Predicting drug trial outcomes

Choosing drug targets based on studies of genetic associations with disease could potentially double the success rate of new drug development, reports a study published online this week in Nature Genetics.

More than half of newly developed drugs fail during clinical trials because they do not have the therapeutic effects expected based on laboratory studies in mice and other animal models. Matthew Nelson and colleagues asked whether other sources of information could be used earlier in the drug development process to increase the rate of success. They looked to genetic association studies, in which particular genes or gene variants are found to be significantly associated with the risk of developing specific diseases. Genetic associations do not always indicate a causative relationship. However, they could potentially increase the chances of selecting a successful drug target when combined with traditional drug selection methods.

The authors created a catalog of gene?disease associations based on genetic studies and correlated this with information about drug trials. They found that successful drugs (those that were eventually approved for treatment of a disease) were more likely to target genes with evidence of association to that disease or a related trait. Based on this information, they estimated that using genetic association data applied to drug target selection could significantly boost the chances of success.

doi: 10.1038/ng.3314

Return to research highlights