NEW YORK (GenomeWeb News) – The data generated by genome-wide association studies so far is more relevant for understanding disease biology than for predicting individual disease risk, according to a new study online today in PLoS Genetics.
Researchers from the University of Pittsburgh and the University of California at Los Angeles used a combination of real data and theoretical examples to evaluate hypothetical genetic tests based on strongly associated SNPs for several conditions. Based on their analyses, the researchers concluded that SNPs identified in GWAS, while informative in terms of disease etiology, cannot yet classify individuals well enough for genetic testing purposes.
"The rapid discovery of new genetic risk factors is giving us vitally important insights into human health, but a strong association between these factors and disease risk may not reliably predict which health issues a specific individual will face in the future," senior author Daniel Weeks, a genetics and biostatistics researcher at the University of Pittsburgh's Graduate School of Public Health, said in a statement.
"[T]hough we can paint a picture of our genetic makeup with current tests, this may not be enough to help us understand our individual risk for disease," he added.
Personalized medicine based on GWA studies relies on accurately translating information about disease risk conferred by genetic variants in the population into individual risk estimates. Such information is intended to predict disease risk, promote better health decisions and, in some cases, guide treatment.
But because many conditions are linked to a limited set of genetic variants that increase disease risk across an entire population, Weeks and his colleagues noted, using GWAS data for genetic testing is not always straightforward.
"The hope is that genetic testing will benefit patients and their families and encourage positive lifestyle changes and guide clinical decisions," they wrote. "However, for many complex diseases it is arguable whether the era of genomics in personalized medicine is here yet."
The researchers attempted to test the clinical validity of tests for several common conditions based on two statistical methods — the risk-based logistical regression method and the classification-based receiver operating characteristic curve analysis method.
Using data from their own lab and published GWA studies, the researchers looked at how well relatively strongly associated SNPs classified individuals’ risk of age-related macular degeneration, type 2 diabetes, prostate cancer, cardiovascular disease, and Crohn’s disease.
In the case of cardiovascular disease, the researchers noted that nine highly significant risk SNPs associated with low- or high-density lipoprotein cholesterol levels did not significantly improve the power of predicting cardiovascular events over a model based on traditional risk factors alone.
Similarly, when the team looked at 12 SNPs related to type 2 diabetes, five SNPs linked to Crohn’s disease, and two SNPs associated with prostate cancer, their analyses indicated that the SNPs lack the power necessary for screening or prognostic applications. Last fall, two other research teams also highlighted the importance of traditional risk factors for predicting type 2 diabetes risk.
The three SNPs used to assess macular degeneration fared slightly better in the team's analysis, but would still be expected to misclassify many individuals.
Based on such results, the authors suggested that scientists will have to find many more genetic variants before they can develop SNP-based genetic tests for common diseases.
"The scientific community should be very cautious to avoid over-hyping association findings in terms of their 'personalized medicine' value before their time, lest we lose the goodwill and support of the general public," the authors warned.
The team conceded that companies currently offering disease risk estimates based on SNP data "make it clear that they are not a clinical service and that their calculations are not intended for diagnosis nor prognosis purposes" and "typically advise their clients to consult their health care provider for more information." Even so, they argue that "few doctors currently have enough genetics training to actually make sense of the risk calculations now commercially offered."
Weeks and his co-workers stressed that their results do not negate the importance and findings of GWA studies. Instead, they emphasized the ability of such studies to increase researchers' understanding of diseases and drive new research hypotheses.
"With more study, our hope is that genetic testing will benefit people and encourage positive lifestyle changes and guide clinical decisions," Weeks said. "In the meantime, we need to take a step back and proceed with caution and allow the insights gained from these new association findings to be used to explore the basic biological causes of disease."