When is a genetic test suitable for prime time? Predicting the risk of prostate cancer as a case-example.

Advances in genomics promise to deliver personalized medicine both for prevention and treatment of disease. Considerable effort is being directed towards translating observed associations between various genetic variants, such as single nucleotide polymorphisms (SNPs), and disease or drug response into clinically useful genetic tests. Unfortunately, because reported associations are usually weak or moderate, tests based on them are generally not accurate enough for use in routine clinical practice, and therefore, ensuring the appropriate use of genetic tests is important. In a recent report, a combination of 5 SNPs was claimed to improve the predictive value of the test for prostate cancer, compared with the individual SNPs. This led the authors to suggest that a 5-SNP-test could be used to predict the risk of prostate cancer. We evaluate the characteristics of the proposed test, comment on it, and summarize the views of others on its potential clinical utility. We hope that this may serve as a case-example for the evaluation of the many new genetic tests being suggested for adoption.