Commercial DNA Testing May Cause Harm, Scientific Studies Show (if You Read Through to the Study Limitations at the End)

A year after the US Food and Drug Administration urged 23andMe to stop the marketing of its personal genome test in the United States, the world's largest direct-to-consumer genetics company relaunched the test in Canada and the United Kingdom.

"Find out how your genetics relate to things like abnormal blood clotting, cystic fibrosis or response to certain medications." [italics added] "Keep in mind that many conditions and traits are influenced by multiple factors. Our reports are intended for informational purposes only and do not diagnose disease or illness."

The UK website of 23andMe opens with clarifying statements about how the company wants us to see their renewed test. "The UK Medicines and Healthcare Products Regulatory Agency ... has determined that we're not a medical product, more an information product," says 23andMe's CEO Anne Wojcicki in The Guardian.

As it seems, 23andMe has relaunched its DNA test as the information service it was in the beginning: a service that offers a personalized exploration about what science knows about your genes, how they relate to your health and what they tell about your characteristics and traits. Yet, the test itself has not been scaled back to what it was in the early days.

The test now includes highly predictive variants such as BRCA1/2 genes mutations for breast and ovarian cancer, and variants in the APOE gene for Alzheimer's disease and the LRKK2 gene for Parkinson disease.

Clinical geneticists argue already for decades that testing highly predictive variants should be accompanied by adequate pre-and post-test information and counseling to help individuals understand and deal with the information. Testing such highly predictive gene variants directly to consumers may cause distress and anxiety when the initial fascination with the broader DNA test is gone.

The potential of harm from direct-to-consumer genetic testing has been topic of debate and research since the launch of the first tests a decade ago. Without major exceptions, the scientific studies show no evidence of harm. Evidently, this is grist to the mill of advocates of direct-to-consumer access to genetic testing, and a major argument against the intervention of the FDA.

Yet, "no evidence of harm" is not the same as "evidence of no harm." All studies had serious limitations that may have made the results more positive than justified, which the researchers often acknowledged. Let me illustrate this for the three studies that are most often cited to support the claim that direct-to-consumer DNA testing has no harmful consequences.

The largest study so far, in over 2,000 customers, investigated how people react to receiving not-so-predictive risk estimates for 22 common diseases. The study showed that the testing "did not result in any measurable short-term changes in psychological health, diet or exercise behavior, or use of screening tests." Yet, changes may have been absent because the participants were higher-educated "early adopters", who may better understand the limited impact that these genes have on the development of common diseases. The researchers emphasized that their "conclusions may not be applicable to other groups of individuals should the market expand, or to the general population."

The second study investigated how people respond when they receive APOE results, which are highly predictive for the risk of Alzheimer disease. Also this study found no evidence for harmful effects, which might be explained by the fact that the participants were children of patients with Alzheimer disease. They were all were very familiar with the illness and had received extensive counseling to explain the results. The researchers warn in their article that "if APOE genotyping had been provided without genetic counseling or to subjects who had no family history of Alzheimer's disease, the results might have been different."

And last year, researchers of 23andMe published the results of their interview study among customers who tested positive for BRCA1/2 gene mutations. Their article, entitled "Dealing with the unexpected: consumer responses to direct-access BRCA mutation testing" reported no "evidence for serious emotional distress or inappropriate actions," but the test results may not have been as unexpected as the title suggests. Most interviewed mutation carriers were aware of their possible genetic predisposition: two-third had a first or second degree relative with breast or ovarian cancer and 22% even knew already that they carried a BRCA1/2 mutation.

The three studies also shared another crucial limitation: 44 percent of the early-adopters declined the invitation to participate, 46 percent of the children of Alzheimer patients agreed to participate but dropped out before the start of the study, and 76 percent of the BRCA1/2 mutation carriers did not want to be interviewed. The authors of the first study comment that "[people] who might have been harmed psychologically by testing may have declined to participate or may have dropped out of the study." Hence, those who did participate may have been those who thought themselves able to deal with the test results, in all three studies.

The absence of harm is not at all scientifically proven; on the contrary, the limitations of the study suggest that harmful consequences may have been hidden by studying selected populations and because of low participation.

When tests cover highly predictive variants, such as those in the BRCA1/2, APOE and LRKK2 genes, companies should anticipate that customers may not handle the test results well.

To prevent unwanted surprises and anticipate adverse consequences, customers need to know what they are buying. To well serve their customers, companies must provide "information that is understandable, timely, verifiable, accurate, complete, truthful and not misleading" and access to counseling, not just when customers access their test results, but also when they consider ordering the test.

Whether a medical test in health care or an online information product, a responsible service of genetic testing for highly predictive variants should discourage people from taking the test if they feel not able to deal with the results.

This commentary is based on "The hidden harm behind the return of results from personal genome services: a need for rigorous and responsible evaluation", published by the author in Genetics in Medicine (online 20 November 2014).