In the news

Source: Pub Med

An expert US group has concluded there is insufficient evidence to support genomic profiling for cardiovascular disease (CVD) risk prediction.

The EGAPP (Evaluation of Genomic Applications in Practice and Prevention) Working Group is a US panel that assesses the evidence about new genetic tests and applications (see previous news). Prevention of CVD is a public health priority, and in recent years new genomic and proteomic biomarkers – including those emerging from discoveries of genetic variants through GWAS - have been evaluated for their ability to improve current prediction methods based on traditional risk factors. The latest recommendation statement by EGAPP reviews currently available evidence, concluding that there is “insufficient evidence to recommend testing for the 9p21 genetic variant or 57 other variants in 28 genes to assess risk of CVD in the general population” [EGAPP Working Group, Genet Med. 2010 Oct 29, Epub ahead of print].   The EGAPP group found that the analytic validity of available genomic profiling tests for individual gene variants was potentially satisfactory, where available, but there was inadequate evidence available. Similarly, evidence for clinical validity was inadequate for 59% of the gene variant/disease associations tested. Only the association between the 9p21 variant and risk of heart disease was reasonably strong, but the improvement in risk prediction from testing for this variant was found to be ‘at best, small’. Overall, there was no evidence of health benefit from using such genomic profiling tests singly or in combination, so the group recommended against use in clinical practice unless new evidence of benefit emerges. 

CVD is an enormous public health issue.With genomic tests for CVD being available to the population through clinicians as well as direct-to-consumer, there need to be well established recommendations for cardiac risk assessments for both physicians and patients; this is a useful contribution to the confusing picture.  A PHG Foundation report released earlier this year, Predicting the risk of coronary heart disease with conventional, genetic and novel molecular biomarkers, similarly concluded that genetic and other biomarkers are in fact ‘unlikely to improve substantially our ability to predict the risk of CHD in individuals’ in the near future. However, risk prediction is a highly complex area, and new work due out soon has been considering the issue of how to reliably assess and compare the impact of risk prediction models, to allow accurate evaluation of how new data – such as genomic profiles – do or do not improve outcomes.