2 March 2010
An important aspect in the prevention of cardiovascular disease involves risk prediction in order to identify individuals at high risk and tailor treatment strategies. Risk prediction is usually based on knowledge of conventional risk factors (e.g. blood pressures, smoking, cholesterol levels) and family history. More recently, genome-wide association studies (GWAs) have led to the identification of many genetic markers associated with cardiovascular disease and it is hoped that inclusion of such information may help refine risk prediction models. However, risk prediction based on genetic factors is a complex process requiring knowledge based on the cumulative effects of multiple common risk variants (see previous news).
In a paper published in JAMA, Paynter et al. have conducted a literature-based selection of genetic markers associated with either cardiovascular disease or intermediate phenotypes (e.g. blood pressure, cholesterol) in order to construct two genetic risk scores [Paynter et al. (2010) JAMA 303(7):631-7]. The genetic markers were selected using the online catalogue from the NHGRI of GWA studies published between 2005 and 2009. 101 SNPs were used in the construction of the primary genetic risk score. The second score, limited to SNPs with a published association with incident cardiovascular disease, included 12 SNPs after pruning.
The predictive ability of these scores was tested using data available from the Women’s Genome Health Study (WGHS), an ongoing project to identify genetic variations in women underlying a range of serious illnesses, including heart disease, stroke, diabetes, breast cancer and osteoporosis. The study is surveying genetic differences among 28,000 initially healthy American women who have been tracked for more than a decade for the development of many common health disorders.
Information from a total of 19 313 participants was analysed after limiting for those of Caucasian background (to avoid population stratification) for whom complete data were available for both the traditional risk factors and for the genetic risk scores. In addition to assessing the predictive ability of genetic information alone, the authors also assessed the predictive ability of this information in combination with known cardiovascular disease risk factors and compared genetic information to information reported on family history.
Both genetic risk scores were associated with increased risk after adjustment for age, but when used alone neither had the ability to discriminate between those who were at increased risk and those who were not. In addition, neither genetic score remained associated with incident cardiovascular disease after adjustment for traditional risk factors whereas family history did. The authors state “Our study finds no clinical utility in a multi locus panel of SNPs for cardiovascular risk based on the best available literature”.
Comment: It has been widely suggested that findings from GWAs could be combined with population data in order to produce estimates of absolute risk for numerous common complex diseases (see previous news). This information could be used to stratify populations into risk categories in order to better target preventative interventions such as screening (see previous news). In addition to associations between genetic factors and risk, knowledge about the mechanisms by which these factors contribute to risk is also needed both for better stratification and, more importantly, for the development of novel treatments. The recent PHG Foundation report Predicting the risk of coronary heart disease with conventional, genetic and novel molecular biomarkers concluded that statistical considerations coupled with available empirical studies make it unlikely that the associations are strong enough to have an important general influence on the discriminatory ability of risk models at present.