Pregnant women and their partners can already peer at an unborn child’s chromosomes: with amniocentesis, they can learn about the presence or, more likely, absence of large-scale genetic defects, often gaining peace of mind. But only a small percentage of parents-to-be take the opportunity, because the procedure is invasive and uncomfortable—a large needle is inserted into the amniotic sac—and causes miscarriage in roughly one in 400 cases.
Researchers have long hoped to develop a noninvasive alternative. Ever since scientists discovered, in the 1990s, that pregnant women’s blood contains substantial amounts of fetal DNA, they’ve theorized that they could use this genetic material to test for fetal abnormalities like an extra copy of chromosome 21, which causes Down syndrome.
That technology has now arrived (see “Prenatal DNA Sequencing,” May/June 2013). Several companies have introduced genetic tests that use blood drawn from the mother. These tests can be performed earlier in pregnancy than amniocentesis is usually done, which means that if the results suggest an abnormality, women and their partners have more time to grapple with whether to have an abortion or prepare for a child with special needs. If the results are reassuring, the cloud of anxiety dissipates sooner.
Given that the risks of having blood drawn are minimal, the tests are likely to be widely used. While today fewer than 5 percent of pregnant women undergo amniocentesis, “I think we could see 50, 60, 70, 80 percent of American pregnancies getting genetic testing,” says Hank Greely, director of the Center for Law and the Biosciences at Stanford.
Noninvasive prenatal screening
The catch, though, is that as the accuracy of these tests continues to improve, they will be able to detect a greater range of genetic variations, including some with murkier implications. For example, rather than indicating something with certainty, they could reveal elevated risks for certain diseases or disorders. These advances could collide with the politics of abortion and raise the ugly specter of eugenics. When, if ever, should parents terminate pregnancies on the basis of genetic results? Do we have the wisdom to direct our own evolution? And perhaps most important, are there limits to how much data parents should have—or want to have—about their children before birth?
The first noninvasive tests to reach the market have screened for the largest-scale genetic defects—namely, abnormal numbers of chromosomes. Sequenom Laboratories, Verinata Health (part of Illumina), Ariosa Diagnostics, and Natera all offer tests that look for trisomies—an extra copy of chromosomes 13, 18, or 21, which cause Patau syndrome, Edwards syndrome, and Down syndrome, respectively. Some also identify an aberrant number of sex chromosomes. This fall, Sequenom expanded its test to encompass additional trisomies as well as selected microdeletions (in which DNA is missing), including those known to cause DiGeorge syndrome, -Cri-du-chat syndrome, and Prader-Willi or Angelman syndrome. The various companies’ tests range in price from less than $1,000 to almost $3,000, though they are covered by some insurance plans. So far, these offerings have not replaced amniocentesis, which remains the gold standard for accuracy. But they can be performed as early as 10 weeks into pregnancy and can help identify women who may need the more invasive test.
Companies will modify these tests to flag an increasing number of genetic conditions, including some that are quite rare. The trend is toward “detecting smaller and smaller mutations,” says Jonathan Sheena, chief technology officer of Natera, who predicts that noninvasive identification of inherited single-gene diseases like cystic fibrosis, Tay-Sachs, and neurofibromatosis will soon become commercial reality. In the laboratory, meanwhile, researchers have already used noninvasive methods to sequence a whole fetal genome. In 2012, geneticist Jay Shendure’s group at the University of Washington analyzed blood from the mother as well as a saliva sample from the father to reach this goal. Also in 2012, Stephen Quake’s group at Stanford used a maternal blood sample alone to derive the fetal exome, which consists of the coding parts of genes. “That’s pretty much the whole ball of wax,” Quake told me. (Shendure and Quake are advisors to Ariosa Diagnostics and Verinata, respectively.) These laboratory efforts were not cheap: Shendure says it cost him around $50,000 to do the full genome. But they represent a clear proof of principle. And as the costs of sequencing continue to plummet, far more parents-to-be will potentially have access to far more genetic data about their future children.
Quake says he hopes the technology will be used to identify and manage conditions that are well defined and for which early intervention can make a difference; he points to metabolic disorders like phenylketonuria, in which children require a strict diet, and certain immune disorders that can respond to early treatment. If babies’ problems can be diagnosed prenatally, he says, “you’re not putting them in distress for the first few weeks” while everyone is “running around trying to figure out what is wrong.” Another example is a condition called dilated cardiomyopathy, in which the heart is enlarged and weakened. This disorder can go undiagnosed until its victims find themselves short of breath or have a heart attack as teenagers or young adults. By treating them from a young age with drugs, physicians can “dramatically change outcomes,” says Euan Ashley, a Stanford researcher who cofounded Personalis, a genetic screening company.
But the moral quandaries are sure to intensify as well. If many more women receive information about genetic disorders like Down syndrome earlier in pregnancy, it’s likely that the number of abortions will rise. Inevitably, some people will object to the testing technology because of their opposition to abortion, says Greely. And some current parents of children with Down syndrome will worry that if fewer people are born with the disorder, medical research and public support will start to dry up. The unease deepens with less severe disorders like Kleinfelter’s syndrome, which is caused by an extra X chromosome in males. Boys with this syndrome often have few noticeable symptoms early on and may not be diagnosed until later in life, when they may experience atypical sexual development, learning difficulties, and infertility. If genetic testing identified more cases prenatally, some of those pregnancies would almost surely be terminated. Even firm supporters of abortion rights may find that thought troubling. Similarly, consider achondroplasia, which is an inherited form of dwarfism. If two parents with achondroplasia wanted a child who looked like them, “would it be wrong for them to terminate a normal-sized fetus?” Greely asks. “These are hard questions.”
For now, testing for intelligence or height or other complex traits that might pique parents’ curiosity appears to be far off: researchers largely seem skeptical that they will be able to predict these traits from an individual’s genome in the foreseeable future. “We’re really bad at it right now,” says Shendure. “In 10 years we’ll probably still be pretty bad at it.”
But the underlying issue will still complicate the abortion debate: to what extent should parents be able to choose the traits of their children—and should the calculus change when the traits in question, like sex or hair color or eye color, are not directly linked to disease? For the most part, we tend to trust parents to make the right decisions for their children, but that prerogative may not be absolute, especially when it comes to nonmedical factors. We can’t know how children’s lives will unfold or how important a whole range of traits might turn out to be to them. We surely don’t have the understanding to guide our own evolution, or even to understand the extent to which individuals’ genomes relate to their health or happiness. And given the disastrous history of eugenics, from forced sterilizations to the Holocaust, we should maintain a healthy fear of even small-scale efforts to select some nonmedical traits over others. This is not merely a theoretical matter: parents in India, China, and South Korea who learn their fetuses’ sex through ultrasound have disproportionately chosen abortion in the case of girls. (Arizona has already made it illegal to abort on the basis of sex or race, though introducing criminal penalties for doctors is not necessarily
Perhaps the biggest question is which information will be meaningful for parents to receive. Genetic interpretation can be a dicey game. It is well known, for instance, that mutations in the BRCA1 gene are strongly associated with breast cancer, but in a disturbingly large number of cases, patients are told they have variants of unknown significance. “It would be very unfortunate if we started delivering ‘variants of unknown significance’ results in the context of reproductive health,” Shendure says. Similarly, when it comes to complex problems like cognitive impairment, it’s not clear how useful it is to test for—or report on—variants that have been associated with disabilities. Research suggests, for instance, that people with specific duplications on chromosome 16 are at higher risk of mental impairment. Some are severely affected, but others are “absolutely, perfectly healthy, functioning normally,” according to Wendy Chung, director of clinical genetics at Columbia University. To date, there is no reliable data on what percentage of duplication carriers fall into each of these categories, meaning that prenatal testing for these variants could greatly increase parents’ anxiety while leaving them at a loss to assess the results quantitatively. Then there are girls with three copies of the X chromosome. They are also at higher risk for cognitive impairment and learning disabilities, but the risk remains small, and the vast majority of them will be normal. How should parents make sense of these possibilities? Most of us find it hard to think about risk, and we are truly bad at predicting how future events will affect us emotionally. And on top of all that, who knows which disorders will be curable or treatable through gene therapy or some other method 20 or 30 years from now? In other words, we’re not ready for the onslaught of information the new tests seem poised to provide.
Nevertheless, that information is coming, and parents will have to figure out what they want to know and how to interpret the choices they’re offered. It is critical, then, that the informed–consent process for testing be exceptionally good, says Greely. Ideally, parents should meet with a genetic counselor to discuss what exactly testing might reveal and what wrenching decisions might follow. If formal genetic counseling isn’t available, obstetricians should step in with extended, thorough conversations that take into account the parents’ values, desire for data, and tolerance for uncertainty. Genetic testing, as Greely puts it, should be made distinct from other forms of prenatal care; it should never be “just one more tube of blood” taken in the course of another whirlwind visit to the doctor.
Amanda Schaffer is a freelance journalist who writes about science and medicine for Slate, the New York Times, and other publications.
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.