Michael Snyder never saw it coming. Healthy, trim, and active at age 54, the Stanford geneticist decided to undergo whole genome sequencing (WGS) in 2010, when the technology was just being established at his university. Snyder figured he’d be a “guinea pig” for WGS not only to satisfy his own curiosity, but because, at the time, “a lot of people were concerned about people’s ability to handle that information.”
What Snyder, director of the Stanford Center for Genomics and Personalized Medicine, couldn’t have known was that secrets revealed by his DNA would pose astonishing health challenges in his own life. Sequencing exposed gene mutations predisposing him to type 2 diabetes — a condition entirely at odds with his physique and lifestyle. This knowledge prompted him to diligently track his blood sugar levels for months afterward, revealing drastic glucose spikes in real time.
Careful dietary tweaks and an uptick in exercise have since helped him keep his diabetes largely in check. Snyder also shared the genetic information with his five siblings, four of whom underwent testing that indicated elevated glucose levels.
“It came out of nowhere for me,” says Snyder, now 60, whose sequencing results also have him on the alert for developing the skin cancer basal cell carcinoma. “I view getting whole genome sequencing as a more active way of getting your family history.”
Health revelations large and small are encoded in our genomes, but only recently has accessing these secrets even been possible. In the scant dozen years since the advent of next-generation sequencing technology, the cost of whole genome sequencing by researchers has plummeted from about $2.7 billion per person to about $1,000 today.
But only a fraction of the several hundred thousand people who have gotten all six billion letters in their DNA sequenced have actually received an interpretation of their whole genome sequencing. In addition, many of these people are scientific research participants. Unlike clinical efforts to help solve baffling birth defects in children or to sequence malignant tumors — which can steer doctors and patients to more effective treatments — deciphering the DNA of healthy people may hold more dubious value, despite its allure, experts say.
“Some healthy people may have a specific interest in their genome because of a family history of rare illnesses, while some are early adopters, and some are just interested in the ultimate selfie, so to speak,” says Muin Khoury, founding director of the Office of Public Health Genomics at the Centers for Disease Control and Prevention in Atlanta. “It’s a multifactorial proposition as to why.”
For starters, not all genome-decoding techniques are created equal, and comprehensiveness varies. Marked differences distinguish whole genome sequencing from whole exome sequencing. Whole genome sequencing examines every single letter in your genome in both the gene regions (which make up, surprisingly, only about 1.5 percent of the genome) and the non-gene regions (the remaining 98.5 percent). Whole exome sequencing examines just the gene regions. While variants, or mutations, in genes are responsible for most rare single-gene disorders such as cystic fibrosis, Huntington’s disease, and sickle cell disease, variants in the non-gene regions of the genome also may be responsible for some common diseases, such as heart disease, cancer, and diabetes, as well as traits like a person’s response to certain drugs. So, while whole exome sequencing examines what is considered the meatiest part of the genome, it doesn’t deliver quite as complete a picture as whole genome sequencing.
Beyond that, direct-to-consumer DNA tests reveal only a subset of variants in certain genes. Last October, for example, the DNA testing firm 23andMe began marketing a $199 test that could indicate whether someone carries genes linked with 36 disorders such as cystic fibrosis or the blood disorder beta-thalassemia, which causes chronic anemia. Such services don’t sequence a customer’s whole genome or exome, but instead perform so-called “targeted genotyping” of about 650,000 known genetic variants among the three to four million such mutations found in the average person’s genome.
“Most people don’t understand that there’s a difference between sequencing six billion letters in the genome versus spot-checking certain parts of the genome or certain letters,” explains Eric Topol, director of Scripps Translational Institute, a professor of genomics at Scripps Research Institute in La Jolla, California, and a member of Genome’s advisory board.
Increasingly valid answers about the power of our DNA are systematically being gathered from decades of research focusing on which of those three to four million gene variants in each person are linked to the development of disease. The number of mutations may sound huge, but 99 percent of everyone’s base pairs are identical, Khoury notes. This reinforces the complexity of scientifically determining the significance — bad, good, or inconsequential — of any individual variant or group of variants.
For healthy people, experts readily point to several reliable uses for genome sequencing — aside from learning about ancestry — that best utilize its established and emerging strengths.
One of those uses is in carrier testing for those rare, single-gene, recessive disorders in which each parent contributes one copy of a variant gene. This test is typically used in reproductive planning by those with a family history of certain conditions, such as cystic fibrosis or sickle cell disease, or hailing from a specific heritage, such as Ashkenazi Jews, who are more prone to the neurological condition known as Tay-Sachs disease. Since 1993, the number of highly predictive gene tests has risen from about 100 tests to more than 4,000 tests now.
A second use for whole genome sequencing is in predisposition testing for so-called highly penetrant dominant disorders, in which people with the gene mutation have a high risk of developing a disease. These include breast and ovarian cancers, for which BRCA mutations confer an extremely elevated risk; colon cancer, for which Lynch syndrome mutations place carriers at an extraordinarily high risk; and some cardiovascular conditions, such as familial hypertrophic cardiomyopathy, which affects 1 in 500 people worldwide and can be life-threatening. Learning whether they’ve inherited such genetic mutations can help patients lower their risks by frequent monitoring or preventive treatment.
Predisposition testing has arguably captured the most public attention, with Angelina Jolie Pitt becoming the poster child for how BRCA mutation testing can present wrenching but proactive choices to avoid disease. Jolie Pitt, whose mother died of ovarian cancer, had her breasts and ovaries removed in two separate surgeries over the past several years in an attempt to escape the same fate — a decision that generated national attention.
Pharmacogenomic testing to predict how patients will respond to specific medications depending on their genetic makeup is another way in which sequencing can be useful. Certain gene variants can indicate that a medication may be more effective for an individual, while others might increase the risk for severe side effects.
Of sequencing’s uses, Topol and Khoury contend that pharmacogenomic data may be the most valuable to the average healthy person. But the ability to actually access this data — all of which isn’t yet clinically useful — still lags.
“There are well over 100 drugs where the interaction between DNA and the drug [may determine] if there are serious side effects, if it will or won’t work, or if the dose has to be adjusted from the norm,” Topol says. “But that information is not really attainable by anyone today, because there’s no good, widespread service to get that information.”
Finally, whole genome sequencing can reliably determine a wide variety of traits that test subjects may find edifying, intriguing, or just plain odd. These include their blood type; whether they’re lactose intolerant, with dairy consumption causing digestive distress; or how quickly or slowly they metabolize caffeine or alcohol.
While some feel drug-specific genomic data isn’t yet, as Khoury puts it, “ready for prime time,” the same is true of many common gene variants thus far identified by scientists for common diseases like diabetes, schizophrenia, and heart disease. These variants tend to increase a person’s risk of disease only slightly, so knowing about them doesn’t help a physician to treat or prevent illness. And geneticists don’t yet know the best way to compile the subtle health risks posed by many variants predisposing people to everyday disorders in a way that accurately predicts a person’s overall risk.
This is one of the reasons the Food and Drug Administration (FDA) shut down 23andMe’s initial $99 kit that tested for hundreds of genetic markers. According to a letter the FDA sent to 23andMe in November 2013, the agency had serious concerns “if test results are not adequately understood by patients or if incorrect test results are reported.” While the FDA approved 23andMe’s new $199 test kit in October 2015, the new kit offers a fraction of the tests in the original kit.
The complexity of these diseases is brought on by a mysterious interplay of genetic, environmental, and lifestyle triggers, says Isaac Kohane, chair of the Department of Biomedical Informatics at Harvard Medical School in Boston. “For most common diseases, the measure of your risk due to heredity is about 50 percent,” Kohane says. “Even with perfect omniscience about the genome, for common diseases like heart attacks, strokes, diabetes, Alzheimer’s, and so on, a big chunk of risk is still determined by other factors.”
Greater use of DNA sequencing tests is also serving to illuminate other increasingly glaring glitches in scientists’ ability to correctly forecast rare disorders. Some genetic variants originally identified in people with rare diseases and thought to be highly penetrant — bestowing a high likelihood of developing the condition — are now being found in healthy people as well.
This means the variants aren’t as accurate at predicting disease as once believed. “It turns out,” Kohane observes, “that we are riddled with abnormalities that will not kill us.”
Snyder experienced this shortcoming firsthand, when his WGS results indicated mutations suggesting he should have aplastic anemia, a rare and life-threatening condition in which the bone marrow shuts down blood cell production. Fortunately, however, that warning was wrong. “There’s no single gene that’s 100 percent predictive of getting something,” Snyder says, though he adds that some results are more alarming than others.
Snyder is a self-described “believer in the whole enterprise” of genome sequencing. He doesn’t share the deep worries of those who fear the potential psychological ramifications of learning what’s encoded in their genes. Indeed, the technology can open a Pandora’s box — both enlightening people about their genetic variations and spurring emotional distress as well as concerns about how others might use the information.
Michael Linderman, an assistant professor in the Department of Genetics and Genomic Sciences at the Icahn School of Medicine at Mount Sinai in New York City, and a member of Genome’s advisory board, led research published in June 2015 in the European Journal of Human Genetics. The research focused on the motivations, concerns, and preferences of 35 healthy people who chose to have their whole genomes sequenced. The majority expressed interest in learning personal disease risk or health-related information, with one-quarter mentioning concerns about the adverse psychological impact of potential results. Nearly 6 in 10 endorsed concerns related to potential privacy issues about their data, and an equal number gave permission for their data to be used by researchers elsewhere.
Indeed, legislation is still catching up to the rising availability of genetic information. The federal Genetic Information Nondiscrimination Act forbids discrimination based on genetic information with regard to health insurance or employment, but doesn’t cover life, disability, or long-term care insurance.
Aside from privacy concerns, experts fret that a deluge of personal genetic data may create a class of “worried well” who engage in ultimately unnecessary diagnostic tests and treatments based on information that’s still of questionable value.
“Until we develop the knowledge, I think we have to be ‘buyer beware,’ ” Kohane says. “Just as critical as we should be of any test, the whole genome presents us with literally millions of tests and each will do something potentially helpful or harmful to you. If you’re healthy, you have more to lose. If you’re not healthy, you have less to lose.”
Despite the current drawbacks of the technology for genome sequencing, Kohane and others in the field contend that the more healthy people who take part, the more valid the data will become, with widespread benefits to society. “There will be a day,” says Kohane, “when we’ll have sequenced thousands of individuals with a variety of diseases who were healthy for a very long time, and we’ll understand the meaning of these variants.”