Could the boom in personal genomics backfire?

|
Image Credit: John Goode Flickr (CC BY 2.0)

By now, you’ve no doubt heard about the advantages of personal genomics. It can tell us a lot about what mutations we carry that could lead to disease or impact behavior, although the data are complicated, difficult to interpret and sometimes disputed.

In a 2012 NPR survey, more than 81 percent of people said they would get their genomes sequenced if they could afford it. But the science behind genomic association studies which form the basis of personal genomics is changing, writes physician and contributor David Shaywitz in Forbes. And what we learn can sometimes be dangerous, because you may be getting vague, unhelpful or sometimes even false information from companies. What’s worse is that in the short term, this may cause the bubble of personalized genomics to burst, setting back the roll-out of truly revolutionary technologies.

The rapid development of ever-cheaper and more efficient gene sequencing and bioinformatics technologies has allowed many startups to build a profitable business around interpreting the genome. At the core of the services offered by these companies are analyses of variations in your genome that, presumably, mean something for you as an individual. They might offer to “read” your genome and interpret your risk for a disease such as diabetes; perhaps tell you how your body reacts to certain drugs, even the caffeine in coffee.

In 2013 however, a very public showdown between the federal government and the personal genomics industry occurred: the FDA issued a warning to Google-backed 23andMe, the poster child for the direct-to-consumer genomics revolution. The FDA stated that their services needed FDA approval. Since then the market has significantly cooled off, but many companies continue to function — especially outside the United States. Some in the US work around the FDA by offering these tests through physicians, an approach that does not require FDA approval and is based on the assumption that healthcare practitioners can reliably evaluate the efficacy and interpret the results of these tests.

Regardless of how they obtain the results, it is important that consumers know that the science behind genetic testing is still evolving; it is far from exact. The human genome is both vast and incredibly complex. Scientists have been able to identify in several cases individual genetic markers that are routinely and reliably used in genetic testing laboratories around the world. But these are more likely to represent ‘low hanging fruit’ as epidemiologist Cecile Janssens points out in The Conversation. Physician David Shaywitz who is also the Chief Medical Officer at a genomics data management company echoed this in the Forbes article:

“While it was originally hoped that the heritable component of most human disease would be accounted for by a relatively small number of relatively common mutations that each exerted a relatively large effect, it soon became clear that in many, perhaps most cases, disease seems to result from a relatively large number of mutations that each contribute only slightly to the overall heritability of the disease “

From a scientific standpoint, researchers do not always set out to identify genetic markers that can be immediately implemented in a diagnostic product. In many cases, they are just looking for clues that can add to a small piece of the puzzle. In fact, when trying to assess if genetic markers for common diseases are clinically significant, one cause of concern that scientists routinely acknowledge is the fact many studies are simply not statistically powerful enough to yield a clear cut answer. More work is almost inevitably needed.

Type 2 diabetes, for example, has known connections to many genetic variants that affect the risk of occurrence in small ways. To be able come up with a measure of risk that takes into account all these inherited factors, researchers predict that a gargantuan study with more than 250,000 people might be needed.  Furthermore, ethnic differences may limit the usefulness even of studies on a massive scale unless they are able to incorporate members of various human populations.

The case for heart disease illustrates how complex the situation is and how fast things are changing. After the human genome was sequenced, numerous genetic variants were identified by researchers that could potentially affect the risk for heart disease. In 2012, one editorial published in a peer-reviewed journal suggested that adding a genetic risk score did not make traditional heart disease risk predictions any better. A year later, a new editorial in a different journal pointed out that the science had advanced enough that a genetic risk score improved traditional methods by a small but significant measure. In the same breath, the authors also acknowledged that the improved genetic assessments, promising though they were, were not yet ready for the clinic.

The science itself is undoubtedly moving quickly, but in contrast to the circumspect attitude of the researchers above, entrepreneurs are eagerly trying to leverage every bit of progress to create a product or a service. You are being told – and will continue to be told — that whatever the newest start-up is hawking is going to help you know everything about yourself, transform your life and take control of your health. For someone who is looking to be actively involved what shapes their health and has some money to spare, this is undeniably an attractive proposal.

The allure, however, must be tempered. Not only for the safety of consumers but to ensure we achieve the promises of the genomics revolution in personalized medicine sooner rather than later.

Jason Kim, CEO of the personal genetics company PHENOMBio perfectly captured the risk over-hype and a lack of skepticism poses to personal genomics on the whole during an interview with GLP contributing writer Layla Katiraee.

“I fear that within the next few years this industry will become rife with modern-day ‘snake oil salesmen’ pushing outrageous claims and over promising on what their tests can do for customers with little to no scientific evidence backing their claims. Needless to say this will end up creating a huge setback in public perception for personalized genomics.”

Arvind Suresh is a science communicator and a former laboratory biologist. Follow him @suresh_arvind

  • vlev

    The confusion of terms “genomics” and “personalized medicine” is the real danger at this time.For many people genomics (as in genome sequencing) is the ultimate goal with no other information required for medical decision-making. This dangerous misunderstanding will become a problem in a couple of years, when another wave of GWAS projects will end up with ZERO results. We have to stop oversimplifying the issue: genome and its sequence can be compared to the blueprint of a building, while functioning of the real structure is determined by a myriad of other factors besides the blueprint. We also have to internalize the fact that disease is a malfunction of the organism with emphasis on FUNCTION, and concentrate on *functional* research rather than on studying the bleprint again and again and again.