Genomic risk prediction based on polygenic risk scores generated from genome-wide association studies (GWAS) to summarize the contribution of anywhere between hundreds to millions of genetic variants to determine the risk of developing multifactorial diseases such as coronary artery disease (CAD). This could help physicians make personalized clinical decisions, but first such methods must be shown to enhance or be superior to traditional risk assessment tools such as Framingham, which includes factors such as cholesterol profile, blood pressure and smoking status.
A study led by Robert A. Hegele of Western University, London, Ontario, Canada, sought to assess the clinical utility of a coronary artery disease genetic risk score named FDR267 that they generated from the UK biobank dataset. The results were published in CJC Open, an official journal of the Canadian Cardiovascular Society. The present study applied the score to the ARIC longitudinal prospective cohort population of 15,792 European and African Americans aged 45 to 64 years recruited at 4 separate field centres in the United States between 1987 and 1989, with follow up until 2013.
There were 836 cardiac events during the follow-up period, producing a total of 1138 cases of CAD when including those with CAD at the start of the study. Adding FDR267 genetic risk score to the Framingham clinical data resulted in a marginal but significant increase in sensitivity (more cases predicted) for the European American dataset. For the African American dataset there was no benefit to adding FDR267, probably reflecting the demographics of the UK biobank.
Interestingly, a strong association of FDR267 with lipid phenotypes in the European sample was not observed in the African American sample, which perhaps suggests differing underlying mechanisms of CAD in the two populations.
People in the top quintile of the FDR267 genetic risk score have about 2x increased risk of CAD compared with the bottom quintile, which is in line with risk associated with self-reported family history of CAD. It is perhaps not surprising that applying a genetic risk score to those 45 and above does not greatly improve over prediction from clinical measures alone. It would be expected to be of greater benefit at an earlier age, when all clinical measurements are normal, as it could inform lifestyle adjustments.
The study is limited by the relatively low number of single nucleotide polymorphisms (SNPs) included in the genetic risk score (267). The authors note that genetic risk scores that incorporate more SNPs have performed better in other studies.
“Our genetic risk score was significantly associated with CAD and provided modest predictive utility for incident CAD. Despite comparable predictive power with family history and improved ability to discriminate prevalent cases of CAD when added to a model with traditional risk factors, FDR267 does not improve on risk assessment. The clinical use of FDR267 is further limited by its inconsistent assessment of risk in a non-European population,” concluded the authors.