Column

Sequencing for Healthy People

We’re addressing the controversy using rigorous scientific study.

By Robert Green featured image Illustration by Lauren Crow

The genomic era is upon us and there are increasingly more options for consumers to have their own DNA tested. For some this is through healthcare providers, and for others it will be in the form of direct-to-consumer (DTC) genetic testing using array technology that examines markers for common conditions. We may have friends or family members who share their experiences with DTC genetic testing, perhaps exclaiming, “The test says I am going to get adult-onset diabetes!” But when we dig deeper, we realize that even the highest diabetes risk factors identified by the DTC tests only increase someone’s lifetime risk from 18 to 20 percent. Information like this may be interesting, and even useful, but it doesn’t usually tell the whole story, and it rarely involves a medical procedure.

On the other hand, now that genome sequencing is commercially available, you might have healthy friends nervously tell you that they received their results, revealing they are at risk for developing a fatal heart rhythm or a rare form of cancer, and that their chances of developing that disease are as high as 70 percent. The medical response to this information could be long-term surveillance with blood tests, X-rays, cardiac studies, or even prophylactic surgery.

While this comparison is a bit dramatic, it’s a very real illustration of how the introduction of genomic testing through DTC companies, which began in 2007, has primarily included mildly increased risks for common complex conditions, whereas today’s genome sequencing can reveal variants that have a much more significant impact on an individual’s future health. The stakes have definitely increased.

The medical community has embraced genome sequencing for individuals with undiagnosed conditions, those undergoing cancer treatment, and prospective parents with family histories of recessive disease who seek pre-conception screening. For these groups, there seems to be agreement that genome sequencing will identify information that can be used productively in medical decision-making and care. But as genome sequencing is becoming an affordable commodity, we are hearing more and more prominent voices making the argument that genomic test results contain the key to future health. They maintain that sequencing has the ability to detect serious disease risk before that disease occurs and can prompt healthcare interventions that lead to better health outcomes. It’s a compelling argument, so why does it arouse so much scientific controversy?

While these pilot results are far from definitive, we have begun laying the evidence-based path forward that can answer the question of whether it is medically and economically responsible to sequence healthy individuals for disease prediction and prevention.

There is a one-word answer to this question: penetrance. The penetrance of a particular variant is the probability that the individual carrying the genetic risk variant will actually develop the condition. And the reason for the controversy is that we don’t fully understand the penetrance of most single-gene disease variants, particularly when they are discovered in people without a family history of the condition. What if we identify thousands of individuals as “at risk” who will, in fact, never get cancer or who will never have a heart problem, prompting unnecessary concern, unnecessary medical evaluations, and even unnecessary surgeries?

Although clinical sequencing for healthy people has not yet been widely adopted, several well-established physician-facing molecular laboratories, as well as a few consumer-facing companies, are refocusing their products to provide this testing. As more healthy people respond to this marketing and ask their doctors about genome sequencing, and as more doctors offer genome sequencing as an option to their patients, many in the genetics and medical communities have concerns about how it will all play out.

Will doctors know what to do with this information? What if the prospect of heightened risk causes some patients to seek preventive measures that will do more harm than good? What happens when individuals learn that they are at increased risk of developing a disease that does not have a prevention or cure? Could this be helpful or just cause anxiety for that person and their loved ones?

Then there are concerns about costs. The cost of genome sequencing and analysis continues to decline, but it is still expensive; follow-up testing, appointments, and preventive measures are not cheap and could create significant burdens for patients and the healthcare system. And these costs will remain even if the variants never show any evidence of causing disease.

Without actual evidence, all of these questions have been unresolved for years. But this is beginning to change.

Over the past decade, our Genomes2People Research Program, at Brigham and Women’s Hospital, the Broad Institute, and Harvard Medical School, has led the way in seeking observational and experimental data to better understand the use of genetic risk information. For example, penetrance is generally described in terms of specific genes or even specific detrimental variants within those genes. But to gather enough rare detrimental variants, you would need an extraordinary number of individuals, perhaps in the millions.

To begin addressing this problem, we recently proposed the concept of “aggregate penetrance,” a rougher measure of penetrance, in which we lump variants and genes together within categories in order to generate sufficient numbers to make inferences. In a paper recently published in Science Translational Medicine, we applied this concept to data from the Framingham Heart Study and Jackson Heart Study to show that if a few thousand people in several population-based cohorts were screened for dominantly inherited monogenic conditions and then followed for decades, a surprisingly high percentage of them would eventually develop the condition at risk. While our affected sample size was small, with fewer than 50 people who carried such pathogenic variants, the proportion of those people who developed the condition over time was higher than expected, demonstrating that in the aggregate, most people carrying detrimental variants would eventually develop the condition in question.

In another experiment, we asked whether these findings would necessarily result in confusion and economic waste. The MedSeq Project is a pilot study that enrolled healthy middle-aged patients and their primary care physicians and examined their experience receiving whole genome sequencing (WGS) results and the associated downstream effects on their medical care. Our analysis spanned nearly 5,000 genes for each patient, including many high-risk variants for rare diseases, a far deeper dive than most commercial laboratories currently offer, but perhaps what can be expected in a few years. Data from the MedSeq Project, published this summer in the Annals of Internal Medicine, suggested that confusion among patients and their primary care physicians could be mitigated with a few hours of training for the primary care providers, and that the medical costs after sequencing were only modestly higher than those in the arm who did not get sequenced.

A particularly surprising result from MedSeq was that 11 of 50 primary care participants (22 percent) who received WGS learned that they were carrying a genetic variant for a rare monogenic disease. Of these 11 participants, only two exhibited symptoms of the associated rare disease, and none of these symptoms were terribly serious: One had an eye disease that impaired night vision, and the other had a skin condition causing rashes and sun sensitivity. The other nine participants with disease variants showed no outward signs of those diseases — but if they do eventually appear, some of the conditions, such as heart rhythm abnormalities, could be life threatening. Results like these are at the center of the debate about the costs and benefits of genome sequencing for healthy people.

Although our results reflected only a small sample in this pilot study, they suggest that with some training and support, mainstream physicians may be able to manage genome sequencing results without becoming geneticists; that patients do quite well managing unanticipated risks; and that downstream health costs may not skyrocket in response to sequencing. While these pilot results are far from definitive, we have begun laying the evidence-based path forward that can answer the question of whether it is medically and economically responsible to sequence healthy individuals for disease prediction and prevention.

So, what’s next? Our research program is currently seeking funding to conduct a full-scale, randomized controlled trial of WGS with several thousand participants. We believe that conducting a rigorously designed study is the best way to resolve the controversies around the sequencing of healthy people and set the stage for integrating genomics into the everyday practice of medicine.