By Allison Proffitt
July 29, 2016 | Researchers mining data in California’s cystic fibrosis (CF) newborn screening registry have identified an additional mutation that is linked to later onset disease, but is usually missed in newborn screening.
“The more information you look for, the more you may be able to say about more unusual forms of CF,” said Richard Parad, a neonatologist at Brigham and Women’s Hospital and one of the authors of the study.
That may seem self-evident, but it’s not the way most newborn screening is currently conducted. Parad believes there is room for improvement, and that CF is a particularly strong testbed.
Parad has been working with California’s cystic fibrosis newborn working group for years. With colleagues from the Children’s Hospital Los Angeles; the Sequoia Foundation, La Jolla, Calif.; and the California Department of Public Health, he published the findings of a retrospective study in Genetic Testing and Molecular Biomarkers last week (doi:10.1089/gtmb.2016.0102).
Parad has seen the state’s testing methodology develop. It’s unique, he told Diagnostics World.
California’s Model
“California has about half a million births a year, so they are able to collect a lot of information and we took advantage of that,” Parad said. “They have a very, very beautifully-designed newborn screening test in part influenced by the state’s Department of Public Health regulations, which state that you can’t do a test that discriminates against any particularly group. It has to be able to pick up a disease equally well in any group.”
To ensure that its newborn screen included disease-causing mutations common in non-Caucasian populations—particularly Hispanic—California sequenced 700 CF patients and built its own panel of the 40 CF-causing mutations in the CFTR gene most common in the state.
In California, newborns that test positive in the initial screening for the protein immunoreactive trypsinogen (IRT) are referred to one of 16 state CF centers for diagnostic confirmation, follow up, and treatment. These children undergo genetic testing against the state’s 40-mutation panel. Newborns with two mutations are diagnosed with CF, but those with only one mutation are generally deemed carriers. In California, the exons of the CFTR gene in carriers are sequenced by Sanger sequencing.
“That’s how we got the information that we wrote the paper from,” Parad explains
Researchers mined those sequencing data retrospectively to look at the health outcomes and genomic findings for 428 of these carriers. This paper specifically looks at a common mild CFTR gene variant that is carried by nearly one of 10 people, the 5T allele, and the TG repeat length adjacent to that mutation.
The researchers were able to generate risk predictions based on the length of the TG repeat. Newborns with the 11 TG showed no signs of CF during eight years of follow-up. However, 6% of babies with the 12 TG developed the disease and nearly 40% of children with the 13 TG were considered to have CF within eight years of birth.
This research, and other work that Barino et al have published, is made possible because of the sequencing data collected for these carrier children. “We had information on this 5T variant and the TG repeat region, which would not be available through any other newborn screening algorithm,” Parad says.
The impact of this knowledge could be great. “Genomic sequencing as part of the newborn screen would allow us to dive deeper into predicting problems that are going to happen later on,” Parad argues. “If you can know that a problem is there before the clinical manifestations have been firmly established—before the horse is out of the barn—the ability for you to intervene and improve the ultimate outcome should be greater.”
This research didn’t prove that identifying the more at-risk children would improve their long term health outcomes, but Parad referenced long term CF work going on in Wisconsin, the first state to do universal cystic fibrosis screening, that has shown that earlier diagnosis can prevent malnutrition and have better outcomes. “We haven’t proven that specifically for these slightly atypical, later-onset, perhaps milder CF forms, but theoretically the same principle should be true.”
Whole Vs. Part
In Massachusetts, Parad is also involved in the BabySeq program that offers whole genome sequencing to all the babies born at Brigham and Women’s Hospital. He confesses a bias toward whole genome sequencing, but acknowledges that there’s much work to be done.
“We’re not ready for prime time [with BabySeq],” Parad emphasizes, “but we’re studying what happens if you try to do that and how it impacts parents and pediatricians and care of the babies… I do think that there’s a lot of information—if we knew how to use it—that we could gain from doing sequencing as part of a newborn screen. But it’s really complicated!” he says.
“Ultimately sequencing the whole genome might not be that much more expensive, but knowing what to do with the information is another issue. Interpreting it, presenting it to parents, what genes do you look at or not look at that may or may not be relevant to the newborn or pediatric care, those are the things that BabySeq is looking at.”
But the 5T findings, and others like, it are important enough to suggest a solution somewhere in the middle.
“In CF screening, perhaps doing IRT and then doing next generation sequencing of the CFTR gene in those babies who have high IRT would be a better algorithm in the future,” Parad proposes. “The cost of that would still be challenging,” he admits, “but there might be ways to work that into a newborn screening algorithm.”
“For the 30 or 50 newborn screens that are done we might want to look at an analyte measured in the blood first, and if that’s abnormal… and we’re going to a second tier, let’s do next generation sequencing of that whole gene,” he says. That will give us more information and may help us pick up more variants of the disease.”