NEW YORK — Researchers have identified 95 novel loci associated with coronary artery disease (CAD) risk in a genetically diverse cohort from the Million Veteran Program.
CAD affects more than 20 million people in the US, and heart disease is a leading cause of death across racial and ethnic groups, according to the Centers for Disease Control and Prevention. While previous genome-wide association studies (GWAS) have identified 208 CAD susceptibly loci, those analyses have largely drawn on European ancestry groups. In a new study appearing in Nature Medicine on Monday, a team led by investigators at the VA Palo Alto Health Care System analyzed genetic risk of CAD in a more diverse population, boosting the number of identified risk loci by nearly 50 percent.
"Overall, our results underscore the pressing need to produce more data for non-white populations and develop more sophisticated analytic methods to eradicate such differences in performance and minimize the potential for exacerbating existing health disparities as PRSs are implemented in clinical practice," the VA's Themistocles Assimes and colleagues wrote in their paper.
For their analysis, the researchers drew upon 95,151 CAD cases and 197,287 controls from the MVP who were non-Hispanic white; 17,202 CAD cases and 59,507 controls who were non-Hispanic Black, and 6,378 cases and 24,270 controls who were Hispanic. Ninety percent of the cohort was male.
In three stratified GWASs encompassing autosomes and the X chromosome, the researchers replicated known loci associated with CAD. By folding in additional data from the largely European ancestry cohorts CARDIoGRAMplusC4D and the UK Biobank with the white MVP participants, the researchers then identified 33 novel loci reaching genome-wide significance for association with CAD.
Meanwhile, a multi-population meta-analysis that also included Black and Hispanic MVP participants and Japanese participants from the BioBank Japan revealed 62 novel loci, for a total 95 novel CAD-linked loci, which included nine located on the X chromosome.
The researchers also conducted a second GWAS of angiographically determined burden of CAD that identified 15 loci of genome-wide significance, all of which had previously been tied to clinical CAD.
The novel CAD risk loci were linked to at least one traditional risk factor for CAD, such as hypertension, diabetes, or smoking, the researchers found in a further phenome-wide association study. Additional gene and pathway-based analyses found that these SNPs were involved not only in known aspects of CAD biology but also in basic cellular processes like the cell cycle, division and replication, and growth.
Their analyses also hinted at a tie between atherosclerosis and oncogenesis as well as cancer biology, as the PheWas analysis indicated that about a third of the novel loci were linked to either cancer or height. BRCA1, they further noted, falls within one of their novel loci.
The researchers noticed, though, that a well-established CAD susceptibility locus at 9p21 did not reach genome-wide significance in Black or Hispanic study participants. By examining the ancestral origin of the high-risk haplotype block, they found that it is largely absent among populations of African origin and that its presence among Black and Hispanic participants is contingent upon the inheritance of a non-African haplotype in that genomic region. Based on this, the researchers noted that the 9p21 region is unlikely to serve as a risk stratifying locus in populations with a high portion of African ancestry.
The investigators additionally examined how well polygenic risk scores predicted CAD in the MVP cohort, finding that they had a reduced performance among Black MVP participants. A new PRS they developed improved risk prediction across populations but did not close the gap in performance between them.
"[O]ur large-scale multi-population GWAS provides important new insights into the genetic basis of CAD and brings us closer to precision medicine approaches for CAD across the diversity spectrum, but follow-up studies are needed to improve the transferability of PRSs for CAD, to identify and understand mechanisms of causal genes, and to develop cross-population and population-specific novel therapies based on this understanding," the researchers wrote.