Whole Genome Sequencing Boosts Diagnosis of Rare Disease in Infants

Whole genome sequencing captured almost twice as many genetic abnormalities that may be responsible for disease in infants, compared with a standard targeted test, researchers found.

In the Genomic Medicine for Ill Neonates and Infants (GEMINI) study, the diagnostic yield of whole genome sequencing was 49%, compared with 27% with NewbornDx, a standard targeted gene panel, Jill Maron, MD, MPH, of Women and Infants Hospital of Rhode Island in Providence, and colleagues reported in JAMA.

Whole genome sequencing did take longer to return results on average -- 6.1 days versus 4.2 days for NewbornDx -- but it actually took less time for the most urgent cases (3.3 versus 4 days), the researchers said.

"Our study shows the ability to understand the presentation of these genetic diseases early in the life course. That's been missing," Maron told MedPage Today, noting that getting answers in rare disease is often typically described as the "6-year diagnostic odyssey."

"A subset of the babies diagnosed in the neonatal period ultimately would have been diagnosed," she said, "but now hopefully we can diagnose them earlier and understand the natural course of these genetic diseases, and hopefully have more interventions for them."

Regarding the time to return results, Maron noted that "if we believed we had a critically ill baby, the whole genome came in faster than the panel, so I think the 2 days [difference] is probably irrelevant."

The work was supported by a grant to Tufts Medical Center in Boston from the Clinical and Translational Science Awards (CTSA) program at the NIH's National Center for Advancing Translational Sciences (NCATS), according P.J. Brooks, PhD, acting director of NCATS' Division of Rare Diseases Research Innovation. The goal of the program is to accelerate the translational research process and increase collaboration across different research and clinical sites.

"This is getting at the whole diagnostic odyssey for rare diseases," Brooks told MedPage Today, noting that since the project was originally funded 5 years ago, "much has changed in terms of the development of gene-targeted therapies. It's of great interest how you go from these molecular diagnoses ultimately to therapeutic intervention."

While the "diagnostic odyssey remains an issue" for rare disease patients, Brooks noted NCATS has "several different projects we're funding that can incorporate genomic sequencing and artificial intelligence approaches to speed that up."

The prospective, multicenter study enrolled 400 hospitalized infants younger than 1 year of age who were suspected of having a genetic disorder at 6 U.S. hospitals from June 2019 to November 2021. Their parents were also included when available, totaling 388 mothers and 318 fathers.

They received simultaneous testing with whole genome sequencing and the more targeted genetic panel. Whole genome sequencing was done by Rady Children's Institute for Genomic Medicine in San Diego, and Quest/Athena Diagnostics conducted the commercially available NewbornDx test, which looks at 1,722 genes associated with disorders that typically present early in life.

Overall, a molecular diagnostic variant was identified in 51% of infants, who had at least 1 genetic variant identified by either test that was deemed as causal (pathogenic or likely pathogenic) or highly suspicious (a variant of unknown significance [VUS]).

A total of 95 infants had their diagnosis identified by whole genome sequencing but not by the targeted test (49% of 195 positive cases), and only 9 infants arrived at their diagnosis via the targeted test but not on whole genome sequencing (8% of 109 positive cases).

Maron and colleagues also found 134 variants that were considered novel, which will be uploaded to the NIH's ClinVar, a database of genetic variants, she said.

The researchers also noted that discordant results were not uncommon. When both labs detected the same variant, the interpretation and classification differed 43% of the time, they found, noting theirs was not the first study to report "discrepant variant interpretation."

Maron provided an example of how inheritance impacts interpretation: "You get a hit from your mom that's pathogenic, and the hit from your father was a variant of unknown significance. The labs may differ and say you need two pathogenic hits to have this disorder, so [one lab may] call this a VUS. But the other lab would say, 'I have a pathogenic allele and a VUS, and I think it's causing a problem. I'm going to call it a pathogenic variant.' ... All of it has to do with human interpretation."

Brooks said the diverging interpretations point to "the need for better standardization of how one interprets variants of unknown significance. In the future, we should encourage data sharing and maybe more concordance about how you interpret these variants of unknown significance. There are various efforts going on toward that but it seems maybe more needs to be done."

Ultimately, about a fifth of patients (19%) had changes in clinical care based on the results of their testing, Maron and colleagues reported. They noted 15 infants "transitioned from palliative care to either a known treatment for the disorder or withdrawal of life support."

"Most of the disorders that we diagnose have no treatment," Maron said. "What we're really doing is getting an answer to put a name to what this is, and contributing to the body of literature saying a lot of babies have genetic disorders, many of which have not been recognized before. Hopefully, that helps to move the field forward."

Indeed, 76% of clinicians reported that they perceived genomic sequencing as useful or very useful, whether or not a diagnosis was provided.

The study was limited because it didn't conduct formal superiority testing to assess whether one test was superior to the other. Researchers should also be cautious about extrapolating the data to real-world settings, partially because "access to the technology and well-developed study protocols likely affected the speed in which a molecular diagnosis was made."

Still, Maron and co-authors concluded that their study "confirms the need to have such testing widely available and covered by Medicaid and commercial insurance."

The cost of genetic testing has long been a problem for rare disease diagnosis, as not all insurers will pay for such testing, Maron said. She estimated that out-of-pocket, whole genome sequencing would cost "about $9,000 per trio -- mom, dad, and proband."

Sometimes a hospital will absorb that cost, she said: "In the end, it will save the hospital money if we can actually get the diagnosis rather than another day in the ICU where we don't know what we're dealing with."

Study co-author Jonathan Davis, MD, chief of newborn medicine at Tufts Medical Center in Boston, noted it "can be frustrating to get all these MRIs, blood tests, and biopsies -- all these tests they'd cover, but this [genetic] test they won't cover."

This prompted his group to conduct a cost-effectiveness study, which will be published in the near future, he said.

Davis noted that genetic testing has also come into play in malpractice cases, which are frequent in the ob/gyn realm.

"In many cases, they're yelling at the obstetrician who should have done a C-section sooner, and we'd say, no, it's something else but we just can't figure it out," Davis said. "You hear upwards of $200 million malpractice awards for one baby, and you know some of the time it's easy to blame the obstetrician or the hospital, but we should be sequencing these kids and seeing whether there's something else there. Some of the courts are starting to allow that."

Comment