Sequencing Not Ready for First-Line, Newborn Screening

Scientists have been exploring the "hows" and "what ifs" of applying sequencing to newborn screening. However, two new National Institutes of Health-sponsored studies presented at the American Society of Human Genetics annual meeting (Vancouver, Oct. 18-22, 2016) are dashing hopes that every newborn will have their genome sequenced any time soon. The NBSeq project found that whole-exome sequencing (WES) does not have "acceptable" accuracy to replace traditional newborn screening, while, a second project, BabySeq, found that parents are not that interested in having their newborns sequenced.

Exome Sequencing Not Accurate Enough for Newborn Screening
The NBSeq project found that WES of newborn dried blood spots misses 25 percent of metabolic disorders found by traditional newborn screening with tandem mass spectrometry (MS/MS). The authors say these findings indicate limits to WES both in newborn screening and diagnostic testing.

MS/MS detects many metabolic disorders with excellent sensitivity, but often does not identify the precise disorder, experts say. Given newborn screening is done at the population-level it must scale and have greater than 99 percent specificity. By contrast, sequencing performed on patients suspected of genetic diseases is cumbersome, requiring manual review from experts, and still has a diagnostic yield between 25 percent and 50 percent.

The NBSeq researchers used archived dried blood spots from all California newborns (July 2005 to December 2013) with disorders diagnosed by MS/ MS plus a selection of false positives (n = 1,600 samples). To date, the group has sequenced 600 samples, and analyzed 184.

The researchers found that WES yields "vastly" more false negatives than MS/MS. In one quarter of newborns with metabolic disorders, WES did not identify rare alleles for genes responsible for their recessive Mendelian disorder. They systematically explored how different WES interpretation protocols impacted the prediction of metabolic disorders in newborns. While tuned analysis pipelines shifted the balance of false negatives versus false positives as part of exploration of how different interpretation protocols impacted identification of metabolic disorders, it did not yield acceptable specificity and sensitivity for newborn screening. While apparently unsuitable for first-line NBS, WES might yet improve NBS in NICU settings or help specify a diagnosis following positive MS/MS screens.

"The success of diagnostic WES is limited in part because patients' disorders may be genetically complex or not genetic," writes lead author Aashish Adhikari, a postdoc at University of California, Berkeley. "Our study suggests limits to identifying even monogenic disease as well; our work focused on some of the best-characterized Mendelian disorders, yet recognized only three-quarters of the cases. This therefore suggests a ceiling to the potential of current diagnostic WES."

The authors add, though, that WES may be useful for diagnostic purposes in conjunction with positive MS/MS newborn screening findings.

Parents Not Very Interested in Sequencing
The BabySeq Project is a randomized controlled trial exploring the utility of sequencing for infants in both the neonatal intensive care unit (NICU) and well-baby nursery at Brigham and Women's Hospital in Boston. Half of infants are randomized to receive genetic sequencing. Families receive results during a session that involves review of state-mandated newborn screening results and family history. The families of infants randomized to sequencing also receive a report that includes pharmacogenomic variants and pathogenic and likely pathogenic variants in 1,800 genes associated with dominant and recessive monogenic conditions with childhood age of onset and high estimated penetrance, but without regard for actionability, the authors say.

To date, participation in the BabySeq project has been offered to the parents of 254 NICU infants and 1,193 healthy infants, of whom only 6 percent and 7 percent, respectively, have enrolled. Time from DNA extraction to issuance of a sequencing report has averaged 50 days.

Parents who declined before meeting with a study genetic counselor most often cited logistics of return visits as their primary reason for not participating. Parents who declined after meeting with a study genetic counselor cited the potential to receive unfavorable or uncertain results (38 percent), insurance discrimination (26 percent), and confidentiality/privacy concerns (22 percent) as their reasons for not participating.

A related BabySeq abstract surveyed participating parents and doctors about their perceptions of sequencing technology, compared to newborn screening.

The majority of parents (81 percent; n=159) and physicians (97 percent; n=67) agreed that there are health benefits associated with newborn screening. A smaller majority of parents (67 percent) and physicians (57 percent) agreed that there are health benefits associated with genomic sequencing. More physicians than parents agreed that there are risks associated with genomic screening (73 percent versus 35 percent, respectively). Both parents and physicians identified similar risks resulting from genomic sequencing, including psychological distress, genetic discrimination, impact on family, and the nature and potential uncertainty of genetic sequencing results. Parents and physicians also agree that genetic sequencing will increase in importance in 10 years, with 82 percent of parents and 90 percent of physicians saying it will be more important in the future, compared to the 61 percent and 64 percent, respectively, that view genetic sequencing as important now.

Takeaway: Despite much enthusiasm for the clinical application of WES, the technology is not yet ready for application to newborn screening.