An epidemiologic assessment of genomic profiling for measuring susceptibility to common diseases and targeting...

The current clinical value of genomic profiling (testing for genotypes at multiple loci) for assessing susceptibility to common diseases and targeting behavioral and medical interventions is questionable. As common diseases result from many gene-environment interactions, epidemiologic studies should be used to examine the value of genomic profiling in terms of clinical validity (future disease positive and negative predictive value stratified by exposure), clinical utility (targeted interventions to reduce disease risk among persons with the profile) and public health utility (comparing reduction of disease burden in the population based on genomic profiling to population-wide interventions).

We investigate these parameters for a hypothetical common disease (5% lifetime risk), for which 3 genetic variants at different loci and one environmental exposure are risk factors.

We show that even for modest effects of each variant alone (risk ratios from 1.5-3.0) and modest interactions between the exposure and the genes, the disease predictive value for people with 2 or more variants (especially 3) can be quite high (50-100%) in the presence of a modifiable exposure. Individual risks can then be reduced by targeted exposure intervention among persons with the genotype. However, the predictive value for multiple genotypes is much lower for rarer diseases (< 1 per 1000). Also, with increasing number of genes in a profile, the population impact of disease reduction for targeted intervention based on genotype will be smaller, especially for rare genotypes, weak associations, and weak interactions.

To assess the value of genomic profiling, well-designed epidemiologic studies are needed to quantify disease risks, in addition to costs, benefits, and risks for testing and interventions.