In DepthPolicy

Baby genome screening needs more time to gestate

See allHide authors and affiliations

Science  28 Oct 2016:
Vol. 354, Issue 6311, pp. 398-399
DOI: 10.1126/science.354.6311.398

Genome sequencing of babies may not quickly replace biochemical tests of blood from heel pricks.


Genetics researchers have begun gingerly approaching an ethically fraught idea: As DNA sequencing gets ever cheaper, should every newborn baby, ill or seemingly healthy, have its genome analyzed? Would sequencing help spot rare diseases and guide future medical decisions? Or would it open a Pandora's box of needless costs and worries?

Two research projects exploring these questions are now delivering early results, and both are giving pause to newborn-sequencing advocates. One suggests that exome sequencing, which reads the protein-coding regions of a genome, is not comprehensive enough to replace standard newborn blood screening for rare diseases. And the other finds that parents are surprisingly uninterested in having their newborns sequenced.

Together, the findings suggest newborn sequencing “as available now is not the answer,” says geneticist Jennifer Puck of the University of California in San Francisco, who co-leads one of the studies presented last week at the annual meeting of the American Society of Human Genetics (ASHG) in Vancouver, Canada.

The projects are among a quartet funded by the National Institutes of Health (NIH) to compare the risks and benefits of sequencing newborns' DNA with those of existing screening programs, which rely on biochemical analysis of newborn blood spots to test for dozens of metabolic diseases (Science, 13 September 2013, p. 1163). NBSeq, co-led by Puck, is sequencing exomes of deidentified samples from 1700 newborn blood spots archived after standard biochemical screening in California.

With 182 samples analyzed so far, the results are sobering: The DNA sequencing missed about 20% of confirmed disease cases that had been flagged by biochemical screening. All were very rare diseases, some of which may have been caused by mutations, perhaps in rearranged DNA or regions regulating gene activity, not detectable by exome sequencing. Whole genome sequencing might have found some of the culprit mutations, Puck says, but it's still too costly and challenging for routine use.

Exome sequencing does appear useful, however, in concert with the traditional screening. It can pin down the gene responsible for an abnormal biochemical result, yielding a definitive diagnosis. “In tricky cases it will help,” Puck says. NBSeq is also finding that sequencing may help identify newborns with inherited immunodeficiency disorders. Only one of those diseases is detected by standard biochemical screening of a newborn's blood; others often only surface later, when the child develops recurrent infections.

Another NIH-funded newborn project aims to show whether sequencing a long list of genes could be useful for routine medical care. The BabySeq team, co-led by geneticist Robert Green at Brigham and Women's Hospital in Boston, analyzes babies' DNA for mutations in roughly 7000 genes implicated in metabolic and other childhood diseases or adverse drug reactions. The effort, which began in May 2015, is hitting an unexpected snag: The parents of only 24 of 345 sick babies in neonatal intensive care, and 138 of 2062 healthy babies, have agreed to join BabySeq, Green reported at the ASHG meeting. That 7% response rate from both groups is much lower than expected based on a survey of hundreds of parents of newborns about 4 years ago, when nearly half declared they would be “very” or “extremely” interested.

Many parents were apparently put off by having to return to the hospital with their newborns to discuss the sequencing results. Others who met with a genetic counselor before being invited to join the study had concerns such as privacy, the prospect of worrisome or unclear results, and insurance discrimination. Although U.S. law prohibits health insurers from denying coverage based on genetic data, the children could be denied life, disability, or long-term care insurance. “A lot of that is literally 50 years in the future and they are concerned,” Green says.

Green's study so far has delivered mixed evidence on whether the genetic data will be useful. Three of the newborns carried inherited mutations implicated in heart disease, but there were no signs of illness in the parents or child. Green's team will track whether the DNA results benefit the family, or lead to unnecessary anxiety and tests. Another baby had mutations linked to an enzyme deficiency that biochemical screening missed; although asymptomatic, the baby had enzyme levels that are slightly below normal and is being treated. A fifth infant carried a mutation in a gene linked to breast cancer. At first, Green's team did not plan to tell families about mutations linked to adult-onset diseases, but they got permission from the study's ethics boards to disclose the result. The mother, who had relatives with breast cancer, “was obviously concerned but grateful,” Green says.

Because of the uncertainties and possible harms of routine newborn sequencing, ASHG's current position is that only babies with an undiagnosed illness should have their DNA read—and preferably only for genes likely to explain the disorder. But genomics policy expert Misha Angrist of Duke University in Durham, North Carolina, thinks the appeal of sequencing every newborn may grow. “These are still very early days,” he says. “If more people do this and the discrimination and confidentiality risks do not materialize, then presumably more people will choose newborn sequencing.”

View Abstract

Stay Connected to Science

Navigate This Article