Incorporating Genetics Into Drug Dosing and Risk Prediction - Ready for Prime Time?

Are we ready to roll out genetic testing to our cardiovascular disease patients?

One of the interesting sessions that I plan to attend at ACC 2010 is where the results of the Medco-Mayo Warfarin Effectiveness Study (MM-WES) trial are being announced.

This large study (>1000 subjects) has compared two warfarin dosing strategies with historical controls. In one arm, subjects have their warfarin dose adjusted along standard lines. In the genetic testing arm, patient genotype is considered and warfarin dosing proceeds according to this. It s known that polymorphisms in the CYP 2C9 gene are associated with more potent anticoagulation because of reduced warfarin metabolism. This can potentially mean more episodes of clinically important bleeding. This trial is the first to determine whether prior knowledge of a patient s CYP 2C9 allele (and another, related polymorphism of VKORC1) will improve thromboembolic and bleeding rates and thereby patient safety and healthcare costs, specifically over a 6 month period.

Another study that focussed on genetic testing for determining cardiovascular risk was recently published in JAMA by Paynter and colleagues ( The authors asked whether genetic testing added anything to cardiovascular risk prediction over and above simple scoring algorithms. Many genetic markers of risk have been identified over the last decade using GWAS techniques; what is unknown is the added benefit of their knowledge for risk prediction.

In over 19,000 women, 101 SNPs known to be associated with an increased risk of CV disease were measured. In this study, over the 12 year follow-up, genetic risk scoring did not improve either discrimination or reclassification compared with ATP-111 scoring. In fact, simple family history of cardiovascular disease had the most predictive power.

This study, and another one by the same group that examined a common SNP at chromosome 9p21 ( have been widely discussed, including by Heartwire genetic expert Dr Eric Topol ( The main issue seems to be that the JAMA study did not examine more SNPs, particularly those for Lp(a), or low frequency SNPs.

So what are your views? Is genetic testing either for directing drug therapy or to predict risk a little premature? Should we be smarter in our searches? Are low frequency SNPs the flecks of gold in the dark mysterious river of risk factors?