Primary carnitine deficiency is an autosomal recessive disorder caused by loss-of-function variants in the SLC22A5 gene, which sits on chromosome 5q31.1 and encodes a protein called organic cation transporter 2, or OCTN2. This transporter is the gatekeeper of carnitine homeostasis: it pumps carnitine into cells across the plasma membrane and reabsorbs it in the renal proximal tubules, maintaining intracellular free carnitine concentrations 20 to 50 times higher than those outside the cell. When both copies of the gene are disabled, carnitine floods out of the body in urine and cells are starved of a molecule they desperately need to import long-chain fatty acids into mitochondria for beta-oxidation, the metabolic pathway that supplies energy during fasting and prolonged exertion.

The clinical consequences are stark. Roughly half of affected children develop hypoketotic hypoglycemia between 3 months and 2 years of age, a dangerous combination of low blood sugar and absent ketone bodies that can spiral into hepatic encephalopathy, lethargy, and hyperammonemia. The other half present between 2 and 4 years old with skeletal muscle weakness, poor exercise tolerance, cardiomyopathy, and arrhythmias. In severe cases, sudden death can occur, and multiple unexplained deaths among young adults in the Faroe Islands, where the disorder affects as many as 1 in 297 people, have been attributed to untreated primary carnitine deficiency. Yet the condition is also one of the most treatable inherited metabolic disorders known: early diagnosis and oral L-carnitine supplementation, typically at 50 to 100 milligrams per kilogram per day, can prevent irreversible damage entirely.

The screening pipeline described in the study relies on tandem mass spectrometry, or MS/MS, to measure free carnitine, abbreviated C0, in dried blood spots collected from newborns within 3 days of birth. Following Chinese expert consensus, the team used a cutoff of 10 micromoles per liter. Of the 439,486 newborns screened over ten years, 1,518, or 0.35 percent, fell below that threshold on the initial test. Nearly all of them, 1,517 infants, were recalled for re-screening, and 154 showed persistently low C0 levels. Of the 152 who completed full diagnostic evaluation, 20 received a clinically relevant diagnosis, a diagnostic yield of 13.16 percent. Seventeen were diagnosed with primary carnitine deficiency, yielding a screening-detected prevalence in Yancheng of 1 in 25,852, while three cases traced back to maternal carnitine deficiency transmitted through the placenta rather than an inherited defect in the infant.

The numbers reveal the central challenge of screening for this disease. Carnitine crosses the placenta freely during pregnancy, and neonatal levels are strongly shaped by maternal factors, so a normal mother can partially buffer a genuinely affected newborn, keeping the infant’s C0 above the minimum value on day 3. Physiological carnitine decline during pregnancy, premature sample collection, insufficient intake from exclusive breastfeeding, prematurity, low birth weight, and cesarean section all inflate the false-positive rate. The statistical analysis made this tension visible: the PCD group’s median C0 fell from 7.33 micromoles per liter at initial screening to 4.98 at re-screening, a significant drop, while the false-positive group’s median barely moved from 8.45 to 8.22, a non-significant change. The maternal deficiency group, unusually, rose from 4.33 to 4.98. But the ranges overlapped across all three groups, which means C0 measurements alone cannot settle a diagnosis.

That is where genetics entered the picture. The team sequenced the exons and flanking intronic regions of SLC22A5 in recalled newborns using the DNBSEQ-T7 platform at an average depth of at least 200-fold, then validated every variant by Sanger sequencing and classified pathogenicity under American College of Medical Genetics and Genomics guidelines. Fifteen of the 17 PCD infants carried two pathogenic variants, four of them homozygous, giving molecular confirmation of the biochemical picture. Across these patients, 32 variants spanning 10 distinct sites were identified, all in exonic regions, and missense changes dominated. One variant stood out: c.1400C>G, which swaps a serine for a cysteine at position 467 of the OCTN2 protein, accounted for 13 of the 32 variants, or 40.63 percent, making it the clear hotspot in this region of Jiangsu Province.

The hotspot finding carries real practical weight, because SLC22A5 variant spectra differ sharply across ethnic groups and regions. The common variant in Italy is c.505C>T; in the United States it is c.136C>T; and c.1400C>G ranks among the most frequent in Japan and has not been reported in non-East Asian populations. Large-scale Chinese studies have flagged c.1400C>G, c.51C>G, and c.760C>T as common, but with regional variation. In Yancheng, c.1400C>G towered over the second most frequent variant, c.51C>G at 21.88 percent, while c.760C>T appeared only once. Knowing which variants dominate locally allows laboratories to build faster, cheaper targeted molecular probes rather than sequencing the entire gene for every recalled infant.

The study also confronted the limits of genetic testing head-on. Two of the 17 PCD infants carried only one detectable variant, yet their C0 levels stayed below 10 micromoles per liter across three consecutive measurements with adequate feeding, their mothers had normal carnitine levels, and no secondary cause could be found, supporting a clinical diagnosis of PCD. This is not unusual: sequencing that covers only coding regions and 10 base pairs of flanking intronic sequence cannot detect variants in regulatory or deep intronic regions, nor methylation changes. Previous research found pathogenic variants in only 84 percent of alleles among people with impaired fibroblast carnitine transport, and in one large series only 56.4 percent of suspected PCD individuals carried biallelic variants. The authors therefore insist that diagnosis must rest on combined biochemical and genetic evaluation rather than either alone.

Follow-up, ranging from 18 months to 10 years, showed the payoff of early detection. None of the diagnosed children experienced acute metabolic decompensation, and all but one showed normal intellectual, motor, and growth development; the exception had poor exercise tolerance. Four families chose not to medicate their asymptomatic children but committed to avoiding fasting, infection, and prolonged strenuous exercise, the classic triggers of metabolic crisis. Among the three maternal deficiency cases, two mothers were themselves found to carry the c.1400C>G variant in homozygous form and were diagnosed with PCD, underscoring how newborn screening can cascade into diagnoses in asymptomatic adults who remain at risk of sudden cardiac death.

Perhaps the most forward-looking result came from a single recurrence-risk family. Their previous child carried two heterozygous SLC22A5 variants, c.865C>T, a premature stop codon, and c.1400C>G, inherited from the mother and father respectively. Sanger sequencing of amniotic fluid DNA in the next pregnancy revealed only the maternal variant, meaning the fetus had an extremely low probability of disease given the autosomal recessive inheritance and the parents’ normal phenotypes. The couple continued the pregnancy, and the newborn’s screening C0 fell within the normal range with normal follow-up. The authors are careful to frame this as a case illustration rather than general evidence, but it demonstrates that SLC22A5 testing can anchor prenatal diagnosis and informed reproductive counseling for families who already know the fear of this disease, offering certainty at a moment when it matters most.

Subject of Research: Newborn screening and prenatal genetic diagnosis of primary carnitine deficiency caused by SLC22A5 variants

Article Title: Newborn Screening of Primary Carnitine Deficiency and Prenatal Genetic Diagnosis for a Recurrence‐Risk Family

Article References: Sun, H., Zheng, J., Zhou, C., Chen, H., Zhang, L., Li, X., Bao, D., Yang, F., & Liu, J. (2026). Newborn Screening of Primary Carnitine Deficiency and Prenatal Genetic Diagnosis for a Recurrence‐Risk Family. Molecular Genetics & Genomic Medicine, 14(10), Article e70316. https://doi.org/10.1002/mgg3.70316

Image Credits: AI Generated

DOI: 10.1002/mgg3.70316

Keywords: primary carnitine deficiency, newborn screening, SLC22A5, OCTN2, tandem mass spectrometry, prenatal diagnosis, genetic testing, carnitine transporter, hypoketotic hypoglycemia, L-carnitine supplementation, autosomal recessive inheritance, China

Mary Ellen Avery’s legacy provides the intellectual template for this kind of work. In the 1950s, Avery demonstrated that respiratory distress syndrome in premature newborns was caused by a deficiency of pulmonary surfactant, the lipoprotein mixture that keeps the tiny air sacs of the lung from collapsing with each breath. That insight, derived from careful basic science, eventually produced surfactant replacement therapy, one of the most consequential interventions in modern neonatology. The lesson embedded in her career is that a mechanistic discovery, pursued patiently over decades, can be converted into a system delivered to every infant who needs it. It is precisely that conversion, from knowledge to infrastructure, that the award celebrates.

Chung’s own work sits at the intersection of genetics, pediatrics, and public health. As a human geneticist who has led large-scale genomic medicine initiatives, she has focused on a question that sounds simple but is operationally formidable: which conditions should be looked for in every newborn, and how should the results be delivered to families in a way that changes outcomes rather than merely generating anxiety? The answer determines whether a diagnosis arrives in time to alter the course of a disease or only in time to explain it retrospectively.

The clearest recent example comes from spinal muscular atrophy, a genetic disorder that was, until recently, the leading inherited cause of infant mortality. SMA is caused by loss or mutation of the SMN1 gene, which produces a protein survival motor neurons require. Infants with the most severe form lose motor neurons progressively in the first months of life, and before effective therapy existed, most died of respiratory failure before their second birthday. The disease, however, has a crucial property that makes it an ideal target for newborn screening: intervention is dramatically more effective when it begins before symptoms appear, because motor neurons that have not yet been lost can be preserved.

The therapeutic landscape changed with the development of disease-modifying treatments, most prominently nusinersen, an antisense oligonucleotide that alters splicing of the related SMN2 gene so that it produces more functional survival motor neuron protein. The pivotal randomized trial published in the New England Journal of Medicine in 2017 showed that infants treated with nusinersen had significantly better motor-function outcomes than those receiving a sham control procedure. Suddenly, a fatal diagnosis became a treatable condition, but only if it was identified early enough. Screening became the bottleneck: without systematic detection at birth, infants would still present with established weakness, at which point the therapeutic window had narrowed.

Chung and colleagues addressed this directly in New York State, where a pilot study of population-based newborn screening for spinal muscular atrophy demonstrated that molecular screening could be integrated into existing public health infrastructure. The technical approach relies on detecting the homozygous deletion of SMN1 in dried blood spots, the same heel-prick specimens already collected from every newborn for dozens of metabolic and genetic conditions. The pilot study, published in Genetics in Medicine, established feasibility, laboratory workflow, and the clinical pathway connecting a positive screen to confirmatory testing and rapid treatment initiation. New York subsequently moved from pilot to universal screening, and the model has since been adopted across a growing number of states and countries.

The same logic has been extended to Duchenne muscular dystrophy, another severe neuromuscular condition in which early identification may enable earlier access to emerging therapies. A two-year pilot study of newborn screening for Duchenne in New York, reported in Annals of Clinical and Translational Neurology, evaluated the performance of screening for elevated creatine kinase and confirmatory genetic testing in dried blood spots. These pilots illustrate a recurring pattern in Chung’s career: rigorous, unglamorous implementation science that determines whether a screening concept can actually function at population scale, with acceptable rates of false positives, clear follow-up protocols, and equitable reach across diverse communities.

Chung’s most ambitious project in this space is the GUARDIAN study, a genomic newborn screening program conducted in New York that screened tens of thousands of infants for hundreds of genetic conditions, many of which are actionable in early childhood. Unlike traditional newborn screening, which targets a fixed panel of biochemical markers, genomic screening reads the infant’s DNA directly, allowing a single platform to detect conditions ranging from metabolic disorders to hearing loss to cardiac arrhythmia syndromes. The study received in-kind contributions from GeneDx and Illumina, and its results are shaping the debate over how much genetic information should be gathered at birth and how families should be counseled about findings that range from urgently actionable to uncertain.

The GUARDIAN study also raises the ethical and logistical questions that any system of universal early care must confront. Screening an entire population inevitably identifies infants with variants of uncertain significance, conditions that may never cause symptoms, and incidental findings unrelated to the reason for testing. The design of the program, with defined gene panels focused on childhood-onset, actionable conditions, reflects an attempt to balance the promise of early intervention against the risks of overdiagnosis and medicalization of healthy infancy. Chung’s framing of the work, as building systems of care rather than simply tests, emphasizes that a screen is only as valuable as the clinical pathway behind it: confirmatory diagnostics, genetic counseling, treatment access, and long-term follow-up.

The deeper argument of the Avery Award lecture is historical and structural. Medicine tends to organize itself around the sick patient who arrives at the clinic, but the greatest gains in pediatric health have come from systems that reach children before illness declares itself: vaccination schedules, congenital hypothyroidism screening, hearing screening, pulse oximetry for critical congenital heart disease, and now genomic screening. Each of these required decades of persistence, coalition-building among scientists, public health officials, advocacy groups, and families, and a willingness to demonstrate value through pilot data before universal implementation. Chung’s career, from her leadership in identifying the genetic basis of conditions to her role in designing and executing population screening programs, exemplifies the long gestational period that careers in physician science often require before the birth of innovation reaches patients. In honoring that journey, the Society for Pediatric Research signals to the next generation of pediatrician-scientists that the highest-impact work may not be a single discovery but the patient construction of systems that ensure every child, starting at birth, benefits from what medicine already knows.

Subject of Research: Newborn screening and systems of early pediatric care, including genomic screening for treatable genetic conditions

Article Title: The journey to establishing systems of care for all children, starting at birth: the Society for Pediatric Research’s 2026 Mary Ellen Avery award

Article References: Chung, W. K. (2026). The journey to establishing systems of care for all children, starting at birth: the Society for Pediatric Research’s 2026 Mary Ellen Avery award. Pediatric Research. https://doi.org/10.1038/s41390-026-05580-6

Image Credits: AI Generated

DOI: 10.1038/s41390-026-05580-6

Keywords: newborn screening, spinal muscular atrophy, Mary Ellen Avery, Wendy Chung, GUARDIAN study, genomic medicine, nusinersen, Duchenne muscular dystrophy, pediatric research, public health, SMN1, early intervention

The technical logic behind the study rests on a well-established principle of population genetics. In small, historically endogamous populations, founder effects can elevate the frequency of otherwise rare disease alleles to levels where homozygosity becomes common. When two carriers of the same recessive variant have children, each pregnancy carries a 25 percent chance of producing an affected child. In admixed continental populations, such variants typically lurk at minor allele frequencies below one percent, making homozygous disease vanishingly rare. But in Nunavut, the numbers tell a different story. The SI c.273_274delAG variant, which abolishes the function of the sucrase-isomaltase enzyme needed to digest sucrose and starch, reached a carrier frequency of one in four, yielding an estimated live birth prevalence of one in 28 for CSID. The IFNAR2 c.157T>C variant, which disrupts the type I interferon receptor and leaves affected infants unable to mount innate antiviral responses, showed a carrier frequency of one in 18 and a live birth prevalence of one in 737. The DNAH11 c.4095+2C>A splice-site variant, responsible for a form of primary ciliary dyskinesia that impairs mucociliary clearance in the airways, appeared at a carrier frequency of one in 13 with a live birth prevalence of one in 1,474.

These are not abstract statistics. Each condition carries concrete, sometimes devastating consequences for infants and toddlers. IFNAR2 deficiency is perhaps the most dramatic: children who inherit two copies of the c.157T>C variant lack a functional receptor for type I interferons, the molecular alarm system that cells use to signal viral infection. The clinical consequence is extreme vulnerability to live attenuated viral vaccines—measles-mumps-rubella-varicella among them—because the weakened vaccine viruses, which a healthy immune system clears effortlessly, can cause severe disseminated disease in these children. The condition is functionally analogous to severe combined immunodeficiency in its implications for vaccination policy, yet it arises from a completely different molecular lesion and would be missed entirely by standard newborn screening for SCID, which relies on T-cell receptor excision circle (TREC) measurements that remain normal in IFNAR2-deficient infants. Identifying affected children before vaccination is therefore a matter of life and death, and the new prevalence estimate of one in 737 live births transforms the risk calculus for immunization programs across the territory.

Primary ciliary dyskinesia caused by the DNAH11 variant presents a different diagnostic challenge. The DNAH11 gene encodes a heavy-chain dynein motor protein essential for the beating motion of cilia, the microscopic hair-like structures that sweep mucus, bacteria, and debris out of the respiratory tract. When the c.4095+2C>A variant disrupts splicing of the gene’s transcript, ciliary motility fails, and affected children accumulate recurrent sinopulmonary infections, chronic wet cough, and neonatal respiratory distress. Without a high index of suspicion, PCD is frequently misdiagnosed as recurrent ordinary pneumonia or asthma, delaying the airway-clearance therapies and aggressive antibiotic strategies that preserve lung function. In a region where respiratory illness already places an outsized burden on pediatric health services and where medevac to southern tertiary centers is often the only route to specialized diagnostics, a prevalence of one in 1,474 live births means that PCD is no longer a zebra diagnosis but a realistic consideration for any Nunavut infant with unexplained neonatal respiratory distress or chronic productive cough.

The most numerically dominant condition, congenital sucrase-isomaltase deficiency, emerged from the study with particular clarity. Eighty-three children in the linked subset carried two copies of the SI c.273_274delAG variant, and their health records confirmed phenotypes consistent with the diagnosis. CSID impairs the brush-border enzymes of the small intestine that break down sucrose and starch, so affected infants develop chronic watery diarrhea, abdominal distension, failure to thrive, and irritability once sucrose and starches enter the diet—typically with the introduction of fruits, juices, and cereals in the first year of life. The symptoms mimic common conditions such as toddler’s diarrhea, cow’s milk protein allergy, and celiac disease, and affected children frequently undergo repeated investigations before the true cause is identified. The study’s quantitative analysis showed that CSID was associated with a significant increase in health care utilization for gastrointestinal illness visits before age two, with an odds ratio of 1.9 (95 percent confidence interval 1.2 to 3.0). In practical terms, children with two copies of the SI variant were nearly twice as likely to require health visits for gastrointestinal complaints during the toddler years, straining families and nursing stations alike.

Yet the study also delivered an unexpectedly hopeful finding. When the researchers examined the health visit patterns of CSID cases, they found that breastfeeding for up to six months was protective, associated with a reduced number of health care visits, with an odds ratio of 0.80 (95 percent confidence interval 0.63 to 1.0). The biological rationale is straightforward: breast milk contains virtually no sucrose or starch, so exclusively breastfed infants with CSID are spared exposure to the substrates their intestines cannot digest during the most vulnerable months of early life. The finding carries immediate public health relevance, reinforcing existing breastfeeding promotion efforts in Nunavut while adding a genetically informed dimension: for infants who later prove to have CSID, prolonged breastfeeding may meaningfully blunt the severity of early symptoms and reduce the cascade of medical encounters that otherwise accompanies diagnosis.

Methodologically, the study is notable for how it was conducted as much as for what it found. The research group combined local and national expertise, integrating Inuit governance through Nunavut Tunngavik Incorporated, the territorial Department of Health, the Qaujigiartiit Health Research Centre, and clinical and laboratory specialists from institutions including the Children’s Hospital of Eastern Ontario, Newborn Screening Ontario, the University of British Columbia, and the University of Manitoba. Dried newborn screening blood spots—already collected routinely from every infant—were genotyped, an approach that sidesteps the logistical barriers of recruiting families for fresh blood draws across scattered Arctic communities. A subset of genotyped children was then linked to a demographic and health database built from medical records spanning birth to five years, allowing the team to verify that genotypes predicted the clinical phenotypes expected for each condition. Live birth prevalence and carrier frequencies were first estimated across the full genotyped cohort of 2,947 infants and then recalculated for the linked subset, with the two approaches yielding consistent results.

The work also fits into a broader infrastructure effort. The genotyping and analysis were funded through the Silent Genomes Project, supported by Genome Canada, Genome BC, and the Canadian Institutes of Health Research, an initiative aimed at building an Indigenous Background Variant Library—a reference resource that catalogs benign and disease-associated variants in Indigenous populations so that clinical geneticists can distinguish pathogenic findings from population-specific benign variation. Without such reference data, Indigenous patients undergoing clinical genome sequencing face a disproportionate burden of variants of uncertain significance, a problem the Nunavut study directly helps to address by documenting which variants in this population are common, expected, and, when homozygous, clinically consequential.

The authors’ conclusion is unambiguous: genetic testing, practitioner education, and Inuit-informed management strategies are urgently needed to address these conditions and improve outcomes. For IFNAR2 deficiency, that means developing pathways to identify affected infants before they receive live attenuated viral vaccines, potentially through carrier screening or targeted newborn testing, and educating immunization providers across the Arctic. For PCD, it means lowering the diagnostic threshold for referral to ciliary function testing in infants with compatible respiratory histories. For CSID, it means recognizing that a child in Nunavut with chronic diarrhea and failure to thrive has roughly a one-in-28 baseline likelihood of the condition—odds that justify empirical trials of sucrase enzyme replacement or sucrose-restricted diets far earlier than would be considered elsewhere. Each of these interventions is inexpensive relative to the costs of delayed diagnosis, repeated medevacs, and preventable morbidity.

Beyond Nunavut, the study sends a message to genomic medicine worldwide. Reference genomes, variant databases, and screening panels built predominantly on European-ancestry populations systematically underperform for everyone else, and the Inuit case demonstrates the scale of the blind spot: three recessive conditions, each effectively invisible in generic screening programs, collectively affecting on the order of one in 25 live births in this cohort when CSID’s high prevalence is included. Population-specific genomics, governed by and conducted with the communities it describes, is not a niche academic exercise—it is the difference between a health system that anticipates the diseases its children actually carry and one that discovers them one crisis at a time. The Nunavut cohort study provides both the evidence and, increasingly, the template for how such work should be done.

Subject of Research: Population-specific autosomal recessive genetic variants and their child health impacts in Inuit children in Nunavut

Article Title: Population-specific genetic variants and related health outcomes in a cohort of Inuit children in Nunavut

Article References: Gonzalez-Lema, J., Allen, J., Agyemang, E., Patterson, M., Miners, A., Sheffield, H., Caughey, A., Pham-Huy, A., Kovesi, T., Shapiro, A. J., Kernohan, K., Yeh, E., Yeh, A., Mears, A., Lacaria, M., Collins, S., Braschel, M., Marcadier, J., Otley, A., … Arbour, L. (2026). Population-specific genetic variants and related health outcomes in a cohort of Inuit children in Nunavut. BMC Pediatrics. https://doi.org/10.1186/s12887-026-07720-7

Image Credits: AI Generated

DOI: 10.1186/s12887-026-07720-7

Keywords: Inuit, Nunavut, genetic variants, IFNAR2 deficiency, primary ciliary dyskinesia, DNAH11, congenital sucrase-isomaltase deficiency, founder effects, newborn screening, autosomal recessive, child health, Indigenous genomics

Hemophilia is an inherited bleeding disorder caused by deficiencies in clotting factor VIII, in the case of hemophilia A, or factor IX, in the case of hemophilia B. Because these proteins are essential for forming stable blood clots, affected individuals bleed for longer than normal after injury and can bleed spontaneously into joints, muscles, and, most dangerously, the brain. The disorder is passed on through the X chromosome, so it overwhelmingly affects boys, while girls are typically carriers. What makes the neonatal period so treacherous is the combination of an immature coagulation system, the mechanical stresses of labor and delivery, and the fact that many affected infants are born into families with no known history of the disease. In roughly a third of cases, hemophilia appears in a child with no prior family history, arising from a new genetic mutation, which means neither obstetricians nor parents may anticipate the bleeding risk before delivery.

Intracranial hemorrhage in a newborn is a medical emergency of the highest order. Bleeding can occur in several compartments of the skull: inside the brain tissue itself, in the subdural space beneath the dura mater, in the subarachnoid space surrounding the brain, or within the ventricles, the fluid-filled cavities deep in the brain. In a baby with hemophilia, even a minor trauma during delivery, such as the pressure of forceps, a vacuum extraction, or a prolonged second stage of labor, can trigger bleeding that a healthy infant’s clotting system would quickly contain. Symptoms may include seizures, lethargy, vomiting, a bulging fontanelle, pallor, or simply poor feeding and irritability, and because these signs overlap with many other neonatal conditions, diagnosis is often delayed. Untreated, the bleed can expand, compressing brain tissue and causing permanent damage or death.

The true scale of the problem has long been difficult to pin down. Previous estimates of neonatal ICH risk in hemophilia have ranged widely across the literature, from figures below one percent to reports suggesting rates several times higher. The authors of the new analysis attribute this fragmentation to heterogeneous study designs, small single-center cohorts, differing diagnostic criteria, and inconsistent surveillance practices. Some studies relied on clinical symptoms alone, others on routine cranial imaging, and the era in which each study was conducted matters as well, since the availability of neonatal ultrasound, computed tomography, and magnetic resonance imaging has changed dramatically over recent decades. By systematically searching PubMed, Embase, and Scopus from their inception through November 2025 and applying consistent eligibility criteria, requiring reported ICH within the first twenty-eight days of life among live-born neonates with hemophilia, the team sought to produce the most reliable estimate to date.

The statistical machinery behind the study is standard for modern evidence synthesis but worth unpacking. Rather than simply averaging the results of the thirteen included studies, the researchers used a random-effects meta-analysis, a model that assumes each study estimates its own true incidence and that these true values are distributed around a common mean. This approach is appropriate when between-study heterogeneity is high, which the authors confirmed it was. Heterogeneity in this context means the studies differed substantially in their reported rates, likely reflecting differences in geography, era, severity of hemophilia represented, mode of delivery, and intensity of diagnostic workup. The random-effects model widens the confidence interval to account for this variability, which is why the pooled estimate of 2.10 percent carries an interval stretching from 1.40 to 2.79 percent rather than a razor-thin range.

One of the more sobering findings concerns mortality. The pooled estimate for fatal neonatal ICH was approximately 0.2 percent, with a 95 percent confidence interval of 0.0 to 1.2 percent. In other words, among every thousand live-born neonates with hemophilia, roughly two may die from a brain bleed in the first month of life, though the wide interval reflects the rarity of the event and the limits of the available data. Fatal bleeds were rare in absolute terms, but the analysis underscores that when ICH does occur, it carries the potential for catastrophic outcomes. Survivors of neonatal ICH can face cerebral palsy, epilepsy, developmental delay, and cognitive impairment, making prevention far more valuable than even the best acute treatment.

Prevention, the authors argue, hinges on three pillars: early diagnosis, careful perinatal management, and avoidance of traumatic delivery techniques. Early diagnosis means identifying at-risk infants before or immediately after birth, ideally through known family history, carrier testing, or, in some health systems, targeted newborn screening of male infants born to carrier mothers. Careful perinatal management involves coordinating obstetric and hematology teams well before delivery, planning the mode of birth to minimize trauma, and avoiding invasive monitoring such as fetal scalp electrodes or vacuum and rotational forceps deliveries when a baby is known or suspected to have hemophilia. For infants diagnosed after birth, prompt replacement of the missing clotting factor can halt bleeding before it spreads, and some guidelines recommend prophylactic factor replacement for high-risk neonates during the vulnerable first days of life.

The study’s findings arrive amid a broader shift in hemophilia care. Prophylactic clotting factor infusions, once reserved for older children with severe disease, are increasingly considered earlier, and newer agents, including extended half-life factor products and bispecific antibodies that mimic factor VIII activity, have transformed long-term management. Yet the neonatal window remains a blind spot in many treatment pathways, because diagnosis itself may lag behind the highest-risk period. The authors call for dedicated neonatal-focused registries and prospective studies to refine risk stratification, arguing that only systematically collected, individual-level data can identify which babies are most likely to bleed and which preventive strategies work best. Such registries could track delivery mode, factor levels at birth, family history, imaging findings, and outcomes in a standardized way across centers and countries.

The meta-analysis also highlights persistent gaps in the evidence base. The thirteen included studies spanned different decades, healthcare settings, and diagnostic standards, and the authors note that the substantial heterogeneity between them limits the precision of any single pooled estimate. Retrospective designs dominated, and few studies distinguished between severity categories of hemophilia, modes of delivery, or the timing and type of diagnostic imaging. Without prospective, standardized surveillance, some bleeds may go undetected, particularly small hemorrhages that produce subtle or absent symptoms in the first days of life. This means the true incidence could, in principle, be higher than the pooled estimate suggests, a possibility the authors acknowledge indirectly through their call for better data.

For clinicians and families, the practical message is clear. A 2.10 percent risk is low enough that most newborns with hemophilia will never experience a brain bleed, but high enough that every birth should be treated as an opportunity for prevention. When hemophilia is known or suspected before delivery, obstetric teams can choose the least traumatic route of birth, neonatologists can maintain a low threshold for cranial imaging, and hematologists can have replacement factor ready at the bedside. When there is no family history, the challenge is harder, but the growing availability of genetic carrier screening and heightened awareness among providers can shorten the path from first symptom to diagnosis and treatment. The study, published open access on October 6, 2026, and citable under DOI 10.1186/s12887-026-07750-1, adds a carefully quantified data point to a field that has long operated on fragmented numbers, and it makes a compelling case that the first twenty-eight days of life deserve far more systematic attention in the care of children with hemophilia.

Subject of Research: Neonatal intracranial hemorrhage risk in infants with hemophilia

Article Title: Risk of intracranial hemorrhage in neonates with hemophilia: a systematic review and meta-analysis

Article References: Zheng, M., Zheng, H., Li, Y., & Tu, J. (2026). Risk of intracranial hemorrhage in neonates with hemophilia: a systematic review and meta-analysis. BMC Pediatrics. https://doi.org/10.1186/s12887-026-07750-1

Image Credits: AI Generated

DOI: 10.1186/s12887-026-07750-1

Keywords: hemophilia, intracranial hemorrhage, neonatology, meta-analysis, systematic review, clotting factors, bleeding disorders, perinatal care, newborn screening, prophylaxis, BMC Pediatrics, birth trauma

The study, published in BMC Pediatrics, was led by a team at the Women and Children’s Medical Center of Shenzhen Maternity and Child Healthcare Hospital affiliated with Southern Medical University. The investigators started with an unusually complete dataset: 2,652 deaths of children under five years of age recorded in Shenzhen between January 2016 and December 2023. From this mortality cohort, they applied prespecified, record-based cause-of-death exclusions and removed deaths occurring within 72 hours of birth, reasoning that very early neonatal deaths typically involve distinct clinical circumstances and are less likely to reflect undiagnosed metabolic disease. That filtering left 1,982 children eligible for metabolic investigation.

The critical resource was the archived dried blood spots, the small cards of filter paper collected from nearly every newborn in China as part of routine screening programs. These spots preserve a biochemical snapshot of the infant’s blood in the first days of life. Dried blood spots were available for 652 of the eligible children, and the team analyzed each one for 43 metabolites using tandem mass spectrometry, a technique that can precisely quantify tiny amounts of amino acids, acylcarnitines, and other small molecules in a single rapid run. The screening panel was designed to flag 48 different conditions, using both individual metabolite concentrations and prespecified metabolite ratios, since ratios often discriminate true disease from benign variation more reliably than any single value.

Tandem mass spectrometry works by ionizing the compounds extracted from the blood spot and then sorting them by their mass-to-charge ratio in two sequential stages, which allows the instrument to distinguish closely related molecules such as the various acylcarnitines that accumulate when fatty acid oxidation is impaired. When a fatty acid oxidation enzyme is deficient, the body cannot efficiently break down fats for energy, and specific intermediate acylcarnitines build up in the blood in characteristic patterns. Elevated concentrations of these markers, or abnormal ratios between them, serve as biochemical fingerprints of the underlying enzymatic block. In this cohort, 50 of the 652 tested children were flagged as positive by the MS/MS panel, a striking yield given that none of them had received a metabolic diagnosis before death.

To determine whether these biochemical signals reflected genuine disease, the researchers turned to whole-exome sequencing, a next-generation sequencing approach that reads the protein-coding regions of all roughly 20,000 human genes. Because most known inborn errors of metabolism are caused by variants in genes encoding enzymes, transporters, or cofactor proteins, the exome captures the great majority of the relevant disease space. The team sequenced all 50 MS/MS-positive cases and interpreted the variants according to American College of Medical Genetics and Genomics and Association for Molecular Pathology criteria, the standard framework for classifying sequence variants as pathogenic, benign, or uncertain.

The molecular work delivered definitive answers in 13 of the 50 cases, or 26 percent. Twelve of these were inborn errors of metabolism, and one child turned out to have Smith-Magenis syndrome, a genomic disorder caused by a deletion on chromosome 17 that is not a metabolic disease at all but which had produced a misleading biochemical signal. One additional case was classified as probable ornithine transcarbamylase deficiency, a urea cycle disorder that can cause lethal hyperammonemia but which often eludes both standard screening and postmortem testing. Among the definitively diagnosed metabolic cases, the spectrum included fatty acid oxidation disorders, which emerged as the dominant category, consistent with the known tendency of these conditions to present as sudden, catastrophic illness in a previously healthy child, sometimes triggered by fasting or infection.

The diagnostic categories named in the study’s abbreviations read like a catalogue of rare but devastating pediatric disease: very-long-chain acyl-CoA dehydrogenase deficiency, carnitine-acylcarnitine translocase deficiency, systemic primary carnitine deficiency, multiple acyl-CoA dehydrogenation deficiency, methylmalonic acidemia, pyruvate dehydrogenase deficiency, ethylmalonic encephalopathy, and mucolipidosis II or III alpha/beta. Each of these conditions has a distinct biochemical signature and a distinct genetic cause, and each is potentially detectable, and in some cases treatable, if identified in time. The fact that they surfaced only after death underscores the central argument of the paper: expanded newborn screening is not universally implemented, and where it is absent or incomplete, lethal metabolic disease can pass through the health system entirely unrecognized.

One of the most clinically informative findings emerged from the comparison between the 13 definitively diagnosed cases and the 37 MS/MS-positive cases that lacked a definitive molecular diagnosis. The diagnosed children had significantly lower rates of low birth weight, 31 percent versus 84 percent; preterm birth, 15 percent versus 78 percent; and obstetric complications, 46 percent versus 78 percent, with P values below 0.001 for the first two comparisons and 0.040 for the third. In other words, the children whose deaths were ultimately explained by inherited metabolic disease were far more likely to have been born at term with normal birth weight and uncomplicated pregnancies. This pattern suggests that a metabolic cause of death should rise higher on the differential diagnosis when a seemingly healthy, full-term baby deteriorates or dies without an alternative explanation, whereas MS/MS abnormalities in preterm or low-birth-weight infants may more often reflect the physiological turbulence of prematurity rather than true inherited disease.

The study also examined how well the biochemical and genetic evidence agreed with each other, a question of real practical importance because discordant findings complicate diagnosis and family counseling. Among the 13 definitively diagnosed cases, clear metabolite-genotype concordance was observed in nine, or 69.2 percent, meaning the biochemical pattern and the pathogenic variants pointed to the same disease. Partial concordance was seen in two cases and non-concordance in two. These discordant cases illustrate a well-known limitation of newborn screening panels: metabolite levels can be influenced by age, feeding, liver function, and prematurity, and a positive screen is not a diagnosis. The authors argue that this is precisely where sequencing adds decisive value, establishing molecular diagnoses and clarifying atypical, nonspecific, or biochemically confusing findings that MS/MS alone cannot resolve.

The implications reach well beyond Shenzhen. The study provides direct evidence that retrospective analysis of archived dried blood spots can convert unexplained pediatric deaths into explained ones, offering families a diagnosis, clarifying recurrence risk for future pregnancies, and identifying relatives who may carry the same condition. Just as importantly, it strengthens the case for broader access to expanded MS/MS newborn screening worldwide, since every undiagnosed case in this cohort represents a child whose condition was, in principle, detectable at birth. The authors advocate a risk-stratified approach to genomic testing, reserving whole-exome sequencing for cases that remain clinically or biochemically unresolved after screening. As sequencing costs continue to fall and blood spot archives grow, the combination of metabolomics and genomics may become a standard tool not only for diagnosing living children but for understanding why some of the most vulnerable never survive their first years.

Subject of Research: Detection of undiagnosed inborn errors of metabolism in unexplained early childhood deaths using newborn dried blood spot screening and whole-exome sequencing

Article Title: Inborn errors of metabolism in unexplained early childhood deaths: a retrospective cohort study combining MS/MS and WES

Article References: Zhou, L., Yang, X., Wu, D., Zhang, L., Tu, X., Zhan, Z., Xie, S., Xu, Y., Liu, C., Wu, B., & Liu, W. (2026). Inborn errors of metabolism in unexplained early childhood deaths: a retrospective cohort study combining MS/MS and WES. BMC Pediatrics. https://doi.org/10.1186/s12887-026-07740-3

Image Credits: AI Generated

DOI: 10.1186/s12887-026-07740-3

Keywords: inborn errors of metabolism, newborn screening, tandem mass spectrometry, whole-exome sequencing, dried blood spots, early childhood mortality, fatty acid oxidation disorders, pediatrics, metabolomics, genetic diagnosis, urea cycle disorders, Shenzhen

The ASPIRE Prize, whose full name is the APEC Science Prize for Innovation, Research and Education, is awarded each year to an outstanding early-career scientist who has demonstrated both scientific excellence and a strong commitment to cross-border collaboration within the APEC region. Each member economy may nominate one scientist under the age of 40, and a single winner is chosen through a regional selection process. The structure of the prize is deliberately designed to spotlight researchers whose work transcends national boundaries, and Luo’s selection at this early stage of his academic career underscores the global reach of the computational tools his laboratory has built. According to APEC, the prize represents not only excellence in science and technology but also the organization’s collective aspiration to address global challenges through collaboration and forward-looking leadership.

This year’s award specifically recognizes outstanding contributions that harness artificial intelligence and data science to strengthen industries and build resilient economies for the future, a theme that sits squarely at the center of Luo’s research program. His work focuses on developing AI-powered computational tools that have transformed the analysis of complex and highly diverse genomic data. Genomic datasets are among the largest and most computationally demanding in modern biology: a single human genome consists of roughly three billion base pairs, and sequencing instruments produce fragments that must be assembled, aligned, and interpreted against reference sequences before any biological meaning can be extracted. The algorithms Luo and his collaborators have developed address precisely this bottleneck, and their breakthroughs have made genomic sequencing significantly faster, more accurate, and highly cost-effective than previous approaches allowed.

The practical consequences of that speed and accuracy extend far beyond the laboratory. Faster and cheaper sequencing lowers the cost per genome to the point where large-scale clinical deployment becomes feasible, which in turn opens the door to applications in healthcare, food security, and biotechnology. Luo’s innovations are described as instrumental in translating complex genomic research into practical, clinical solutions for disease prevention, precision medicine, and clinical decision support. In precision medicine, the ability to rapidly characterize a patient’s genome can inform drug selection and dosing; in disease prevention, population-scale sequencing can reveal inherited risk factors before symptoms appear; and in agriculture and food security, genomic tools underpin the breeding of resilient crops and livestock and the monitoring of pathogens that threaten food supplies.

The adoption figures attached to Luo’s work illustrate how a research algorithm can evolve into global infrastructure. The open-source AI tools developed by his laboratory have been downloaded more than 10 million times worldwide, with approximately 60 percent of those downloads originating from other APEC economies. Open-source distribution of this kind is significant for two reasons. First, it means the tools are not locked behind commercial licensing, so research groups, hospitals, and public health agencies across economies with very different resource levels can integrate them into their own pipelines. Second, the concentration of downloads within the APEC region demonstrates the kind of cross-border scientific diffusion that the ASPIRE Prize is explicitly designed to reward, with the laureate’s software functioning as shared technical infrastructure across the very economies the prize seeks to connect.

Those tools are now widely integrated into global healthcare, agricultural, and medical biotechnology industries, and Luo has described how they are moving from research settings into clinical environments. In a statement accompanying the announcement, he said that the algorithms developed by his lab are now being integrated into clinical decision-support systems that can help detect early-stage cancer and rare diseases in newborns. For him, he added, that is what innovation is about: turning research into tools that can help save lives. The two clinical scenarios he identified are among the most demanding in genomic medicine. Early-stage cancer detection requires algorithms sensitive enough to distinguish faint tumor-derived signals from abundant background genetic material, while newborn screening for rare diseases depends on rapid turnaround, because many rare genetic conditions are most treatable within days or weeks of birth.

The technical challenge that Luo’s career has been built around is a familiar one to anyone working in bioinformatics: raw sequencing data is voluminous, noisy, and fragmented, and the quality of every downstream conclusion depends on how well it is processed. Sequence assembly and alignment, the core problems his algorithms target, are computationally hard problems that scale with both genome size and dataset volume. As sequencing technology has shifted toward producing ever larger volumes of data at lower cost, the computational step has increasingly become the limiting factor in genomic workflows. AI-driven approaches, including machine learning methods that can correct errors, resolve repetitive regions, and classify variants, have emerged as a way to push past those limits, and Luo’s laboratory has positioned itself at the leading edge of that shift, producing tools that are simultaneously faster, more accurate, and cheaper to run than the alternatives.

Luo’s path to this recognition reflects the international character of contemporary computational biology. He completed his PhD training in bioinformatics at the University of Hong Kong between 2010 and 2015, working with Professor Tak-Wah Lam, and then undertook postdoctoral training with Professors Steven Salzberg and Michael Schatz at the Center of Computational Biology at Johns Hopkins University from 2016 to 2017. Salzberg and Schatz are among the most prominent figures in genome informatics, and that training environment placed Luo at the intersection of algorithm design and large-scale genome assembly at a moment when the field was being transformed by high-throughput sequencing. He is now an Associate Professor in HKU’s School of Computing and Data Science, in addition to his leadership roles in the AI and Data Science Division, where his work spans bioinformatics algorithms and clinical informatics.

His publication record and the recognition it has attracted trace the trajectory of a researcher whose influence has grown rapidly. Luo has published more than 100 papers, ten of which have each achieved over a thousand citations, a benchmark that places those papers among the most influential in their fields. He has been identified as a Top 1 Percent Scholar Worldwide by Clarivate Analytics every year since 2019, was selected by Baidu Research as one of the Worldwide Top 150 Chinese Young Scholars in AI, was named one of the Top 10 Innovators Under 35 in Asia Pacific by MIT Technology Review in 2019, and was recognized in Forbes’ 30 Under 30 Asia list in the healthcare and science category in 2017. Taken together, these honors chart a career in which algorithmic innovation, clinical application, and international collaboration have advanced in parallel rather than in sequence.

For the Asia-Pacific research community, the 2026 ASPIRE Prize carries a message about where the region’s scientific priorities are heading. APEC’s framing of this year’s award, emphasizing artificial intelligence and data science as tools for strengthening industries and building resilient economies, signals that the organization sees computational biology not as a niche discipline but as strategic infrastructure for public health, agriculture, and biotechnology across its member economies. Luo’s own account of his motivation, that innovation matters when research becomes a tool that can help save lives, captures the ethos the prize is intended to celebrate. With his open-source tools already embedded in clinical decision-support systems and downloaded millions of times across the region, the 2026 laureate offers a concrete example of how early-career research, when built openly and aimed at real clinical problems, can ripple outward into healthcare systems and industries far beyond the laboratory where it began.

Subject of Research: Artificial intelligence technologies for genomic analysis and clinical decision support

Article Title: HKU scholar named sole recipient of 2026 APEC ASPIRE Prize

Article References: HKU scholar named sole recipient of 2026 APEC ASPIRE Prize. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: APEC ASPIRE Prize, Ruibang Luo, genomics, artificial intelligence, bioinformatics, precision medicine, clinical decision support, open-source software, The University of Hong Kong, newborn screening, cancer detection, data science

The study, published in the journal Advances in Therapy, drew on the Komodo Health Research Database, one of the largest longitudinal healthcare claims datasets in the country, covering approximately 330 million individuals across commercial insurance, Medicare, and Medicaid programs. Researchers led by Walter Toro Jimenez of Novartis Pharmaceuticals Corporation, together with colleagues at Analysis Group, examined claims recorded between January 2016 and October 2024 to reconstruct how patients with spinal muscular atrophy actually move through the therapeutic landscape. Because the data capture routine clinical care rather than the carefully controlled conditions of a clinical trial, the findings offer a rare window into how the promise of these therapies is being translated into everyday practice.

The biology underlying the disease explains why treatment has evolved so rapidly. Spinal muscular atrophy results from homozygous deletions or mutations in the SMN1 gene, which deprives the body of survival motor neuron protein and triggers the progressive degeneration of motor neurons in the spinal cord. A second gene, SMN2, produces small amounts of the same protein, and the severity of disease generally tracks with how much functional protein the SMN2 copies can generate. All three approved therapies work by increasing the output of that backup gene, but they do so through strikingly different mechanisms. Nusinersen, approved in 2016, is an antisense oligonucleotide injected into the spinal canal that corrects the splicing of SMN2 transcripts. Risdiplam, approved in 2020, is an oral small molecule that achieves the same splicing modulation through daily dosing. Onasemnogene abeparvovec, approved in 2019 as an intravenous infusion for children under two, delivers a functional copy of the SMN1 gene directly into cells using a viral vector, effectively replacing the missing genetic instructions in a single administration.

The clearest signal in the new analysis comes from the youngest patients. Among 257 children who initiated their first disease-modifying therapy at age two or younger during 2022 through 2024, a period after newborn screening had become widely available across most of the United States, 64.6 percent began with onasemnogene abeparvovec, 24.9 percent with risdiplam, and only 10.5 percent with nusinersen. The contrast with the immediate post-approval era is stark. An earlier study using a different claims database covering January 2017 through March 2021 found that just 16.2 percent of children under two received the gene therapy, while 83.8 percent received nusinersen. The researchers attribute the reversal largely to newborn screening, which was introduced in the US in 2018 and now identifies affected infants before symptoms appear, allowing treatment to begin at the stage when clinical trials have shown the greatest benefit.

Equally revealing is what happens after the first treatment. Children who started with the gene therapy were the least likely to add a second therapy type: only 12.0 percent of them did so, compared with 20.3 percent of those who began with risdiplam and 55.6 percent of those who began with nusinersen. Among the subset of 209 children with the most severe, type 1 form of the disease, the pattern held, with 8.3 percent of gene therapy recipients adding a second therapy versus 60.9 percent of nusinersen recipients and 18.5 percent of risdiplam recipients. The implication is that a single gene replacement appears to satisfy treatment needs for most infants, whereas patients on the chronic, repeat-dosing therapies more often end up adding or switching to another modality, most frequently the gene therapy itself.

To understand sequencing across the full age spectrum, the researchers assembled a second cohort of 341 patients with types 1 through 3 of the disease who received at least two different therapy types between May 2019, when at least two options became available, and October 2024. This group was markedly older, with a mean age of 15.4 years at the start of their first recorded therapy, and reflected an era when nusinersen was the only option: 75.4 percent of them began with the antisense oligonucleotide, 15.2 percent with risdiplam, and 9.4 percent with the gene therapy. Within this cohort, risdiplam was the most common second treatment, received by 76.2 percent of patients, followed by nusinersen at 13.2 percent and the gene therapy at 10.6 percent. Only five patients, or 1.5 percent, progressed to a third therapy type. The median interval between the first and second treatments was 18.9 months, though the timing varied by starting therapy, from 21.7 months for nusinersen starters to 13.2 months for those who began with risdiplam.

The picture changes again when the analysis is restricted to the 44 children who received multiple therapies and began treatment at age two or younger during 2022 through 2024, the true newborn screening generation. In this group, first treatments were distributed far more evenly, with 36.4 percent starting on the gene therapy, 34.1 percent on nusinersen, and 29.5 percent on risdiplam. When these children did move to a second therapy, the gene therapy dominated: 52.3 percent received onasemnogene abeparvovec as their second treatment, 45.5 percent received risdiplam, and just 2.3 percent received nusinersen. All of the children who had started on risdiplam, and most of those who had started on nusinersen, went on to receive the gene therapy. The median time from first to second treatment in this young cohort was 13.4 months, and the researchers suggest that some patients may have received the oral or injectable therapies as a bridge while awaiting gene therapy, while others switched because of the demonstrated value of gene replacement in previously treated patients.

The data also expose the growing complexity of managing a disease that now has multiple viable treatment pathways. A 2023 community survey cited in the study reported that 47.4 percent of children with spinal muscular atrophy had received more than one therapy type, with families and physicians citing reasons that included wanting a different treatment, physician recommendation, the desire to pursue every available option, and loss of motor function. The claims analysis adds a further wrinkle: among young children treated with multiple therapies, 20.0 percent of those who started on nusinersen and 38.5 percent of those who started on risdiplam later received the same therapy again after initiating a second one. Whether this reflects concurrent use, re-initiation after a gap, or something else cannot be determined from claims data alone, but the authors note that it points to increasing treatment complexity and burden for patients and caregivers alike.

The study has important caveats, which the authors acknowledge candidly. Claims databases lack the clinical measures, such as motor function scores, disease severity, and genetic modifiers like SMN2 copy number, that would allow researchers to connect treatment sequencing to outcomes. The specific reasons behind any switch, whether efficacy concerns, adverse effects, or issues of access and insurance coverage, are invisible in billing records. Disease type itself had to be inferred from an established algorithm based on age at diagnosis and patterns of clinical activity, raising the possibility of misclassification, particularly among pre-symptomatic infants identified by screening in whom distinguishing clinical features have not yet emerged. Birth-year service dates are masked for privacy, which forced the exclusion of the small fraction of patients, 5.2 percent, who received multiple therapy types in their first year of life, and small subgroup sizes limit the precision of some estimates.

Even with those limitations, the analysis stands as the first large-scale, real-world description of how patients with spinal muscular atrophy navigate an era of multiple disease-modifying options, and its message is unambiguous: newborn screening has reshaped the field, gene replacement therapy now anchors early treatment, and sequential multi-therapy strategies have become an ordinary feature of care rather than an exception. The landscape is still shifting. In 2025, regulators approved an intrathecal formulation of onasemnogene abeparvovec for patients aged two and older, meaning that for the first time all three therapeutic modalities are available across the full age range. Future studies, the authors write, will need to examine treatment sequencing across that broader population, control for the confounding factors that claims data cannot capture, and, most critically, link sequencing strategies to long-term clinical outcomes as the therapeutic story of this once untreatable disease continues to unfold.

Subject of Research: Real-world treatment sequencing patterns among patients with spinal muscular atrophy receiving multiple disease-modifying therapies

Article Title: Real-World Treatment Patterns in Patients with Spinal Muscular Atrophy Receiving Multiple Disease-Modifying Therapies

Article References: Toro Jimenez, W., Yang, M., Song, W., Ye, D., Lemus Wirtz, E., Meehan, J., & Reyna, S. (2026). Real-World Treatment Patterns in Patients with Spinal Muscular Atrophy Receiving Multiple Disease-Modifying Therapies. Advances in Therapy. https://doi.org/10.1007/s12325-026-03710-4

Image Credits: AI Generated

DOI: 10.1007/s12325-026-03710-4

Keywords: spinal muscular atrophy, onasemnogene abeparvovec, nusinersen, risdiplam, gene therapy, newborn screening, disease-modifying therapy, treatment patterns, SMN1, SMN2, motor neuron disease, real-world evidence

The underlying biology is straightforward but consequential. Bilirubin is the yellow pigment produced when the heme portion of hemoglobin is catabolized, and every molecule of bilirubin generated corresponds to one molecule of carbon monoxide released as heme oxygenase does its work. In newborns, whose livers are still maturing and whose red cell lifespan is shorter than in older children and adults, any acceleration of red cell destruction can push bilirubin concentrations upward rapidly. When bilirubin crosses the blood-brain barrier in sufficient quantities, it can cause acute bilirubin encephalopathy and permanent kernicterus, a devastating outcome that modern screening and phototherapy are designed to prevent. Identifying hemolysis early therefore gives clinicians a chance to intervene before bilirubin levels reach neurotoxic territory.

The study at the center of the dispute, conducted by Burch and colleagues, compared the direct antiglobulin test, commonly known as the DAT or Coombs test, against end-tidal carbon monoxide corrected for ambient carbon monoxide, abbreviated ETCOc, across three threshold levels. The DAT has long been the traditional tool for flagging immune-mediated hemolysis, particularly in infants whose mothers have antibodies that cross the placenta and coat the fetal red cells. It detects antibody bound to the infant’s red blood cells, which is an indirect and imperfect proxy for whether red cells are actually being destroyed at an accelerated rate. Many infants with clinically significant hemolysis test negative, and many DAT-positive infants never develop problematic bilirubin elevations.

The performance numbers cited in the correspondence are striking. Schutzman endorses the Burch group’s finding that the DAT had a sensitivity of only 13 percent for identifying hemolysis as a neurotoxicity risk factor, meaning the vast majority of hemolyzing infants went undetected by the blood test. ETCOc, by contrast, achieved a sensitivity of 88 percent at a correction threshold of 2.1 parts per million. The logic of the breath test follows directly from heme catabolism: because carbon monoxide is exhaled in proportion to bilirubin production, measuring it in the terminal portion of the infant’s exhaled breath, after subtracting background carbon monoxide in the ambient air, provides a direct physiological readout of the rate of red cell destruction. The approach has been refined over decades, and nomograms establishing reference ranges for term and late-preterm infants, including work in Chinese newborn populations, have helped define what constitutes abnormal heme catabolism in the first days of life.

Where Schutzman parts company with the original study is in its conclusion that ETCOc screening would lead to overdiagnosis, overtreatment, and higher costs. He argues that this conclusion rests on prevalence-dependent quantities, specifically the rates of indicated testing and phototherapy observed in the study cohort, and that those rates cannot be generalized to routine screening. Although all nursery admissions were nominally eligible for the Burch study, the authors themselves acknowledged that their cohort was enriched with DAT-positive and hyperbilirubinemic infants. In other words, the sample was not representative of an ordinary well-baby nursery population, and any cost or treatment projections drawn from it inherit that bias.

The magnitude of the bias, as Schutzman lays it out, is considerable. The Burch cohort showed a phototherapy rate of 9 percent, which he notes is more than fourfold the 2.1 percent rate reported in the same eight-hospital network when an unselected population of 22,455 infants was examined in earlier work by Sarathy and colleagues. That earlier study, published in Pediatrics, examined bilirubin measurement and phototherapy use after the American Academy of Pediatrics issued its 2022 newborn hyperbilirubinemia guideline, providing a large benchmark for what phototherapy utilization actually looks like across a broad, unselected nursery population. A selective sample, Schutzman contends, simply cannot support conclusions about what would happen if ETCOc were adopted for routine or universal screening.

To make the counterpoint concrete, the correspondence points to the only study to date that evaluated a large, continuous cohort of all admissions to a well-baby nursery. That investigation, by Wells and colleagues and published in the Journal of Perinatology in 2025, examined end-tidal carbon monoxide monitoring for significant hemolysis in the routine management of newborn hyperbilirubinemia. Its findings ran in the opposite direction from the overdiagnosis concern: when ETCOc was used across an unselected population, it was associated with reductions in serum bilirubin sampling, in phototherapy use, and in readmissions. The mechanistic explanation is plausible. If a breath test reliably identifies which infants are truly hemolyzing, clinicians can reserve blood draws and treatment for those who need them, while confidently observing infants whose bilirubin production is normal, rather than reflexively testing and treating based on an insensitive proxy.

The correspondence also situates the debate within a broader effort to establish a definitive reference standard for neonatal hemolysis. A 2023 review by Christensen, Stevenson, Bhutani, and coauthors asked whether any test could serve as a gold standard for diagnosing and quantifying hemolysis in neonates and infants, highlighting the inherent difficulty of the problem. Hemolysis is a dynamic process, its rate changing hour by hour in the first days of life, and no single static measurement captures it perfectly. Earlier collaborative work by Bhutani, Maisels, Schutzman, and colleagues published in Acta Paediatrica in 2018 addressed the identification of risk for neonatal hemolysis directly, helping to frame ETCOc as a physiologically grounded candidate for that role. The current exchange is thus not an isolated squabble over a single study but part of a decade-long effort to replace indirect markers with a direct measure of heme catabolism.

What hangs on this methodological dispute is far from academic. Severe neonatal hyperbilirubinemia remains one of the most common reasons for newborn readmission, and kernicterus, while rare in well-resourced settings, is entirely preventable when at-risk infants are recognized in time. A screening strategy that misses nearly nine out of ten hemolyzing infants, as a 13 percent sensitivity implies, leaves a substantial reservoir of risk undetected, particularly among infants whose hemolysis is not immune-mediated and would never be flagged by a DAT. Conversely, an approach that indiscriminately expands testing and treatment burdens families and health systems. The correspondence argues that the evidence points toward the breath test as the more faithful instrument, and that fears of overtreatment rest on a biased sample rather than on data from the populations in which routine screening would actually be deployed.

For now, the field awaits the kind of large, unselected, prospective studies that both sides agree would settle the question. The correspondence itself declares no funding, and its author discloses only an unpaid advisory role with a company in the field, Capnia, Inc., alongside affiliations with Jefferson Einstein Philadelphia Hospital, Thomas Jefferson University Hospital, and Nemours Children’s Hospital Delaware. Whether ETCOc moves from research tool to standard nursery practice will depend on whether the reductions in blood sampling, phototherapy, and readmissions seen in the Wells cohort can be replicated at scale. But the core physiological insight at the heart of the debate, that every bilirubin molecule announces itself in a puff of exhaled carbon monoxide, continues to make the breath test a compelling candidate for catching the hemolyzing newborn before jaundice becomes dangerous.

Subject of Research: Comparison of the direct antiglobulin test and end-tidal carbon monoxide measurement for detecting hemolytic hyperbilirubinemia in newborns

Article Title: Hemolytic hyperbilirubinemia

Article References: Schutzman, D. L. (2026). Hemolytic hyperbilirubinemia. Journal of Perinatology. https://doi.org/10.1038/s41372-026-02919-z

Image Credits: AI Generated

DOI: 10.1038/s41372-026-02919-z

Keywords: neonatal jaundice, hyperbilirubinemia, hemolysis, end-tidal carbon monoxide, direct antiglobulin test, phototherapy, newborn screening, bilirubin, kernicterus, neonatology, Journal of Perinatology, diagnostic sensitivity

The therapeutic landscape has changed dramatically over the past several years. Treatment options now span substrate reduction and replacement, cofactor replacement, enzyme replacement, RNA-based therapies, cell therapy, gene therapy, and gene editing. The general trend, Whitehead notes, is a movement away from ongoing metabolic replacement strategies such as dietary modification, cofactor supplementation, and enzyme replacement therapy toward single-instance genetic interventions like gene therapy and, eventually, gene editing. To keep pace with this rapidly evolving field, clinicians can consult a dynamic, continuously updated resource known as the Metabolic Treatabolome, which catalogs treatable inherited metabolic disorders. For the practicing radiologist, however, the essential message is simpler: certain reproducible patterns on MRI and MR spectroscopy should trigger immediate consideration of a treatable neurogenetic condition, because these diseases arise from selective tissue vulnerability that produces identifiable and characteristic imaging signatures.

The first challenge is knowing when to suspect a metabolic disorder at all. The imaging characteristics of neurogenetic brain diseases often overlap with far more common entities, particularly hypoxic-ischemic injury and encephalitis. Profound hypoxic-ischemic injury in neonates typically affects the posterior putamen, the posterior limb of the internal capsule, and the ventrolateral thalamus simultaneously, because these regions have high energy demands. When patterns diverge from this expectation, metabolic disease should rise on the differential. Specific red flags include deep gray nuclear lesions that spare the thalami or basal ganglia, temporally disparate lesions, brainstem-predominant abnormalities, unexplained reduction in brain volume, malformations, and chronic lesions in a newborn, none of which fit the typical timeline or distribution of oxygen deprivation or infection.

Among the white matter-dominant disorders, X-linked adrenoleukodystrophy stands out as the single most important diagnosis to know. Caused by variants in the ABCD1 gene, a peroxisomal defect, it is the most common leukodystrophy, it is treatable, and it shows a moderately specific imaging pattern: in roughly 80 percent of cases, demyelination begins in the splenium of the corpus callosum, spreads to the forceps major, descends along projection fibers, and ultimately involves auditory and visual white matter tracts. Active inflammation is marked by reduced diffusivity and contrast enhancement. Alternative phenotypes include frontal onset, deep diffuse, and projectional patterns. With the addition of ABCD1 to the recommended uniform screening panel in the United States, 47 states and the District of Columbia now perform newborn screening, and MRI surveillance identifies the fewer than half of affected boys who develop the cerebral form. Treatment, most effective early when few brain lesions exist, is allogeneic hematopoietic stem cell transplantation, with autologous gene-modified stem cell transplant as a second-line option. Patients with typical lesions and a Loes score below 10 may qualify for therapy, and post-treatment stability is defined as a Loes score below 10 or an increase of less than 6 points from baseline.

Two other leukodystrophies, Krabbe disease and metachromatic leukodystrophy, illustrate how clinical context and subtle imaging distinctions guide diagnosis. Juvenile-onset Krabbe disease can closely mimic X-linked adrenoleukodystrophy with poorly marginated splenial and forceps major abnormality, while the infantile form looks different, featuring lesions in the dentate nuclear hilus, deep cerebral hyperattenuation on CT from galactosylceramide deposition, thalamic T2 shortening, optic chiasm thickening, and enhancement of cranial nerves and the cauda equina. A striated, tigroid pattern in the corona radiata from perivascular myelin sparing is typical but not specific. Hematopoietic stem cell transplant works best before symptoms appear, ideally in the first month of life, which makes newborn screening for GALC enzyme activity, now implemented in 18 states, critically important. Metachromatic leukodystrophy shares several features with infantile Krabbe disease but usually spares the chiasm and hindbrain, may involve the genu earlier, and lacks the deep cerebral hyperattenuation. Markedly elevated myoinositol on spectroscopy and gallbladder disease, even in pre-symptomatic patients, provide additional clues. The preferred treatment is gene therapy using lentivirally modified ARSA, available for children under seven who are pre-symptomatic or only mildly symptomatic, and six states have added MLD to newborn screening panels since its endorsement.

Maple syrup urine disease demonstrates how imaging can track treatment response in real time. This aminoacidopathy typically presents at the end of the first neonatal week with encephalopathy and striking white matter edema, with symmetric, markedly reduced diffusivity and swelling in myelinating and myelinated tracts of the cerebrum, cerebellum, and brainstem. Spectroscopy reveals elevated branched-chain amino acids and ketoacids as a broad doublet near 0.9 to 1 ppm, though detecting these peaks reliably at 3 tesla requires careful echo time selection due to anomalous J-coupling. While protein restriction is necessary and leucine, isoleucine, and valine cannot be eliminated entirely from the diet, liver transplantation is considered a definitive treatment that permits a more relaxed diet. A favorable post-transplant scan shows resolution of the diffusion abnormalities and intramyelinic edema, resolution of lactate, normalization of NAA, and reduction of the branched-chain amino acid peaks, findings documented in a patient imaged seven years after curative transplantation. The mucopolysaccharidoses, a group of multi-system lysosomal storage disorders, offer another treatable example: enlarged perivascular spaces combined with skeletal changes such as a J-shaped sella, frontal bossing, and odontoid hypoplasia suggest the diagnosis, and weekly enzyme replacement therapy has been shown to resolve white matter signal abnormality and stabilize neurological symptoms in eligible patients.

Menkes disease, an X-linked copper transport disorder caused by ATP7A variants, produces one of the most visually distinctive findings in pediatric neuroradiology: diffuse marked elongation and tortuosity of the intracranial arteries, present almost always, even in neonates, because copper is essential for vessel wall integrity. Supporting findings include delayed myelination, transient tumefactive temporal lobe vasogenic edema between one and ten months of life, basal ganglia lesions with reduced diffusion, and skeletal changes such as Wormian bones and fractures. Early daily subcutaneous copper histidinate injections, which bypass the defective copper transporter, are essential for preventing severe neurodegeneration, reducing seizures, and prolonging life. On the gray matter side, Wilson disease from ATP7B variants shows basal ganglia, thalamic, midbrain, and pontine lesions with a characteristic mixed increased-and-decreased signal, where T2 signal correlates with neurological dysfunction and T1 signal with hepatic dysfunction, and the famous face of the giant panda in the midbrain may appear. Copper reduction through diet, zinc supplementation, and chelation, or curative liver transplantation, addresses the underlying metabolic derangement.

Several other gray matter disorders reward pattern recognition. Biotin-thiamine-responsive basal ganglia disease, from SLC19A3 variants, causes thalamic and cortical edema with prominent putaminal involvement, and responds, as its name promises, to lifelong biotin and thiamine supplementation. Glutaric aciduria type 1 announces itself structurally before the basal ganglia lesions appear, with underopercularization and enlarged temporopolar subarachnoid spaces, and early l-carnitine with an arginine-restricted, lysine-free diet is highly effective at preventing permanent striatal injury. Urea cycle defects produce hyperammonemic brain injury with peri-insular and deep cortical lesions that spare the thalami, a pattern unusual in hypoxic-ischemic injury, alongside elevated glutamine and glutamate with reduced myoinositol on spectroscopy; dietary management and, in neonatal-onset ornithine transcarbamylase deficiency, curative liver transplantation are the mainstays. Molybdenum cofactor deficiency type A responds to substrate replacement with cyclic pyranopterin monophosphate, and neuronal ceroid lipofuscinosis CLN2, the most common childhood neurodegenerative disease group, can be slowed by bi-weekly intraventricular cerliponase alfa, identifiable on imaging by cerebellar-predominant atrophy, thalamic T2 shortening, and subinsular signal change.

Finally, MR spectroscopy deserves special emphasis because it can be diagnostic even when the MRI appears entirely normal. Fatty acid metabolism disorders such as CPT1 deficiency produce elevated lipid peaks at 0.9 and 1.3 ppm on a normal-appearing scan, and respond to dietary modification with frequent high-carbohydrate meals and medium-chain fatty acid supplementation. Galactosemia yields a large doublet at 3.7 ppm from toxic galactitol accumulation, which inverts at intermediate echo times and normalizes with dietary change. Cerebral creatine deficiency syndromes show markedly reduced creatine peaks at 3 and 3.95 ppm, and creatine supplementation can normalize the spectrum in AGAT and GAMT deficiency. The review recommends a low threshold for adding short-echo single-voxel spectroscopy to any suspected neurometabolic case, with the voxel placed in the deep gray nuclei for poliodystrophies and parietal white matter for suspected leukodystrophies. The overarching conclusion is unambiguous: symmetric gray or white matter lesions in patterns atypical for hypoxic-ischemic injury and infection, or a characteristic spectroscopic signature, should prompt consideration of these not-to-be-missed disorders, because failure to recognize them promptly delays lifesaving and morbidity-reducing treatment.

Subject of Research: Neuroimaging patterns of treatable neurogenetic and inherited metabolic brain diseases in children

Article Title: Don’t miss this treatable neurogenetic disease!

Article References: Whitehead, M. T. (2026). Don’t miss this treatable neurogenetic disease!. Pediatric Radiology. https://doi.org/10.1007/s00247-026-06796-8

Image Credits: AI Generated

DOI: 10.1007/s00247-026-06796-8

Keywords: neurogenetic diseases, leukodystrophy, MRI, MR spectroscopy, X-linked adrenoleukodystrophy, Krabbe disease, metachromatic leukodystrophy, maple syrup urine disease, Wilson disease, newborn screening, gene therapy, pediatric neuroradiology

Prader-Willi syndrome arises from errors in the expression of paternally inherited genes on chromosome 15, affecting roughly one in ten thousand to one in thirty thousand births. Infants with the condition typically show weak muscle tone, a poor suck that makes feeding difficult, and mild facial differences such as a narrow forehead, almond-shaped eyes, and a downturned mouth. As childhood progresses, the picture changes dramatically: an unrelenting drive to eat emerges, leading to life-threatening obesity if not managed, alongside cognitive and behavioral difficulties. Early recognition matters enormously, because growth hormone therapy, feeding support, and structured developmental care initiated soon after birth can measurably improve outcomes. The catch is that the neonatal facial phenotype, the very signal clinicians might rely on, is faint and easily overlooked in a squirming, sleeping infant.

The Fudan team, led by Yunqian Zhu and Wenhao Zhou, set out to build a diagnostic aid that works precisely where existing tools falter. Their strategy rested on a pragmatic observation: photographs of older children with the syndrome are relatively abundant, while images of newborns are scarce and expensive to collect. Could a model trained on the faces of older children still detect the syndrome in newborns whose facial features are far less pronounced? To find out, the researchers assembled a training cohort of 113 children drawn from a public platform, of whom 21 carried confirmed diagnoses of Prader-Willi syndrome, and a separate validation cohort of newborns at their own hospital consisting of ten affected infants and thirteen controls.

The technical approach combined multiple feature sets and learning algorithms. The team extracted a detailed panel of facial measurements, including distances, angles, and shape descriptors capturing everything from eye spacing to lip contour, alongside age. They then trained six different classification algorithms, spanning from logistic regression to Random Forests and AdaBoost ensembles, on three versions of the feature set: the complete panel, a top-ten subset of the most informative features, and a clinically selected subset chosen for relevance to the syndrome. Performance was judged using the area under the precision-recall curve, a metric well suited to datasets where positive cases, like rare diseases, are outnumbered by controls. Across all feature sets, the Random Forest and AdaBoost models emerged as the strongest performers.

The headline result came from the Random Forest trained on the complete feature set. In cross-validated testing on the training cohort, it achieved a mean sensitivity of 0.97, with a 95 percent confidence interval stretching from 0.84 to 1.00, and a specificity of 0.95, correctly identifying nearly every child with the syndrome while rejecting nearly all children without it. The model was also well calibrated, meaning its predicted probabilities tracked actual likelihoods rather than being systematically overconfident or timid, a property the researchers emphasize as essential for any tool intended to support real clinical decisions rather than inflate benchmark scores.

Peering inside the model revealed a face in numbers. Age, face width, and features of the upper lip carried the greatest predictive weight, consistent with the well-described dysmorphic profile of the syndrome. More intriguingly, the analysis surfaced four facial features not previously reported in the context of Prader-Willi syndrome: jaw symmetry, chin shape, nose width, and jaw width. A sensitivity analysis showed that 26 facial features contributed more to classification than age alone, suggesting that the model was genuinely reading facial structure rather than simply learning that children with the condition tend to be photographed at certain ages. These newly identified dimensions of the syndrome’s facial signature could inform clinicians and researchers alike, expanding what is understood about how a single genetic lesion reshapes craniofacial development.

The decisive test, however, was neonatal. When the researchers applied their child-trained model to the newborn validation cohort, they first had to adjust the decision threshold, because the probability scores that fit older children needed recalibration for the subtler neonatal face. After this re-estimation, the Random Forest model detected 80 percent of affected newborns while correctly clearing 77 percent of controls. To benchmark this performance, the team compared it against Face2Gene, a widely used commercial facial analysis app for genetic syndromes. On the same newborn images, Face2Gene achieved only 70 percent sensitivity and a mere 15 percent specificity, meaning it flagged almost every baby as suspect. At a false positive rate of 0.25 or less, the Fudan model delivered comparable sensitivity but dramatically higher specificity, a distinction that matters in a screening context where alarmingly frequent false alarms would erode trust and swamp follow-up genetic testing capacity.

Beyond the immediate clinical promise, the study carries a broader methodological message for the field of computational phenotyping. Artificial intelligence tools for rare disease detection, including deep learning systems such as DeepGestalt and GestaltMatcher, have demonstrated impressive diagnostic power in clinics and genetic consultation settings, but their training data overwhelmingly feature older children and adults. Newborns, with their rounded faces, abundant subcutaneous fat, and physiologically underdeveloped craniofacial structures, represent a distribution shift that most models handle poorly. The Fudan results suggest a practical workaround: rather than waiting years to accrue a newborn dataset of sufficient size, researchers can train on pediatric data and transfer the learned phenotype knowledge to neonates through careful threshold adjustment. That could accelerate screening tools for a whole family of genetic conditions where the newborn window of intervention is narrow and the data pipeline begins nearly empty.

Caveats remain, and the authors are candid about them. The validation cohort numbered just 23 newborns, a pilot sample sufficient to demonstrate feasibility but far short of what a population-level screening program would demand. The control group’s size and composition limit how confidently the specificity estimates can be generalized, and the model must eventually prove itself across diverse ethnic populations, camera conditions, and gestational ages. The researchers also note that facial analysis is intended as a triage tool, not a replacement for the methylation testing that definitively diagnoses the syndrome. A positive screen would simply channel scarce genetic testing resources toward the babies most likely to benefit, shortening what can otherwise be a diagnostic odyssey measured in months or years.

Still, the trajectory of the work is hard to ignore. With a new drug recently approved to treat hyperphagia in older patients and advances such as hypothalamic organoids offering laboratory models of the underlying biology, the therapeutic landscape for Prader-Willi syndrome is shifting, and the value of catching the condition at birth rises in step. A camera and an algorithm cannot replace a pediatric geneticist, but they can stand watch over every delivery ward, quietly and cheaply, flagging the one baby in ten thousand whose face holds a message that human eyes, on their own, too often miss. If larger prospective studies replicate the Shanghai results, facial phenotype screening could become the first step of a standard newborn pathway, ensuring that children with Prader-Willi syndrome receive growth hormone, feeding support, and early intervention before the syndrome’s most damaging consequences take root.

Subject of Research: Machine learning-based facial phenotype screening for early detection of Prader-Willi syndrome in newborns

Article Title: Early identification of Prader-Willi syndrome in newborns using facial phenotype-based machine learning

Article References: Zhu, Y., Chen, H., Chen, J., Yang, L., Dong, X., Ni, Q., & Zhou, W. (2026). Early identification of Prader-Willi syndrome in newborns using facial phenotype-based machine learning. Pediatric Research. https://doi.org/10.1038/s41390-026-05491-6

Image Credits: AI Generated

DOI: 10.1038/s41390-026-05491-6

Keywords: Prader-Willi syndrome, machine learning, facial phenotype, newborn screening, rare disease, random forest, Face2Gene, genetic diagnosis, dysmorphism, pediatric research, artificial intelligence, early diagnosis