Skip to content

Large, International Study Identifies Genes and Causal Mechanisms for Coronary Artery Disease

Key findings

  • In a discovery analysis involving 210,841 cases of coronary artery disease (CAD) among 1,378,143 participants, the CARDIoGRAMplusC4D Consortium identified 279 genetic variants associated with CAD, including 92 reported for the first time
  • Notably, 38 of the new variants were identified when the team incorporated genome-wide association study results based on only 29,000 CAD cases from Biobank Japan, suggesting multi-ancestry analyses will enhance the yield of genetic discovery for CAD
  • Using an integrative framework for determining causal genes that incorporates eight complementary lines of evidence, the researchers identified 123 candidate genes that had three or more concordant predictors
  • As a demonstration, the team successfully used CRISPR-Cas9 genome editing and cell-based experiments to verify that one of the putative causal genes does mediate CAD risk

More than 200 genetic loci have been statistically associated with coronary artery disease (CAD) susceptibility. However, translating these associations into potential therapeutic targets requires pinpointing causal genes and mechanisms, and that work is lagging far behind.

Now, an international collaboration, the CARDIoGRAMplusC4D Consortium, has demonstrated a powerful new method for identifying causal genes for CAD and potentially other diseases.

Krishna G. Aragam, MD, MS, a physician–investigator in the Cardiovascular Research Center and Center for Genomic Medicine at Massachusetts General Hospital, served as lead author of the group's report in Nature Genetics.

Large GWAS Meta-analysis

The researchers pursued a genome-wide association study (GWAS) and meta-analysis of data on 1,165,690 participants—including 181,522 participants with CAD—from the CARDIoGRAMplusC4D Consortium, the UK Biobank, and nine other studies from around the world. More than 95% of participants were of European ancestry.

241 variants proved to be significantly associated with risk of CAD at genome-wide significance, including 54 newly identified variants. However, only four of the 54 new associations strongly affected CAD (OR per CAD risk allele, 1.30–1.44). It's possible that few, if any, common genetic variants with substantial effects on CAD risk remain to be found by studying principally European populations.

A genome-wide polygenic risk score (PRS) derived from the GWAS improved the ability to predict the risk of CAD compared with a widely used PRS previously derived at Mass General from about 61,000 CAD cases. Similar to the issue of gene discovery, though, further increases in European-ancestry GWAS sample size may only modestly improve the predictive ability of the PRS.

Cross-Ancestry Analysis

The researchers combined the meta-analysis with summary statistics from a recent GWAS of 29,319 CAD cases and 183,134 controls from Biobank Japan. 199 of the 241 variants from the primary meta-analysis were also found in this analysis, along with 38 new variants reaching genome-wide significance.

Exploring Causation

Moving beyond association studies, the researchers set out to determine what genes might cause CAD. They systematically applied eight complementary prediction methods to all 279 variants (the 241 identified in the primary meta-analysis and the 38 found in the cross-ancestry analysis).

220 genes were pinpointed as candidate causal loci because two or more predictors supported them. 123 of those genes were supported by three or more predictors.

Sample Validation

As a demonstration, the team chose one newly-identified candidate gene, MYO9B, and explored whether it has a mechanistic relationship to CAD. MYO9B is contained within a vascular tissue enhancer, and using CRISPR-Cas9 to edit out the enhancer resulted in reduced expression of not just MYO9B but also HAUS8.

Aortic endothelial cells with a knockout of either MYO9B or HAUS8 exhibited impaired wound healing, suggesting the regulatory effect of the enhancer contributes to CAD risk through impaired wound healing in endothelial cells.

The integrative approach pioneered in this study for prioritizing likely causal variants, genes, and biological pathways will likely provide testable hypotheses for experimental follow-up in other diseases researched with GWAS.

Learn more about the Cardiovascular Disease Prevention Center

Refer a patient to the Corrigan Minehan Heart Center


Current guideline-based strategies for estimating a patient's risk of coronary artery disease do not account fully for inherited susceptibility.


Akl C. Fahed, MD, MPH, Amit V. Khera, MD, MSc, and colleagues found that carriers of familial hypercholesterolemia gene variants who followed healthy lifestyle practices had minimal change in LDL cholesterol yet were at 86% less risk of coronary artery disease than carriers whose lifestyle was unhealthy.