(CNN) Every baby born in the United States is given a routine blood test to screen for dozens of inherited medical conditions. Now, the U.S. National Institutes of Health is exploring whether to use DNA sequencing to screen newborn babies for additional genetic abnormalities and disorders. Such DNA testing would likely complement, but not replace, the current routine blood tests.
However, before routine genetic screening of infants even approaches reality, many questions need answers, including whether genetic sequencing can accurately identify babies who will develop a disease, according to Dr. Joseph A. Bocchini Jr., chairman of the Advisory Committee on Heritable Disorders in Newborns and Children. The committee evaluates scientific evidence and makes recommendations to the secretary of the Department of Health and Human Services, which in turn provides a recommended uniform screening panel for newborns to the states.
The field is evolving swiftly, Bocchini told CNN: "It's clear the data is becoming available quite quickly, so potential changes [to the recommended uniform screening panel] may occur within the next few years. But it's too early to say."
More clarity is also needed on issues surrounding newborn DNA testing, including consent, accessibility, data privacy and the potential changes to medical practice and costs.
Published Thursday in the American Journal of Human Genetics, one of the several NIH-sponsored scientific studies found that 9.4% of the 159 sequenced babies participating in the research had mutations predictive of a genetic condition or disease.
"The question, though, is: 'Do we really think that all these babies are going to get sick in the future based on what we found?' " said Alan Beggs, co-author of the study and director of the Manton Center for Orphan Disease Research at Boston Children's Hospital. "And the answer is, 'Probably not.' "
This is the current "dilemma" with genetic testing, said Beggs, this "unknown sensitivity and specificity." While it may be true, for example, that everyone with a certain illness shares a specific genetic mutation, it may also be true that others also have that mutation, but might never become sick. In genetics, this concept is called "reduced penetrance."
His study, said Beggs, is really exploring, "How do we best communicate this type of uncertainty to families and to their doctors?"
Another uncertainty with genetic sequencing is that it reveals genes that don't get "expressed," meaning the protein the gene codes for doesn't get made, said Dr. John Lantos, the director of the Children's Mercy Hospital Bioethics Center in Kansas City, Missouri. "The whole process of going from gene to protein is controlled by all sorts of things we don't yet understand," said Lantos, who did not participate in Beggs' study but conducted his own infant sequencing project for the NIH. "Every attempt to link some specific genome sequence variant with some specific disease runs into all these qualifiers and modifiers and expressivity and penetrance."
Lantos pointed out that the entire field of genomics is a relative newborn itself.
"The first genome was sequenced in 2003 and cost $3 billion to do," Lantos recalled. Today, sequencing is "a combination of computer applications and people," he explained.
First, a genome machine "spits out" the millions of base pairs of an individual genome, said Lantos. Next, a computer program sifts through these pairs and boils them down to a subset. This "preliminary cut" might show a 100 genetic mutations that appear to be disease-causing, 19,000 that look harmless and 1,000 mutatations of unknown significance, he said.
For the last step in the process, a trained genetic scientist looks at that result and does an analysis "that is more art than science," said Lantos.
It's only become technically feasible to do full-scale sequencing studies, including the several infant studies sponsored by the NIH, in the last five years, he said: "The question is, can we develop a way to use this innovation that does more good than harm?"
One such harm, as envisioned by Beggs, would be false predictions of disease based on findings in a newborn's DNA.
"We might be causing unnecessary anxiety and stress if it turns out they're not going to get [the disease] after all," said Beggs. The wrong genetic forecast could also lead to unnecessary medical testing, which would have an economic impact and might involve medical risk. For example, a doctor could order a biopsy and the patient could have a bad response to anesthesia or the procedure.
"The risks on a per-patient basis are extremely tiny, but if you were to scale this to millions of babies, then there would probably be an occasional bad outcome," Beggs said. "Are there enough good outcomes and benefits to outweigh the potential for an occasional bad outcome?"
Jeantine E. Lunshof, a philosopher and ethics collaborator in the Church lab at Harvard Medical School and an assistant professor at the University of Groningen in The Netherlands, told CNN in an email that genetic screening of newborn children "differs significantly" from genetic diagnostic testing in a sick child, where researchers look at only a part of the genome for a specific mutation known to cause symptoms or a disease.
One of the "key ethical issues" when using genome sequencing as a screening tool in newborns (or children in general), she wrote, "is that a comprehensive genetic profile is established without the person's consent and without a clinical indication. However, this is a thorny issue, as parents decide all sorts of very important things for their children that sometimes have lasting consequences."
"The issue with genetic information is, that once generated, it cannot be made 'undone,' " wrote Lunshof, who was not involved in Beggs' study. "If genetic information (that is often probabilistic) is on file, will it be used similar to 'existing disorders' and lead to denial of health insurance coverage?"
Both the Health Insurance Portability and Accountability Act (HIPAA) and the Genetic Information and Non-Discrimination Act (GINA) offer some protections. Medical record privacy was enhanced under HIPAA, while GINA prohibits health insurance companies from using genetic information to make coverage decisions. However, GINA does not cover life insurance or long-term care or disability insurance.
Lunshof explained that when genetic screening is offered to adults, there's an "ethical requirement of informed consent. Adults can weigh the benefits and disadvantages and consider issues of privacy and access to their data. A newborn cannot provide informed consent so it's "more difficult to ethically justify the screening of babies and children," she said.
Dr. Cynthia M. Powell, a professor of pediatrics and genetics and director of the Medical Genetics Residency Program at The University of North Carolina at Chapel Hill, said, "We could be taking away that child's autonomy to decide for themselves when they are older whether or not they want this information. The other ethical issue is taking away that child's right to an open future."
Powell, who is conducting her own NIH-sponsored study of newborn screening, said infant sequencing not only raises concerns about potential future insurance discrimination but also potential future "employment discrimination or social discrimination."
Yet, catching genetic disorders in children before they develop symptoms could positively change lives since the earlier treatment begins, the better, she explained. Powell worries about access: "It's not fair if only those children born to families who can afford to pay for it have the ability to receive it."
"But could we handle it on a public health scale basis? My main concern is we can overwhelm the system," said Powell. "There's a shortage of trained geneticists and trained counselors out there and if we open Pandora's box, will it be in the best interests of the child?"
Privacy, both personal and extended, is the main concern of Lee Tien, a senior staff attorney at the Electronic Frontier Foundation, a nonprofit that works to ensure rights and freedoms are protected as the use of technology grows.
"The genome isn't just about you -- it contains information about your parents, your siblings, and your own progeny," Tien told CNN in an email. "So from a privacy standpoint, DNA data is a far more sensitive kind of health information than a sports injury, and it challenges our conventional norms of consent because you're effectively making decisions about other people's DNA."
"We don't know what the data means, we just think we do -- and we may not be very good at dealing with the information," Tien wrote. "Will I, as a parent, treat my kid differently than I otherwise would have because I believe from newborn sequencing that he or she has an above-average chance of developing a type of mental illness?"
It seems right to consider the "best interests of the child," Tien wrote, "but parents have their own interests (and may not even agree with each other), and that assumes both parents are involved in the decision." He also questioned, along with Lunshof and Powell, whether the newborn genome sequence data could be kept private and secure.
This is the focus of Lantos' NIH study, which looked at whether newborn genomic sequencing for babies in the NICU (neonatal intensive care unit) could be turned around quickly enough to affect their medical care, he explained. In at least a few cases, doctors thought it made a difference.
Unexpectedly to Lantos, rather than providing information that guided treatment for a child's recovery, the sequencing results more frequently led to discussions with parents about withdrawing life support, he said: Newborn sequencing "confirmed a bleak diagnosis" and so prevented pain and a prolonged dying process.
Beggs said, "I don't advocate for sequencing kids now," at least not while scientists are still learning to predict the consequences of what they find when peering into the mysteries of an individual's genetic code. He described the families who enrolled in his study as "early adopters" willing to take on the risk of living with uncertainty to help advance science.
Beggs believes that over time "the level of uncertainty is going to go down" and newborn sequencing will become a "standard of care." Another possibility, though, is that unless there's a medical reason, sequencing will be deferred until a child reaches age 18 and parents might be sequenced instead; in the coming years, this might occur during pregnancy or as a routine part of their own health care, he said.
"If you sequence both the parents, then you know most of what you need to know about the baby -- not everything because all of us carry a small handful of new mutations that occurred during our own fetal development," said Beggs. "But by and large if the parents have been sequenced there will be much less urgency for sequencing the child."
The newborn period is a very stressful time for parents, so it's not really the best time to undergo this process, said Beggs. And by waiting until children are legal adults, you preserve their autonomy.
"Part of what we're learning is how much more complicated it is than what we thought 10 years ago," Lantos said. "It's like exploring a new continent."