Genomic newborn screening catches conditions conventional tests can miss
Genomic newborn screening is moving from concept to clinical operations, and the SNUseq Baby Project offers one of the clearest looks yet at what happens when sequencing enters the nursery. In a prospective multicenter pilot across the Seoul National University Hospital network, investigators used a curated 419-gene panel to screen newborns for disorders selected on clinical severity, early onset, an asymptomatic window, and the existence of effective treatment options. The goal was not to replace conventional screening wholesale, but to test whether a genome-based layer could identify infants who would otherwise pass through standard biochemical screening unnoticed. It did.
Genome sequencing was performed on 399 newborns, with six monozygotic twin pairs counted as single data points. The cohort included 385 inborn neonates and 14 infants referred after abnormal conventional newborn screening results. Among 393 newborns included in the screen-positive analysis, the positive rate was 1.3% (5/393). Two of those positive cases involved variants in SLC25A15 and MYO15A, both missed by conventional screening. Three additional cases carried pathogenic variants in arrhythmia-associated genes, a category that conventional newborn screening is not designed to detect at all.
Those findings are clinically consequential because they expose the boundary of biochemical screening rather than its failure. Conventional newborn screening is built to capture a limited set of metabolic and inherited disorders through measurable analytes or physiologic surrogates. Sequencing expands that aperture to disorders that may not generate a newborn biochemical signature but still carry immediate implications for surveillance, confirmatory testing, counseling, and early intervention. In practice, the genome can reveal disease risk before symptoms emerge, and for some conditions that window may be the only chance to prevent irreversible injury.
One sentence captures the operational reality: genomic screening found actionable disease biology beyond the reach of standard newborn testing.
A 419-gene panel shaped by treatability and early-onset risk
The gene list was not assembled as a broad fishing expedition. A multidisciplinary expert team curated the 419 genes using criteria that favored early childhood onset, disease severity, an asymptomatic interval, and local treatment availability. That framework matters because the strongest argument for newborn screening has always been tied to actionability. Finding a disease early has limited value if no intervention exists or if treatment cannot meaningfully alter the trajectory.
This is where genomic screening differs from classic population testing while still inheriting its public health logic. Wilson and Jungner’s screening principles remain embedded in the design: identify conditions with a recognizable presymptomatic stage, and intervene before harm becomes fixed. Sequencing simply changes the detection method. Instead of waiting for a biochemical abnormality to surface, clinicians can interrogate the genome for pathogenic variants linked to conditions that may still be silent at birth. The Seoul pilot shows that this model can be operationalized in a hospital network, not just discussed in theory.
The inclusion criteria also explain why arrhythmia genes were part of the panel. Sudden cardiac conditions can be clinically silent in the neonatal period, yet pathogenic variants may still dictate future monitoring or family evaluation. Conventional screening has no practical way to capture those infants unless symptoms appear. Genomic screening, by contrast, can identify inherited risk before a first clinical event, which is precisely where preventive pediatrics has its greatest leverage.
Negative genomic results after abnormal conventional screening point to a different kind of utility
A separate and equally important signal came from the 49 neonates with abnormal conventional newborn screening results. All of them were negative on genomic screening. That result does not diminish the value of conventional screening; it clarifies the complementary role of sequencing in a dual-track system. Biochemical screening remains essential for disorders defined by metabolite abnormalities, enzyme deficiencies, or transport defects that may not always be cleanly resolved by DNA analysis alone. Genomic screening, meanwhile, can help sort out abnormal biochemical findings and may reduce diagnostic ambiguity in selected cases.
For families and clinicians, this distinction affects the first weeks of life, when uncertainty can be overwhelming. A positive biochemical screen often triggers urgent confirmatory testing and counseling. A negative genomic result in the setting of an abnormal conventional screen may narrow the differential, but it does not eliminate the need for disorder-specific follow-up. What it does offer is a second lens on risk, one that can sometimes prevent overinterpretation of ambiguous metabolic signals or redirect attention toward non-genetic causes.
The cohort data suggest that the sequencing layer is not merely additive. It can be clarifying. In a newborn screening system already built around urgent decision-making, clarity has measurable clinical value even when it does not change the initial positive flag.
A 26-day turnaround still leaves room for faster clinical action
The median turnaround time from sample collection to final report was 26 days. In a newborn screening context, that is a meaningful operational benchmark, but it is also a reminder that implementation is not yet equivalent to bedside immediacy. For some conditions, four weeks is an acceptable interval for risk stratification and outpatient planning. For others, especially those with narrow treatment windows or rapidly evolving symptoms, a 26-day median can feel long.
Speed will matter even more as genomic NBS moves from pilot settings to broader adoption. Sequencing pipelines, variant interpretation, reporting frameworks, and post-test counseling all have to function reliably at scale. A genomic result is only as useful as the system that can return it in time for intervention. That is why the study’s emphasis on multidisciplinary curation is more than an administrative detail. Variant classification changes over time, gene-disease validity evolves, and the clinical meaning of a finding can shift as evidence accumulates. Regular review is not optional if genomic newborn screening is going to remain clinically credible.
The logistics are as important as the biology. Without a reliable workflow, even accurate sequencing risks becoming a delayed research artifact rather than a screening tool.
Longitudinal follow-up will decide how far newborn sequencing can go
The strongest conclusion from the SNUseq Baby Project is not that genomic newborn screening should replace conventional programs. It is that sequencing can uncover clinically relevant conditions beyond the scope of biochemical screening while preserving a favorable precision profile in a pilot setting. The study also shows that curation and implementation shape performance just as much as the sequencing technology itself. A panel chosen too broadly would invite noise; one chosen too narrowly would miss disease.
Long-term follow-up will be essential. Newborn screening is judged not only by how many cases it detects, but by whether early detection changes outcomes, family decisions, and downstream utilization of medical care. That endpoint cannot be answered by a single pilot report. Infants with pathogenic variants will need longitudinal observation to determine how often genomic findings alter care in a durable way, whether early surveillance prevents symptoms, and how families respond to results that may forecast disease before any clinical sign appears.
For now, the Seoul cohort provides a practical demonstration that genomic screening can add information conventional newborn screening cannot reach. The unanswered issues are embedded in the implementation details: which genes belong on a screening panel, how quickly results can be returned, how variant reinterpretation should be handled, and how much clinical benefit emerges after the first report lands in the chart.