Methods: The study population consisted of 15,891 participants with whole genome sequencing and CAC data in TOPMed Freeze 5b, including 8,919 European, 6,044 African, and 928 Hispanic ancestry participants. Phenotype harmonization was performed by the TOPMed data coordinating center. log(CAC+1) was further transformed within each study-ancestry subset by 1) creating residuals after adjusting for age and sex differences, 2) performing inverse-normal transformation, and 3) rescaling the inverse-normal transformed residuals by multiplying each value by the subset specific standard deviation. We performed single variant as well as gene-based aggregate analysis on ENCORE using EPACTS. Significant associations were explored using OASIS.

Results: In single variant analyses, variants at six loci were significantly associated with CAC (P<5E-8), including 3 known loci (9p21, PHACTR1, and APOE), and 3 novel loci (ARSE, GRM7, and SYNPO2). Index variants at all three novel loci were rare (MAF<1%) in at least one ancestry group: rs149951959 at GRM7 was only found in participants of African ancestry, while variants at ARSE and SYNPO2 were present across ancestry groups but rare in European and African ancestry participants, respectively. No gene-based aggregation units were associated with CAC (P<1.7E-6).

Conclusion: Using whole genome sequencing we identified three novel loci for CAC that were missed by earlier genome-wide association studies. In future analyses we will incorporate data on further participants in TOPMed Freeze 6. We will also expand gene-based aggregate analysis to include regulatory variants.