Get an inside look at how Polin and Fox Fetal and Neonatal Physiology approaches genetic variants and neonatal disease — from exome and genome sequencing in the NICU to the genomic risk factors behind common prematurity-related morbidities.
The most complete text available in this complex and dynamic field Order your copy today at elsevierhealth.com/9780443128233 From Polin and Fox Fetal and Neonatal Physiology, 7th Edition Exclusive preview: Genetic Variants and Neonatal Disease CHAPTER 168 — GENETIC VARIANTS AND NEONATAL DISEASE 2159 (NHGRI)-funded Clinical Genome Resource (ClinGen). 79 ClinGen’s mission is to define the clinical relevance of genes and variants to be used in research and precision medicine (Fig. 168.1). ClinGen has developed tools to evaluate clini- cal validity of gene-disease associations and pathogenicity of genetic variants for use in clinical care, as well as educational material for clinicians and patients to better understand the process of curation of gene- and variant-disease associations. Curation is a term used to describe the analysis of the exist- ing data and evidence about a specific gene or gene vari- ant and its relation to a specific phenotype. The tools used allow for a quantitative approach to weighing the evidence supporting gene-disease and variant-disease associations to reach summary categories: pathogenic, likely pathogenic, uncertain significance, likely benign, and benign. ClinGen’s work and impact are made directly available on their web- site (https://clinicalgenome.org/) and additionally, via another resource, ClinVar. ClinVar is the NCBI archival database that aggregates information about genomic varia- tion and relationships to human health that are provided by researchers, clinical laboratories conducting sequenc- ing, expert groups, clinics, and patient registries. In addi- tion, with genomeconnect.org (an online registry designed by ClinGen), patients may also submit their clinical genetic test reports and health information to increase research- ers’ and healthcare providers’ overall understanding of the relationship between genetics and health. While single investigators may submit a variant suspected of association with a specific disease, ClinVar submissions are scored based on the number and types of sources submitting the same data, and validation by working groups and expert panels that examine the genetic epidemiologic evidence, as well as supportive evidence from experimental model systems. These curations, as well as user interfaces, are constantly evolving, as data from multiple sources become available for consideration by expert panels. 81,82 Currently, there are 28 working groups, 54 gene curation panels, and 69 vari- ant curation panels working within the broader ClinGen effort (https://clinicalgenome.org/). The ClinGen website includes links to multiple educational modules and publicly available browser tools (Fig. 168.2). Clinicians should be aware that different automated cura- tion tools in use in academic and commercial laboratories may make different “calls” of significance of associations between specific variants and disease. Clinicians should also be aware that data continue to accumulate, and the number of expert panels that have been trained using ClinGen pro- grams is also growing, with the aim to decrease the hetero- geneity of interpretation about the significance of gene/ variant/disease associations. As the legitimacy of the data- bases increases with replication, clinicians must continue to be vigilant and cautious, as increasing knowledge may some- times lead to a change in the classification of a variant. As a cautionary example, a recent report analyzing variants that Table 168.1 Web Resources Providing Genetic and Genomic Information and Training. ClinGen: a National Institutes of Health (NIH)-funded resource dedicated to building a central resource that defines the clinical rel- evance of genes and variants for use in precision medicine and research: https://clinicalgenome.org/ ClinVar: freely accessible, public archive of reports of the relationships among human variations and phenotypes, with supporting evi- dence: https://www.ncbi.nlm.nih.gov/clinvar/ Database of Genotype and Phenotype (dbGaP): an archive of data from genome-wide association studies on a variety of diseases and conditions accessible through the NCBI: https://www.ncbi.nlm.nih.gov/gap/ DECIPHER: a database of reported copy number variants and linked phenotypes: https://decipher.sanger.ac.uk/ Ensembl: a genome browser for vertebrate genomes that supports research in comparative genomics, evolution, sequence variation, and transcriptional regulation. Ensembl annotates genes, computes multiple alignments, predicts regulatory function, and collects disease data: https://www.ensembl.org/index.html GeneMatcher: a website that enables connections between clinicians who have a patient with a candidate or ultra-rare gene and researchers who have an interest in that gene: https://genematcher.org/ GeneReviews: a clinical resource for many genetic conditions that provides clinically actionable information, including diagno- sis, inheritance, and management, as well as a differential diagnosis of related conditions: https://www.ncbi.nlm.nih.gov/books/ NBK1116/ GenomeConnect: an online registry designed by the Clinical Genome Resource (ClinGen) for people who are interested in sharing de-identified genetic and health information to improve understanding of genetics and health: https://www.genomeconnect. org/ Online Mendelian Inheritance in Man (OMIM): a searchable database of clinical features, phenotypes, and genes: https://omim.org/ Unique: a website with patient-/family-facing resources regarding chromosome and gene disorders: https://www.rarechromo.org/ University of California Santa Cruz Genome Browser: a website created initially to ensure public access to the initial human genome assembly; has now evolved to include a broad collection of vertebrate and model organism assemblies and annotations, along with a large suite of tools for viewing, analyzing, and downloading data: https://genome.ucsc.edu/ Leiden Open Variation Database (LOVD): An online gene-centered collection and display of DNA variants; its purpose is to provide a flexible, freely available tool for genomic variant and phenotype collection, display, and curation . LOVD allows both patient-centered and gene-centered views. LOVD is open source, released under the GPL license, and is actively being improved. On the server in Leiden, LOVD offers free hosting and support of LOVD-powered gene variant databases. https://www.lovd.nl/ 2160 SECTION 25 — PATHOPHYSIOLOGY OF NEONATAL DISEASES had been classified as pathogenic or likely pathogenic, and also occurred in more than 0.5% of the population, found that of 217 variants in 173 genes that were selected for cura- tion, 87 (40%) of the variants were downgraded to benign, likely benign, or variant of uncertain significance. 84 Obtaining informed consent prior to testing is critical for the purpose of educating patients and families about possible testing outcomes. 27,85,86 Some outcomes can cause unexpected difficulties, including incidental but clini- cally relevant findings not related to the phenotype that prompted the test but linked with adult-onset diseases; VUS, which can lead to ambiguity and possible frustration; and candidate gene discovery that may require many years of research-based functional analysis before the gene’s rele- vance to human disease is conclusively determined. The case for the adoption of widespread GS will arise from its successful utilization across a variety of clinical indications and patient cohorts. As researchers who have successfully built the infrastructure to support effective application of this technology continue to expand their networks, patients, families, and caregivers may begin to see the benefit of this testing through earlier diagnosis and potentially a change in clinical management. These tests will also necessitate con- tinued monitoring of variants either for evolving evidence of clinical significance, in the case of VUS, or the confirma- tion of persistent classification among variants determined to cause disease. 87 Currently, most labs perform ES/GS using Illumina’s state-of-the-art “short-read” sequencing-by-synthesis. 88 Short- read ES/GS has enabled the identification of a plethora of pathogenic variants in Mendelian genes. Additionally, it has fostered the discovery of novel disease-associated genes, thus finally providing an accurate diagnosis to patients and families that have spent years on a diagnostic odyssey. 89 Despite advances in clinical genetic testing and the utiliza- tion of the most comprehensive genetic testing clinically available, an estimate of approximately 50% of patients with a suspected genetic disorder remain undiagnosed. One of the factors contributing to this modest diagnostic rate is the use of short-read sequencing, which fails to pro- vide coverage in clinically challenging genomic regions such as repeat expansion regions. Emerging technologies, such as long-read DNA sequencing, are able to overcome this limitation since they use long DNA fragments ranging in size from 1000 to 20,000 bases. Voted the method of the year in 2022, long-read sequencing (LRS), also described as “third-generation sequencing,” aims to interrogate whole human genomes and uncover novel variants inaccessible with short-read sequencing. 90 When compared to short- read genome sequencing (SRS), LRS has many theoretical advantages. Along with increased read length, LRS has the ability to sequence native DNA through repetitive genomic regions such as long homopolymer stretches, pseudogenes, and short tandem repeat regions. Additionally, it is capable of performing haplotype phasing (variants are assigned to the homologous paternal or maternal chromosomes), epi- genetic detection (with the power to determine the methyla- tion status of CpG sites), and RNA sequencing. Lastly, LRS enables de novo genome assembly to allow for more effec- tive evaluation of structural variation. 91 Pacific Biosciences (PacBio) and Oxford Nanopore Technologies (ONT) are leaders in the LRS arena, although other companies, includ- ing Illumina, have also developed LRS methods. 91 ONT uses nanopore sequencing, where an ionic cur- rent can drive single-stranded DNA (originally dsDNA dis- sociated via an adaptable motor protein) or RNA through a nanopore membrane. Alterations in charge determine nucleotide type as the nucleic acid passes through the nano- pores. 92 Advances in nanopore sequencing technology have resulted in enhanced sequencing speeds reaching ~450 base pairs/second, increased read lengths of up to 2.3 MB, and accurate resolution of complex structural variants. 93,94 Raw sequencing and methylation data can be provided in real time following base calling using nanopore sequencing. This sequencing technology also has an adaptive sampling ability, allowing for the enrichment of specific regions of the genome. Additionally, nanopore sequencing allows for direct RNA sequencing, providing rapid and concise tran- scriptomic analysis. 91,95,96 PacBio HiFi LRS works by ligating hairpin adapters at both ends of the DNA fragment, creating a circular mole- cule (also known as SMRTbell) that is repeatedly sequenced. This technology is able to generate highly accurate reads of DNA molecules ( > 99.9%; QV > 30) with an attached cost of $1000 per genome. 97 Combined, both PacBio and ONT technologies have increased the number of pathogenic vari- ants identified and previously missed by clinical genetic testing. As the capabilities of nanopore sequencing have improved, so has its implementation in the clinic. Recent studies have demonstrated that nanopore sequencing, in conjunction with the appropriate analysis tools, can sequence a human genome in less than 2 hours, provide candidate variants for disease within 8 total hours, and provide a 42% diagnos- tic rate. 98 A striking example of this was the identification Improved patient care through genomic medicine Building a genomic knowledge base ClinVar and other resources ClinGen’s critical questions Sharing genetic and health data Patients Clinicians Laboratories Researchers Is this gene associated with a disease? Clinical validity Is this variant causative? Pathogenicity Is this information actionable? Clinical utility Fig. 168.1 Clinical genomic (ClinGen) resource framework. ClinGen’s overall mission is to build a genomic knowledge base to improve patient care. The ClinGen framework provides a semiquantitative mea- surement for the strength of evidence of a gene-disease relationship that correlates to a qualitative classification: “Definitive,” “Strong,” “Moderate,” “Limited,” “No Reported Evidence,” or “Conflicting Evidence.” Within the ClinGen structure, classifications derived with this framework are reviewed and confirmed or adjusted based on the clinical expertise of appropriate disease experts. Detailed guidance for utilizing this framework and access to the curation interface are available on the ClinGen website. 80 (From https://clinicalgenome.org/ start/#loc_1550536143-7476-1.) CHAPTER 168 — GENETIC VARIANTS AND NEONATAL DISEASE 2163 subjects with the disease phenotype are ideal, as have been used for multiple complex common diseases of adulthood, such as prostate cancer and Alzheimer disease. 129,130 GWAS studies in the NICU population have numbered in the hun- dreds and rarely thousands, and therefore are much smaller than the adult cohorts. 131–136 As a comparison, in a GWAS of 2727 cases and 3336 controls for Alzheimer disease, APOE gene variants were found to be associated with Alzheimer disease with a P value of 2.52 × 10 − 53 . 137 An additional challenge for genomic analysis among extremely preterm cohorts is the differences in minor allele frequency among study cohorts of differing ancestry. While studies on samples from relatively homogenous nations, such as Iceland, may produce fairly homogenous, replicable results, results from cohorts of mixed ancestry, which is more common in the United States, present challenges to analyz- ing associations between variants that differ in prevalence in cohorts of different ancestry. Acknowledging those strate- gies and challenges, studies in cohorts of premature infants for genetic associations using candidate gene and SNP anal- yses and more agnostic GWAS approaches have produced some findings of interest that reach genomic significance within the study cohort. These results, while encouraging, continue to be only tempting because of the severe limita- tion of examining relatively small cohorts compared to the more typical larger population analyses that can be done for common, complex diseases that emerge later in life. 138,139 RETINOPATHY OF PREMATURITY Recent reviews of genetic risk factors for ROP document the lack of any emergent, identifiable, strongly associ- ated genetic variants with severe ROP risk among preterm infants. Certainly, ROP-like phenotypes are linked with rare inherited variants or spontaneous mutations causing diseases like familial exudative vitreoretinopathy (FEVR) and Kabuki syndrome, but for the vast majority of prema- ture infants, candidate gene/SNP studies have not iden- tified variants with genome-wide significance. Multiple attempts have been made to identify associations that are plausibly associated with the pathophysiologic mecha- nisms of ROP, including VEGF and VEGF receptors. One report, which tested a panel of over 1000 variants in over 100 genes related to inflammation and organ develop- ment, identified two intronic variants in the brain-derived neurotrophic factor (BDNF) gene with severe ROP versus nonsevere or no ROP with a P value of less than 5 × 10 − 7 . 140 Lower serum levels of BDNF have been associated with higher likelihood of ROP in reports testing associations of serum levels of various cytokines and growth factors with outcomes in premature infants. This adds some plausibility to an association with BDNF variants that might influence expression, or an association with some other gene prod- uct/component of a pathway important for neurovascular development that includes BDNF. 141–143 While this is prom- ising, the impact of specific variants on BDNF circulating levels, or in situ levels in the developing retina, has not been described. More recently, analysis beyond genotyping has identified that placental CpG methylation, a measure of epigenetic modification, of 12 different genes is associated with the development of pre-threshold ROP. Interestingly, the genes with methylation changes associated with ROP included BDNF. 144 A recent GWAS study performed with patients enrolled in the iROP consortium revealed a novel ROP risk association with the SNP GLI3 rs2058019 reach- ing genome-wide significance ( P = 4.90E − 09). 145 This is the first study to identify an ROP-associated variant. Further studies are needed to validate these findings. NECROTIZING ENTEROCOLITIS In a cohort study using a GWAS approach, minor allele(s) in a cluster of SNPs spanning a 43-kb region of chromo- some 8 (8q23.3) conferred an odds ratio of 4.72 (95% con- fidence interval [CI]: 2.51–8.88) for elevated risk of NEC, with multiple SNPs associated with P < 10 − 8 (Fig. 168.4). Two smaller clusters on chromosomes 14 and 11 exhib- ited P values of 10 − 7 to 10 − 8 . Like many gene association studies done in the extremely preterm population, this analysis was limited by a small sample size ( n = 751, only 30 with surgical NEC), from multiple sites in the United States, with significant ancestry admixture. Interestingly, the increased risk was similar for all three genetic ances- tries represented in this population. 146 The investigators attempted to validate the associations of the SNPs in the chromosome 8 region with NEC in a separate cohort ( N = 1018, 26 with surgical NEC) of premature infants enrolled in a study using GWAS to identify variants associated with severe intraventricular hemorrhage. 134 In that cohort, one of the SNPs (rs13252246) was associated with NEC with SNP 3 Associated to disease SNP 3 SNP 2 No association to disease SNP 2 SNP 1 No association to disease Using a chip can genotype 500,000–5 million SNPs SNP 1 Individuals with disease Individuals without disease Fig. 168.3 Genome-wide association studies (GWAS). In GWAS, genotypes of single nucleotide polymorphisms, that is, loci with variants identified in 1% or more of the population, are identi- fied across the entire genome. Prevalence of the variants among individuals with a condition is compared with prevalence among individuals without the condition to identify candidate genes and variants associated with disease. SNP , Single nucleotide poly- morphism. (From www.genome.gov/about-genomics/fact-sheets/ Genome-Wide-Association-Studies-Fact-Sheet.) CHAPTER 168 — GENETIC VARIANTS AND NEONATAL DISEASE 2161 of a pathogenic heterozygous variant in a 3-month old who presented with status epilepticus and seizure semiol- ogy. Within 9 hours, a likely pathogenic variant in CSNK2B was detected, resulting in a definitive diagnosis of Poirier- Bienvenu neurodevelopmental syndrome. 99 Other studies have demonstrated the importance of nanopore sequencing in several genetic disorders, including Angelman syndrome, Duchenne muscular dystrophy, and retinitis pigmentosa. Nanopore sequencing allowed for the identification of a complex structural variant in DMD , confirming a genetic diagnosis for an individual, even after routine genetic test- ing, transcriptomic, and dystrophin protein analysis failed to yield a genetic diagnosis. 100 Additionally, the clinical sig- nificance of a UBE3A variant in an individual with Angelman syndrome was confirmed by identifying the parental allele of this variant using nanopore sequencing. 101 The application of nanopore sequencing also allowed for the identification of likely pathogenic structural variants in individuals with Gene-disease clinical validity curation The ClinGen Gene curation working group has developed a framework to standardize the approach to determine the clinical validity for a gene-disease pair. This framework: • Defines the criteria needed to assess clinical validity, • Describes the evidence supporting a gene-disease association in a semi-quantitative manner, and • Allows curators to use this information to methodically classify the validity of a given gene-disease pair. The ClinGen Gene-disease clinical validity curation process involves evaluating the strength of evidence supporting or refuting a claim that variation in a particular gene causes a particular disease. Gene-disease validity Powerpoint slides, videos, handouts, etc. for those interested in curating gene-disease pairs using the ClinGen method. Educational and training materials Documents and announcements related to Gene- Disease Clinical Validity Curation. Documents and announcements Current gene-disease pairs that have been evaluated by ClinGen for clinical validity. Gene-disease clinical validity results This module offered through the ACMG Genetics Academy is intended to provide learners with educational credit for participating in ClinGen gene curation activities. Claim educational credit for ClinGen gene- disease validity curations Detailed documentation outlining the gene disease validity process. Current standard operating procedure Currently available for ClinGen biocurators and expert panels. Click here to view a demo version. Gene curation interface Further information and policy for data contributors Interested in sharing data with ClinGen? Training materials Documents Interface Browse curations Learn more Learn more Learn more Learn more Learn more Learn more Learn more Fig. 168.2 ClinGen resource link page. Publicly accessible web resource with links to an explanation of the process for gene and variant- disease curation. The links bring users to manuals and training resources related to ClinGen’s gene curation process, which is designed to aid in evaluating the strength of a gene-disease relationship based on publicly available evidence. Genetic, experimental, and contradictory evidence curated from the literature is compiled and used to assign a clinical validity classification per criteria established by the ClinGen Gene Curation Working Group. 80,83 (From www.clinicalgenome.org.)
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 2155 168 Genetic Variants and Neonatal Disease Jennifer L. Cohen, Ludmila Francescatto, Eric W. Miller, and C. Michael Cotten INTRODUCTION Infants with complex phenotypes and infants born at early gestational age are two of the most challenging groups of patients cared for in neonatal intensive care units (NICUs). Since the initial reports of the sequence of the human genome, 1,2 clinicians have hoped that the “genomic revolu- tion” would lead to an increasing likelihood of diagnoses for infants with complex phenotypes and identification of com- mon variants associated with prematurity-associated mor- bidities. In both situations, early identification of variants linked with disease would hopefully lead to novel, genetically informed prevention and treatment strategies. Progress in molecular methodologies coupled with collaborative efforts to compile, link, and analyze phenotype and genomic data from large cohorts is enabling the identification of rare genetic variants that are likely causative of disease, and is informing care in many patient populations. 3 These efforts, with validated clinical data, collaborative data accumula- tion, and advanced analysis approaches, have improved the identification of variations in the genome that are likely to contribute to or cause disease and malformations in neo- nates. 4,5 For preterm infants, there has been limited success in identifying common genetic variants with associations to common complex morbidities such as retinopathy of prema- turity (ROP), intraventricular hemorrhage (IVH), necrotiz- ing enterocolitis (NEC), and bronchopulmonary dysplasia (BPD) that would provide insights into pathophysiology for most infants with these problems. In this chapter, we will discuss: (1) the expanding application of genetic testing in infants in the NICU with complex phenotypes and (2) the ongoing investigations to identify genetic risk factors for common morbidities seen in extremely preterm infants. GENETIC DISEASE IN NEONATES WITH COMPLEX PHENOTYPES Congenital malformations and genetic disorders have long been recognized as major contributing factors to pediatric hospitalization, morbidity, and mortality, 6–8 and this finding extends to the patient population treated in the NICU. 9–12 Gene discovery for ultra-rare Mendelian disorders is ongo- ing, and large-scale genomic technologies and international collaborative efforts have allowed for advancement in this field. 13 Along with simply identifying phenotype-genotype associations, rapid detection of genetic disorders through more expansive genomic sequencing has changed the medi- cal management of neonates with suspected genetic disease and presents new possibilities and challenges. As genomic medicine advances, the efficiency with which we are able to conduct precision pediatric medicine—accurate diagno- sis followed by tailored management based on this growing genomic information linked with clinical phenotypes—has improved and has reminded us of the importance of multidis- ciplinary teams for diagnosis and subsequent care for these infants. 5,14 As an indicator of the rapid increase in knowledge since publication of the first draft of the human genome, as of October 2001, the Online Mendelian Inheritance in Man (OMIM) database included approximately 2610 disorders with associated genetic loci. 15 The OMIM database was placed online in 1995 and now encompasses 4909 genes with phe- notype-causing mutations, and 7528 phenotypes for which the molecular basis is known (updated May 03, 2024). 16 CURRENT GENETIC TESTING APPROACHES IN THE NEONATAL INTENSIVE CARE UNIT AND THEIR EVOLUTION Initially, cytogenetic and molecular (DNA-based) genetic tests were considered separate specialties, but with the advent of new sequencing technologies, the boundaries between fields have blurred. Traditionally, cytogenetic tests focused on aneu- ploidies and large structural chromosomal rearrangements— both balanced and unbalanced—and used techniques such as G-banding karyotype (often referred to clinically as a “karyo- type”) and fluorescence in situ hybridization. The develop- ment of chromosomal microarrays (CMA) in the early 2000s maintained the ability to detect aneuploidies and unbalanced chromosomal rearrangements and added the ability to detect large regions of homozygosity and smaller copy number abnormalities such as microdeletions and microduplications, as well as identify the positions of breakpoints for all chro- mosomal alterations. CMA is still largely applied as a first-tier test in clinical cases of multiple congenital anomalies, devel- opmental delay, intellectual disability, and behavioral differ- ences like autism spectrum disorder. 17,18 In order to find even smaller insertions and deletions (indels) at the exon level and simultaneously uncover single nucleotide variants (SNVs), DNA-based sequencing tech- nology is required. Initially, this was accomplished through Sanger sequencing, and later, next-generation sequenc- ing (NGS). In high-resource settings, providers can order
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 2156 SECTION 25 — PATHOPHYSIOLOGY OF NEONATAL DISEASES therapeutics or enrollment in clinical trials/experimental therapeutics in some instances. 22–25 While there are many diagnostic and clinical advantages to broader tests such as ES and GS, aspects to consider and plan for include pretest and posttest counseling for secondary and incidental find- ings (unrelated to the clinical indication for sequencing but potentially medically actionable), how patient data can be used for data reanalysis and research, or the impact testing could have on future insurance discrimination. 26–28 In 2020 the Pediatric Exome Sequencing/Genome Sequencing Guideline Work Group set out to develop new evidence-based guidelines for the use of exome and genome sequencing in patients with congenital anomalies, develop- mental delay, and intellectual disability. This practice guide- line from the American College of Medical Genetics and Genomics (ACMG) concluded with a strong recommenda- tion for exome and genome sequencing as a first-tier or sec- ond-tier test (after CMA or focused testing) for patients with one or more congenital anomalies less than 1 year of age, or for patients with developmental delay/intellectual disability less than 18 years of age. 29 UTILIZATION AND SUCCESS RATES OF VARIOUS GENETIC TESTING METHODOLOGIES IN THE NEONATAL INTENSIVE CARE UNIT In a study aimed solely at detecting chromosomal abnormali- ties through array-based comparative genomic hybridization (a type of CMA analysis), investigators reported a detection rate of 17% for clinically significant chromosomal altera- tions among neonates with birth defects. 30 Studies aimed at identifying pathogenic variants in the protein-coding regions of Mendelian genes with broad NGS technology— ES or GS—cite diagnostic rates ranging from approximately 35% to 60%. A study of ultrarapid exome sequencing, with a goal of return of results within 5 days, in critically ill patients where the median age was 28 days cited a diagnostic rate of 51%. 31 ES in critically ill children mostly less than 1 month old found a diagnosis in 43%. 5 Infants less than 100 days old in intensive care units (ICUs) demonstrated a 36.7% diag- nostic yield with ES. 32 Acutely ill children with median age 28 days showed a diagnostic rate of 52.5% with ES. 33 Infants less than 2 years old, not necessarily in an acute care setting, demonstrated a diagnostic rate of 57.5%. 34 In a report on 307 infants from three tertiary centers in Shanghai, China, ES or sequence panels identified pathogenic genetic etiolo- gies in over 40% of the patients, including genetic etiolo- gies identified in over 60% of the infants who died during the 180-day follow-up period. Of note, and indicative of the importance of selection criteria for application of ES or GS, four clinical traits had a higher likelihood of identify- ing genetic diagnoses: integument abnormalities, complex immune-related phenotypes, mixed nervous system pheno- types and congenital anomalies, and mixed metabolism and nervous system phenotypes. 35 GS studies in critically ill infants cite diagnostic yields ranging from 20% to more than 50%. 22,24,36,37 A study of criti- cally ill infants and children cited a 42% diagnostic rate. 38 Pediatric patients less than 18 years old in the Hospital for Sick Children in Toronto, Ontario, Canada, not necessarily in the critical care setting, were found to have GS diagnos- tic rates of 41%. 39 In the report by Lionel and colleagues, these DNA-based tests as phenotype-specific gene sequenc- ing panels with copy number variant (CNV) analysis, thus merging the realms of cytogenetics and molecular genetics. When the phenotype is both complex and broad, the same technology can be used in a test known as exome sequenc- ing (ES) to investigate almost all known genomic exons and a limited number of introns (altogether accounting for approximately 1.5%–2% of the genome). ES will detect the majority of known disease-causing variant types (SNVs, indels, and CNVs). 14,19 ES can be coupled with CMA to ensure fuller coverage of exon-level copy number changes. In the clinical setting, ES can also be ordered simultaneously with sequencing of the patient’s mitochondrial DNA, if indicated. The main pur- pose of these tests is to investigate for monogenic disorders in which a pathogenic variant or biallelic pathogenic vari- ants confirm a clinical diagnosis. CMA and gene sequence panels currently maintain an active role in the field of neo- natology for well-characterized genetic diseases (e.g., micro- deletion syndromes), or a phenotypic presentation with a targeted differential diagnosis (e.g., skeletal dysplasias or neonatal seizures); in these cases, a precise test may still be an efficient path toward diagnosis, if the expense for ES is prohibitively high and time to result for a center’s cur- rently available ES test platforms remains weeks to months. Another consideration that has become available in addi- tion to the gene sequencing panels based on specific pheno- types and ES is genome sequencing (GS) as a clinical test. 4 Recognizing the possible prolonged time-to-result and resource-related hurdles of ES and GS, a study to test the potential for applying a broad sequencing panel to patients with suspected genetic disease was conducted among 20 NICU patients who had been referred to the medical genet- ics or metabolic inpatient consult services and had features suggesting an underlying genetic or metabolic condition. Twelve infants had been discharged from the NICU, and eight were enrolled prospectively. Subjects underwent a broad genomic sequencing panel identifying sequences of 4813 “disease-relevant” genes that had known associ- ated clinical phenotypes either in OMIM or the United Kingdom’s Human Gene Mutation Database (www.hgmd. cf.ac.uk/ac/index.php). 20 The investigators found a diag- nostic rate of 40%, suggesting that utilizing a broad panel for analysis of a broad list of Mendelian genes can produce high diagnostic yields at reduced costs. Of note, only 2 of the 8 infants had genetic diagnoses made by standard clini- cal approaches, inclusive of sequencing of one suspected gene in one patient and studying a limited panel of 18 sequenced genes in the other. 21 These authors cite another report of 35 infants less than 4 months old in a single ter- tiary center with suspected genetic conditions, in whom 57% had genetic diagnoses from rapid GS and data analy- sis. 22 With costs associated with sequencing and data storage decreasing and the speed of performing testing and ana- lyzing results increasing, the field of genomic medicine has been moving toward ES and GS analyses. 4 The widespread adoption of these technologies has led to gene discovery and novel syndrome characterization and has broadened the phenotypic spectrum of known disorders. The increas- ing efficiency with which these tests are able to produce a diagnosis has also allowed for individual changes in medical management, including earlier administration of targeted
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 CHAPTER 168 — GENETIC VARIANTS AND NEONATAL DISEASE 2157 testing 103 children in nongenetic service clinics, all with conditions suspected to be genetically related, all the sub- jects with molecular diagnoses eventually made by conven- tional methods were captured by GS. The 18 new diagnoses made with GS and not by conventional methods included structural and nonexonic sequence variants not detectable with ES, which were confirmed in gene-disease associations that had been recently identified, highlighting the impor- tance of periodically reviewing/reanalyzing a patient’s GS and ES results. This review of results, along with new lab- oratory-based genomic technologies, can help account for recent gene discovery and new evidence linking variants and disease. 39 In another study of 100 pediatric patients referred for genetic testing, GS identified genetic variants meeting clinical diagnostic criteria in 34% of cases, compared to 8% identified by CMA alone and 13% identified by CMA plus targeted sequencing of a small number of suspect genes. GS identified all rare clinically significant CNVs that were detected by CMA. 40 A NICU-based study from Cambridge, UK of 194 infant/parent trios used GS and relatively per- missive selection criteria and reported a diagnostic rate of 25%, 41 and that expected HPO terms were commonly lack- ing in neonates. A meta-analysis of 37 studies compared the diagnostic rate of the broadest NGS technologies—ES and GS—with chromosomal microarray and concluded that GS/ES should be considered as a first-line genetic test in children, based on the greater diagnostic and clinical util- ity of both, when compared to chromosomal microarray ( < 20% diagnostic yield). 14 Interestingly, this analysis also found that the diagnostic utility of GS (41%) was not sig- nificantly different than ES (36%). Not surprisingly, the meta-analysis found that the availability of parental samples for “trio” analysis and hospital-based interpretation (deep phenotypic information and communication between clini- cian and lab) enhanced diagnostic utility. 14 Other studies have shown higher diagnostic rates for GS among pediatric patients when parental samples are available and when the test is performed on samples obtained from a hospitalized patient. 42 Further work is required to fully elucidate optimal testing inclusion criteria and algorithms, to determine high- yield disease presentations and which critically ill children (including those without suspicion for genetic disease) are most likely to benefit from ES and GS, given the variations in cost, reimbursement, and turnaround time. 43 While GS and ES categorically appear to be advanta- geous, the ability to complete accurate sequencing as well as gene-variant characterization and clinical interpretation utilizing publicly available databases in less than 3 days, as first reported in 2012, provides compelling evidence for considering this approach as a first-line test, in sites where this approach is available. 44 A well-cited advantage for rapid ES or GS is a curtailment of the diagnostic odyssey and the ability to forgo the previously utilized stepwise approach of increasingly broader testing if initial laboratory investiga- tions are nondiagnostic or inconclusive. 44,45 The theoreti- cal diagnostic advantages of GS compared to ES are that through examination of greater than 90% of the genome, it has the capability to: (1) discover a greater percentage of single-exon CNVs, (2) uncover balanced chromosomal structural variations, (3) diagnose repeat expansions impor- tant to specific disorders such as congenital myotonic dys- trophy, and (4) discover nonexonic regulatory and splicing variations, all of which can theoretically increase the diag- nostic yield. 14,46,47 At present, however, not all clinical GS are validated to diagnose certain genetic disorders—namely repeat expansion disorders, genes with a known pseudo- gene, or imprinting disorders that rely on methylation analysis, for instance. 18,26 The technology is moving toward capturing some of these capabilities (specifically the repeat expansion disorders and pseudogenes) as additional posi- tive cases become available for test validation and new soft- ware tools are developed. 46,47 Therefore GS will likely be able to deliver the advantages of ES while still providing cru- cially important genomic information typically uncovered by cytogenetic techniques like CMA, which can be absent/ incomplete from ES. 14,18,48 In one study comparing GS and ES, 25.7% of patients who had a negative exome analysis were found to have a diagnostic variant on subsequent GS. 39 Additionally, the raw data generated for a single patient through GS is more comprehensive than the data generated by ES. This allows for future reanalysis that may be benefi- cial, as our knowledge of the genome increases to include a fuller understanding of the nonexonic regions. 19 In 2023, the Human Pangenome Reference Consortium (HPRC), a project funded by the National Human Genome Research Institute, released a new reference human genome com- prised of genome sequences of 47 ethnically diverse individ- uals. The pangenome, along with improved variant calling methods, holds the promise of further improving clinical diagnostic testing in the future. 49,50 Several previously published studies have shown the high clinical impact that reanalysis of ES data has on confirming a genetic diagnosis and subsequent patient care. Reanalysis of variants of uncertain significance (VUS) after a period of 1 year results in an increase of diagnostic yield of approxi- mately 10% to 20%. This yield can further increase after 2 years, reaching approximately 30% while also eliminating some VUS as likely benign or benign. Moreover, the inclu- sion of new phenotypic data and clinical features can fur- ther enhance the likelihood of a higher diagnostic yield. It is also important to take into consideration the value of resequencing, which one study showed to be responsible for 29% of novel genetic diagnoses. 51–53 Sequencing technology continues to improve at a rapid pace, and advancements in chemistry and bioinformatics allow for better coverage and higher-quality data. Thinking toward the future, a patient’s RNA may be used to conduct clinical RNA sequencing, which may one day complement DNA analysis and help elucidate certain diagnoses. 54–57 RNA sequencing (RNA-seq) adds remarkable value to the clinical setting and diagnostic rate as it can func- tionally characterize the effect of VUS, including noncanon- ical splice variants and noncoding variants affecting gene expression. As demonstrated by Yépez and colleagues, RNA- seq provided a genetic diagnosis to 16% of ES-inconclusive cases. 58 One of the highest hurdles that remains with such broad testing is the large number of genomic variants uncovered by sequencing and the associated time- and labor-inten- sive variant analysis and interpretation. Researchers are actively pursuing ways to partially automate this process and decrease the turnaround time by utilizing clinical natural language processing to parse the clinical phenotype from the patient’s electronic medical record. 4 Studies conducted
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 2158 SECTION 25 — PATHOPHYSIOLOGY OF NEONATAL DISEASES within the NICU patient population have thus far reported successful outcomes, quoting diagnostic percentages of 42% to 57%, clinical management alterations (as a result of the diagnosis) in 30% to 72%, and a change in outcomes experi- enced by 24% to 34% of patients in the studies who received rapid GS. 4,22,24,36,38 A recent review study by Kingsmore and colleagues summarized the findings of 44 studies utiliz- ing rapid genomic sequencing. In addition to providing a genetic diagnosis in 37% of children in ICUs with diseases of unknown etiology or with suspected genetic disorders, it changed the clinical management in 26% of the cases, and significantly reduced the net healthcare costs. 43 Rapid ES (rES) and GS (rGS) in the NICU have become an essen- tial tool to identify and confirm genetic diagnoses. One recent prospective study from a multicenter group in the Netherlands showed that rES in neonates has a higher diag- nostic yield, faster diagnosis, and an increase in cost savings when compared to conventional genetic testing. 59 However, rGS has been shown to have these same advantages in addi- tion to a higher diagnostic yield when compared to rES. 36,60 Moreover, ultrarapid GS provides results within as little as 3 days, allowing for quicker diagnoses, proper clinical treat- ment, and decreased hospitalization stays. 61,62 Recently, mul- tiple Israeli medical centers performed the first prospective trio rGS study for neonates in critical condition. With a turn- around time (TAT) of approximately 7 days and a 50% to 62% diagnostic rate, the incorporation of trio rGS into the NICU holds valuable clinical utility. 63 IMPLEMENTATION AND IMPLICATIONS OF BROADER GENOMIC TESTING A concern that arises is that the cost and informatics demands of broader genomic technologies like ES and GS remain greater than more targeted NGS or cytogenetic testing. 64 Research has shown, however, that rapid use of these broader technologies when implemented with some degree of automation to their variant interpretation may reduce overall costs for healthcare systems. 24 Recent studies have replicated a net healthcare cost savings with the use of rapid ES. 65 Savings have been directly related to the speed with which a diagnosis is made, providing additional evi- dence that utilization of rapid GS in the critically ill infant as a first-tier test may have economic as well as clinical ben- efits. Many of the diagnoses discovered by rapid ES or GS are rare genetic diseases, for which there are few reported cases and published literature, and for which a targeted dif- ferential diagnosis or curated NGS gene panel may simply not exist. For patients without a diagnosis following the broader tests of GS or ES, having the ability to periodically repeat analysis of the patient’s GS or ES data in the setting of expanding genotype-phenotype databases is another ben- efit of these broader genomic tests being performed early in life. Online networks such as GeneMatcher (see Table 168.1 for this and other useful genomic websites) allow pro- viders to submit through an online portal a candidate gene (a gene of uncertain clinical significance) in an attempt to connect with other providers and researchers who may have a patient that matches their own patient’s phenotype and has the same gene affected. In this way, patient cohorts can be collected and functional studies regarding the gene can be undertaken to determine whether the candidate gene is responsible for human disease, and ultimately may lead to the characterization of new ultra-rare Mendelian condi- tions. Separately, progress regarding the genomic variants associated with more common medical concerns, such as preterm birth, may also result from assembling accurately phenotyped cohorts inclusive of thousands of unaffected infants and affected newborns, and applying some of these more advanced genomic technologies such as GS, RNA sequencing, and DNA methylation analyses to these patients on a broad scale. 66 Researchers are working on methods to overcome some of these limitations, in part by creating par- tially automated methods for phenotyping. 67 As has been noted, GS is only as useful as the clinical phenotypic descrip- tion provided, the molecular methodologies used, and the variant analysis and interpretation conducted. Genetics and genomics resources are growing, but the growth and availability of tests and support staff inevitably fall short of demand, creating limited access to testing, which leads to healthcare inequities. 21,27 In 2023, insurance companies, including United Health Care, expanded coverage to enable GS as a clinical test in the pediatric setting. Rapid GS is also covered for hospitalized infants in a few hospitals in the United States by Medicaid. Despite recent changes in the coverage status of GS testing, access can be limited due not only to the geographic region but also to a lack of understanding of the strengths and limitations of this technology. Therefore there is an unde- rutilization of GS testing, which in the hospital setting can result in diagnostic delay, potentially affecting treatment management, and creating a burden to families and health- care providers involved. 68 Furthermore, access to genetic testing is not readily available to the entirety of the popula- tion, and the availability of clinical geneticists and genetic counselors at healthcare institutions can be limited. This is especially true in regions with smaller populations and in more rural areas. Additionally, access to genetic services at larger institutions in more populated areas is often not easily accessible. Regardless of socioeconomic status or insurance coverage, underrepresented populations are frequently at a disadvantage. 69–71 Genetic disorders can often go untreated, as the treating physician may not be familiar with the com- plexity of different genetic diseases. 71,72 Moreover, the lack of genetic testing across underrepresented populations con- tributes to a lack of diagnostic yield, reduced reporting rates of pathogenic variants, and disparities in population genetic studies. 71,73,74 Another aspect of inequity currently inherent in all NGS- based testing is the reliance, in part, on large population databases and variant frequency in determining pathogenic- ity of a particular variant. Large population databases have an underrepresentation of ancestral minority groups, often leaving patients from underrepresented minorities with a higher percentage of VUS on their clinical test report, and oftentimes leading to a nondiagnostic test result. 74–77 As broad NGS technology becomes more widely available, labo- ratories will have a greater number of control data that may help begin to resolve some of these VUS. Additionally, new methodologies are being actively pursued that may further enhance our ability to resolve variants. 78 The effort to build consistency into up-to-date classifica- tion of the gene-disease and variant-disease relationship is through the National Human Genome Research Institute
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 CHAPTER 168 — GENETIC VARIANTS AND NEONATAL DISEASE 2159 (NHGRI)-funded Clinical Genome Resource (ClinGen). 79 ClinGen’s mission is to define the clinical relevance of genes and variants to be used in research and precision medicine (Fig. 168.1). ClinGen has developed tools to evaluate clini- cal validity of gene-disease associations and pathogenicity of genetic variants for use in clinical care, as well as educational material for clinicians and patients to better understand the process of curation of gene- and variant-disease associations. Curation is a term used to describe the analysis of the exist- ing data and evidence about a specific gene or gene vari- ant and its relation to a specific phenotype. The tools used allow for a quantitative approach to weighing the evidence supporting gene-disease and variant-disease associations to reach summary categories: pathogenic, likely pathogenic, uncertain significance, likely benign, and benign. ClinGen’s work and impact are made directly available on their web- site (https://clinicalgenome.org/) and additionally, via another resource, ClinVar. ClinVar is the NCBI archival database that aggregates information about genomic varia- tion and relationships to human health that are provided by researchers, clinical laboratories conducting sequenc- ing, expert groups, clinics, and patient registries. In addi- tion, with genomeconnect.org (an online registry designed by ClinGen), patients may also submit their clinical genetic test reports and health information to increase research- ers’ and healthcare providers’ overall understanding of the relationship between genetics and health. While single investigators may submit a variant suspected of association with a specific disease, ClinVar submissions are scored based on the number and types of sources submitting the same data, and validation by working groups and expert panels that examine the genetic epidemiologic evidence, as well as supportive evidence from experimental model systems. These curations, as well as user interfaces, are constantly evolving, as data from multiple sources become available for consideration by expert panels. 81,82 Currently, there are 28 working groups, 54 gene curation panels, and 69 vari- ant curation panels working within the broader ClinGen effort (https://clinicalgenome.org/). The ClinGen website includes links to multiple educational modules and publicly available browser tools (Fig. 168.2). Clinicians should be aware that different automated cura- tion tools in use in academic and commercial laboratories may make different “calls” of significance of associations between specific variants and disease. Clinicians should also be aware that data continue to accumulate, and the number of expert panels that have been trained using ClinGen pro- grams is also growing, with the aim to decrease the hetero- geneity of interpretation about the significance of gene/ variant/disease associations. As the legitimacy of the data- bases increases with replication, clinicians must continue to be vigilant and cautious, as increasing knowledge may some- times lead to a change in the classification of a variant. As a cautionary example, a recent report analyzing variants that Table 168.1 Web Resources Providing Genetic and Genomic Information and Training. ClinGen: a National Institutes of Health (NIH)-funded resource dedicated to building a central resource that defines the clinical rel- evance of genes and variants for use in precision medicine and research: https://clinicalgenome.org/ ClinVar: freely accessible, public archive of reports of the relationships among human variations and phenotypes, with supporting evi- dence: https://www.ncbi.nlm.nih.gov/clinvar/ Database of Genotype and Phenotype (dbGaP): an archive of data from genome-wide association studies on a variety of diseases and conditions accessible through the NCBI: https://www.ncbi.nlm.nih.gov/gap/ DECIPHER: a database of reported copy number variants and linked phenotypes: https://decipher.sanger.ac.uk/ Ensembl: a genome browser for vertebrate genomes that supports research in comparative genomics, evolution, sequence variation, and transcriptional regulation. Ensembl annotates genes, computes multiple alignments, predicts regulatory function, and collects disease data: https://www.ensembl.org/index.html GeneMatcher: a website that enables connections between clinicians who have a patient with a candidate or ultra-rare gene and researchers who have an interest in that gene: https://genematcher.org/ GeneReviews: a clinical resource for many genetic conditions that provides clinically actionable information, including diagno- sis, inheritance, and management, as well as a differential diagnosis of related conditions: https://www.ncbi.nlm.nih.gov/books/ NBK1116/ GenomeConnect: an online registry designed by the Clinical Genome Resource (ClinGen) for people who are interested in sharing de-identified genetic and health information to improve understanding of genetics and health: https://www.genomeconnect. org/ Online Mendelian Inheritance in Man (OMIM): a searchable database of clinical features, phenotypes, and genes: https://omim.org/ Unique: a website with patient-/family-facing resources regarding chromosome and gene disorders: https://www.rarechromo.org/ University of California Santa Cruz Genome Browser: a website created initially to ensure public access to the initial human genome assembly; has now evolved to include a broad collection of vertebrate and model organism assemblies and annotations, along with a large suite of tools for viewing, analyzing, and downloading data: https://genome.ucsc.edu/ Leiden Open Variation Database (LOVD): An online gene-centered collection and display of DNA variants; its purpose is to provide a flexible, freely available tool for genomic variant and phenotype collection, display, and curation . LOVD allows both patient-centered and gene-centered views. LOVD is open source, released under the GPL license, and is actively being improved. On the server in Leiden, LOVD offers free hosting and support of LOVD-powered gene variant databases. https://www.lovd.nl/
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 2160 SECTION 25 — PATHOPHYSIOLOGY OF NEONATAL DISEASES had been classified as pathogenic or likely pathogenic, and also occurred in more than 0.5% of the population, found that of 217 variants in 173 genes that were selected for cura- tion, 87 (40%) of the variants were downgraded to benign, likely benign, or variant of uncertain significance. 84 Obtaining informed consent prior to testing is critical for the purpose of educating patients and families about possible testing outcomes. 27,85,86 Some outcomes can cause unexpected difficulties, including incidental but clini- cally relevant findings not related to the phenotype that prompted the test but linked with adult-onset diseases; VUS, which can lead to ambiguity and possible frustration; and candidate gene discovery that may require many years of research-based functional analysis before the gene’s rele- vance to human disease is conclusively determined. The case for the adoption of widespread GS will arise from its successful utilization across a variety of clinical indications and patient cohorts. As researchers who have successfully built the infrastructure to support effective application of this technology continue to expand their networks, patients, families, and caregivers may begin to see the benefit of this testing through earlier diagnosis and potentially a change in clinical management. These tests will also necessitate con- tinued monitoring of variants either for evolving evidence of clinical significance, in the case of VUS, or the confirma- tion of persistent classification among variants determined to cause disease. 87 Currently, most labs perform ES/GS using Illumina’s state-of-the-art “short-read” sequencing-by-synthesis. 88 Short- read ES/GS has enabled the identification of a plethora of pathogenic variants in Mendelian genes. Additionally, it has fostered the discovery of novel disease-associated genes, thus finally providing an accurate diagnosis to patients and families that have spent years on a diagnostic odyssey. 89 Despite advances in clinical genetic testing and the utiliza- tion of the most comprehensive genetic testing clinically available, an estimate of approximately 50% of patients with a suspected genetic disorder remain undiagnosed. One of the factors contributing to this modest diagnostic rate is the use of short-read sequencing, which fails to pro- vide coverage in clinically challenging genomic regions such as repeat expansion regions. Emerging technologies, such as long-read DNA sequencing, are able to overcome this limitation since they use long DNA fragments ranging in size from 1000 to 20,000 bases. Voted the method of the year in 2022, long-read sequencing (LRS), also described as “third-generation sequencing,” aims to interrogate whole human genomes and uncover novel variants inaccessible with short-read sequencing. 90 When compared to short- read genome sequencing (SRS), LRS has many theoretical advantages. Along with increased read length, LRS has the ability to sequence native DNA through repetitive genomic regions such as long homopolymer stretches, pseudogenes, and short tandem repeat regions. Additionally, it is capable of performing haplotype phasing (variants are assigned to the homologous paternal or maternal chromosomes), epi- genetic detection (with the power to determine the methyla- tion status of CpG sites), and RNA sequencing. Lastly, LRS enables de novo genome assembly to allow for more effec- tive evaluation of structural variation. 91 Pacific Biosciences (PacBio) and Oxford Nanopore Technologies (ONT) are leaders in the LRS arena, although other companies, includ- ing Illumina, have also developed LRS methods. 91 ONT uses nanopore sequencing, where an ionic cur- rent can drive single-stranded DNA (originally dsDNA dis- sociated via an adaptable motor protein) or RNA through a nanopore membrane. Alterations in charge determine nucleotide type as the nucleic acid passes through the nano- pores. 92 Advances in nanopore sequencing technology have resulted in enhanced sequencing speeds reaching ~450 base pairs/second, increased read lengths of up to 2.3 MB, and accurate resolution of complex structural variants. 93,94 Raw sequencing and methylation data can be provided in real time following base calling using nanopore sequencing. This sequencing technology also has an adaptive sampling ability, allowing for the enrichment of specific regions of the genome. Additionally, nanopore sequencing allows for direct RNA sequencing, providing rapid and concise tran- scriptomic analysis. 91,95,96 PacBio HiFi LRS works by ligating hairpin adapters at both ends of the DNA fragment, creating a circular mole- cule (also known as SMRTbell) that is repeatedly sequenced. This technology is able to generate highly accurate reads of DNA molecules ( > 99.9%; QV > 30) with an attached cost of $1000 per genome. 97 Combined, both PacBio and ONT technologies have increased the number of pathogenic vari- ants identified and previously missed by clinical genetic testing. As the capabilities of nanopore sequencing have improved, so has its implementation in the clinic. Recent studies have demonstrated that nanopore sequencing, in conjunction with the appropriate analysis tools, can sequence a human genome in less than 2 hours, provide candidate variants for disease within 8 total hours, and provide a 42% diagnos- tic rate. 98 A striking example of this was the identification Improved patient care through genomic medicine Building a genomic knowledge base ClinVar and other resources ClinGen’s critical questions Sharing genetic and health data Patients Clinicians Laboratories Researchers Is this gene associated with a disease? Clinical validity Is this variant causative? Pathogenicity Is this information actionable? Clinical utility Fig. 168.1 Clinical genomic (ClinGen) resource framework. ClinGen’s overall mission is to build a genomic knowledge base to improve patient care. The ClinGen framework provides a semiquantitative mea- surement for the strength of evidence of a gene-disease relationship that correlates to a qualitative classification: “Definitive,” “Strong,” “Moderate,” “Limited,” “No Reported Evidence,” or “Conflicting Evidence.” Within the ClinGen structure, classifications derived with this framework are reviewed and confirmed or adjusted based on the clinical expertise of appropriate disease experts. Detailed guidance for utilizing this framework and access to the curation interface are available on the ClinGen website. 80 (From https://clinicalgenome.org/ start/#loc_1550536143-7476-1.)
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 CHAPTER 168 — GENETIC VARIANTS AND NEONATAL DISEASE 2161 of a pathogenic heterozygous variant in a 3-month old who presented with status epilepticus and seizure semiol- ogy. Within 9 hours, a likely pathogenic variant in CSNK2B was detected, resulting in a definitive diagnosis of Poirier- Bienvenu neurodevelopmental syndrome. 99 Other studies have demonstrated the importance of nanopore sequencing in several genetic disorders, including Angelman syndrome, Duchenne muscular dystrophy, and retinitis pigmentosa. Nanopore sequencing allowed for the identification of a complex structural variant in DMD , confirming a genetic diagnosis for an individual, even after routine genetic test- ing, transcriptomic, and dystrophin protein analysis failed to yield a genetic diagnosis. 100 Additionally, the clinical sig- nificance of a UBE3A variant in an individual with Angelman syndrome was confirmed by identifying the parental allele of this variant using nanopore sequencing. 101 The application of nanopore sequencing also allowed for the identification of likely pathogenic structural variants in individuals with Gene-disease clinical validity curation The ClinGen Gene curation working group has developed a framework to standardize the approach to determine the clinical validity for a gene-disease pair. This framework: • Defines the criteria needed to assess clinical validity, • Describes the evidence supporting a gene-disease association in a semi-quantitative manner, and • Allows curators to use this information to methodically classify the validity of a given gene-disease pair. The ClinGen Gene-disease clinical validity curation process involves evaluating the strength of evidence supporting or refuting a claim that variation in a particular gene causes a particular disease. Gene-disease validity Powerpoint slides, videos, handouts, etc. for those interested in curating gene-disease pairs using the ClinGen method. Educational and training materials Documents and announcements related to Gene- Disease Clinical Validity Curation. Documents and announcements Current gene-disease pairs that have been evaluated by ClinGen for clinical validity. Gene-disease clinical validity results This module offered through the ACMG Genetics Academy is intended to provide learners with educational credit for participating in ClinGen gene curation activities. Claim educational credit for ClinGen gene- disease validity curations Detailed documentation outlining the gene disease validity process. Current standard operating procedure Currently available for ClinGen biocurators and expert panels. Click here to view a demo version. Gene curation interface Further information and policy for data contributors Interested in sharing data with ClinGen? Training materials Documents Interface Browse curations Learn more Learn more Learn more Learn more Learn more Learn more Learn more Fig. 168.2 ClinGen resource link page. Publicly accessible web resource with links to an explanation of the process for gene and variant- disease curation. The links bring users to manuals and training resources related to ClinGen’s gene curation process, which is designed to aid in evaluating the strength of a gene-disease relationship based on publicly available evidence. Genetic, experimental, and contradictory evidence curated from the literature is compiled and used to assign a clinical validity classification per criteria established by the ClinGen Gene Curation Working Group. 80,83 (From www.clinicalgenome.org.)
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 2162 SECTION 25 — PATHOPHYSIOLOGY OF NEONATAL DISEASES retinitis pigmentosa, ultimately securing them a molecular diagnosis. 102 Similarly, several studies have demonstrated the clinical utility of HiFi sequencing. A study performed on six pro- band-parent trios affected with neurodevelopmental disor- ders revealed a de novo L1-mediated insertion, resulting in duplication of exon 3 in CDKL5 , in one of the probands, and a de novo complex structural variant (SV) affecting at least three chromosomes, in another proband. 103 Another study identified potential pathogenic variants in 16/96 (16.7%) undiagnosed patients with negative ES/GS. 104 Kilich and colleagues have recently demonstrated the power of using HiFi LRS as a single assay to simultaneously detect a dele- tion, provide phase, and biallelic hypermethylation of the MEG3:TSS-DMR region in a patient with Kagami-Ogata syn- drome. 105 These are just a few studies illustrating PacBio LRS as a critical technology to improve diagnostic yield. Due to recent improvements in data quality/accuracy, higher throughput, and lower costs, LRS has become more attractive in the clinical setting. Although over the past few years, several publications have demonstrated the clinical utility of these technologies, widespread adoption of LRS in the clinical setting will require thorough validation from a lab perspective to ensure accurate and reproducible results. In an effort to increase the diagnostic rate in pediatric patients with genetic conditions, PacBio and GeneDx recently entered into a research collaboration with the University of Washington. Specifically, this study will harness the power of HiFi long- read GS (increased accuracy, read-length, and methylation) and compare the diagnostic rate between short- and long- read GS. Future studies will continue to demonstrate the advantages of utilizing LRS and determine whether these technologies will complement or eventually replace conven- tional genome-wide testing, including in the NICU setting. GENOMICS OF COMMON COMPLEX DISEASES ASSOCIATED WITH PREMATURITY While NGS-based testing continues to inform clinicians and families about the genes and variants that contribute to complex, severe phenotypes in NICU patients, identifying genes and variants associated with the more common com- plex disorders of prematurity, such as ROP, IVH, NEC, and BPD, remains a challenge. Why look for genetic variants? Morbidities of prematurity are strongly associated with mortality and longer-term neurode- velopmental disabilities. 106–108 Identifying whether genetic variations play a role in the pathophysiology of these diseases could allow for the development of mechanistic, individu- alized prevention and treatment approaches based on the individual patient’s genotype. Dosing and treatment strate- gies based on genotype exist in several diseases, usually associ- ated with adult diseases, including cancer, with various tumor genotypes guiding therapeutic decisions. 109 In addition, when management trials targeting these morbidities are under- taken, taking into account genetic variation as a potential confounder on the effect of specific management strategies may be important, as in the interactions of certain CYP2C19 genotypes in combination with clopidogrel and aspirin in adults with minor stroke and transient ischemic attack. 110,111 Of note, for preterm infants, there is a suggestion, but not definitive evidence, of genotypic influence on response to medical therapy for the patent ductus arteriosus. 112,113 The first mandatory steps to investigate before going to the trouble of testing genetic variants for associations with disease include testing whether or not inherited fac- tors are likely contributing to the risk of a specific disease, either directly or via genetics-environment interactions. Quantitatively, this association is measured as heritability, which is a quantitative measure of the extent to which genetic factors account for the phenotypic variance. The process, which includes testing disease prevalence among monozygotic versus dizygotic twins, has been carefully out- lined by Bhandari and Gruen. 114 In studies with mono- and dizygotic twins, significant heritability has been identified for ROP, NEC, sepsis, IVH, and BPD. 115–120 With the establishment of an inherited basis for these common complex diseases, researchers started to test asso- ciations between relatively common variants, known as “sin- gle nucleotide polymorphisms” or “SNPs” that have been identified in 1% or more of tested populations. Having the minor (less common) allele can sometimes increase or decrease the functionality of the gene product. Because they are relatively common, methodologies to genotype SNPs were characterized fairly early in the “genomics era” after the successful sequencing of the entire human genome. Much of the early work on diseases in neonatology focused on assessing associations between SNPs in genes for TNF α or IL-6, which are associated with immune response or regula- tion of inflammation, and diseases such as BPD or late-onset sepsis, often with disparate results for the same SNP. 121,122 Tests for associations with a handful of SNPs in relatively small cohorts dominated early reports, and some identi- fied associations with statistical significance with P < .05; however, with 3 billion nucleotides comprising the genome, over 30 million known SNPs, thousands of variants in the genome of any one individual, and relatively small sample sizes, replication has proven difficult, and the concept of the candidate gene analysis, or, more specifically, candidate SNP analysis, approach was called into question unless study cohorts included thousands of cases and controls. 123,124 With early evidence encouraging a more agnostic approach, broader arrays of genes and variants were devel- oped, and multiple candidate genes across the genome could be tested. With knowledge of the genome expanding, SNPs that are in linkage disequilibrium (the occurrence of a set of SNPs forming a haplotype) could be predicted accu- rately without actually genotyping every SNP. This allows for “imputation” of multiple “tagSNPs,” in which actual genotyp- ing of one SNP allows prediction of the genotypes of 5 to 10 neighboring SNPs (i.e., haplotype). 125 In the genome-wide association studies (GWAS) approach, hundreds of thou- sands, even millions, of SNPs are genotyped and tagSNPs imputed, and then individually tested for association in well-phenotyped cohorts (Fig. 168.3). Statistical approaches have had to be adapted to the millions of tested associations using GWAS. The standard has been to determine “genome- wide significance” when the P values identified with the com- parison reached lower than 5 × 10 − 8 , much lower than the conventional P < .05 used with debatable authority in many epidemiology studies and clinical trials. 126–128 To realistically test hypotheses using GWAS, large cohorts with thousands of
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 CHAPTER 168 — GENETIC VARIANTS AND NEONATAL DISEASE 2163 subjects with the disease phenotype are ideal, as have been used for multiple complex common diseases of adulthood, such as prostate cancer and Alzheimer disease. 129,130 GWAS studies in the NICU population have numbered in the hun- dreds and rarely thousands, and therefore are much smaller than the adult cohorts. 131–136 As a comparison, in a GWAS of 2727 cases and 3336 controls for Alzheimer disease, APOE gene variants were found to be associated with Alzheimer disease with a P value of 2.52 × 10 − 53 . 137 An additional challenge for genomic analysis among extremely preterm cohorts is the differences in minor allele frequency among study cohorts of differing ancestry. While studies on samples from relatively homogenous nations, such as Iceland, may produce fairly homogenous, replicable results, results from cohorts of mixed ancestry, which is more common in the United States, present challenges to analyz- ing associations between variants that differ in prevalence in cohorts of different ancestry. Acknowledging those strate- gies and challenges, studies in cohorts of premature infants for genetic associations using candidate gene and SNP anal- yses and more agnostic GWAS approaches have produced some findings of interest that reach genomic significance within the study cohort. These results, while encouraging, continue to be only tempting because of the severe limita- tion of examining relatively small cohorts compared to the more typical larger population analyses that can be done for common, complex diseases that emerge later in life. 138,139 RETINOPATHY OF PREMATURITY Recent reviews of genetic risk factors for ROP document the lack of any emergent, identifiable, strongly associ- ated genetic variants with severe ROP risk among preterm infants. Certainly, ROP-like phenotypes are linked with rare inherited variants or spontaneous mutations causing diseases like familial exudative vitreoretinopathy (FEVR) and Kabuki syndrome, but for the vast majority of prema- ture infants, candidate gene/SNP studies have not iden- tified variants with genome-wide significance. Multiple attempts have been made to identify associations that are plausibly associated with the pathophysiologic mecha- nisms of ROP, including VEGF and VEGF receptors. One report, which tested a panel of over 1000 variants in over 100 genes related to inflammation and organ develop- ment, identified two intronic variants in the brain-derived neurotrophic factor (BDNF) gene with severe ROP versus nonsevere or no ROP with a P value of less than 5 × 10 − 7 . 140 Lower serum levels of BDNF have been associated with higher likelihood of ROP in reports testing associations of serum levels of various cytokines and growth factors with outcomes in premature infants. This adds some plausibility to an association with BDNF variants that might influence expression, or an association with some other gene prod- uct/component of a pathway important for neurovascular development that includes BDNF. 141–143 While this is prom- ising, the impact of specific variants on BDNF circulating levels, or in situ levels in the developing retina, has not been described. More recently, analysis beyond genotyping has identified that placental CpG methylation, a measure of epigenetic modification, of 12 different genes is associated with the development of pre-threshold ROP. Interestingly, the genes with methylation changes associated with ROP included BDNF. 144 A recent GWAS study performed with patients enrolled in the iROP consortium revealed a novel ROP risk association with the SNP GLI3 rs2058019 reach- ing genome-wide significance ( P = 4.90E − 09). 145 This is the first study to identify an ROP-associated variant. Further studies are needed to validate these findings. NECROTIZING ENTEROCOLITIS In a cohort study using a GWAS approach, minor allele(s) in a cluster of SNPs spanning a 43-kb region of chromo- some 8 (8q23.3) conferred an odds ratio of 4.72 (95% con- fidence interval [CI]: 2.51–8.88) for elevated risk of NEC, with multiple SNPs associated with P < 10 − 8 (Fig. 168.4). Two smaller clusters on chromosomes 14 and 11 exhib- ited P values of 10 − 7 to 10 − 8 . Like many gene association studies done in the extremely preterm population, this analysis was limited by a small sample size ( n = 751, only 30 with surgical NEC), from multiple sites in the United States, with significant ancestry admixture. Interestingly, the increased risk was similar for all three genetic ances- tries represented in this population. 146 The investigators attempted to validate the associations of the SNPs in the chromosome 8 region with NEC in a separate cohort ( N = 1018, 26 with surgical NEC) of premature infants enrolled in a study using GWAS to identify variants associated with severe intraventricular hemorrhage. 134 In that cohort, one of the SNPs (rs13252246) was associated with NEC with SNP 3 Associated to disease SNP 3 SNP 2 No association to disease SNP 2 SNP 1 No association to disease Using a chip can genotype 500,000–5 million SNPs SNP 1 Individuals with disease Individuals without disease Fig. 168.3 Genome-wide association studies (GWAS). In GWAS, genotypes of single nucleotide polymorphisms, that is, loci with variants identified in 1% or more of the population, are identi- fied across the entire genome. Prevalence of the variants among individuals with a condition is compared with prevalence among individuals without the condition to identify candidate genes and variants associated with disease. SNP , Single nucleotide poly- morphism. (From www.genome.gov/about-genomics/fact-sheets/ Genome-Wide-Association-Studies-Fact-Sheet.)
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 2164 SECTION 25 — PATHOPHYSIOLOGY OF NEONATAL DISEASES a P value of 0.02. In in silico analysis, a term that means use of computer modeling systems to predict gene prod- ucts and their potential activities and pathologies, the NEC-associated region of chromosome 8 appears to be evolutionarily conserved, but without any previously iden- tified genes. Pathway analysis, testing variants in multiple pathway-linked genes for associations with NEC, identified associations with over 50 pathways, including pathways involved with immune response, growth regulation, and G-protein signaling. Interestingly, the LTB4R gene, close to the chromosome 14 region with a strong association with surgical NEC, encodes an eicosanoid receptor. Eicosanoid receptor signaling was found to be the second most promi- nent pathway affected by NEC-associated SNPs. In terms of physiological plausibility, LTB4 plays a significant role in a toll-like receptor 4 and cyclooxygenase-2–mediated mecha- nism of intestinal ischemia/reperfusion injury. 146 INTRAVENTRICULAR HEMORRHAGE Several candidate gene studies have been conducted in relatively small cohorts testing associations between SNPs in inflammation, complement, and coagulation pathways. 147,148 One case report of two preterm siblings suggested a novel mutation in the COL4A1 gene may be associated with severe IVH. 149 In the precursor to a GWAS study, the Gene Targets for Intraventricular Hemorrhage study group conducted a candidate gene study of SNPs in 7 genes, in 224 preterm infants with grade III-IV IVH, and 389 matched controls. Only SNPs in the methylenetetra- hydrofolate reductase (MTHFR) gene gave even equivo- cal results of associations with severe IVH. 150 In the GWAS analysis, the group tested over 600,000 SNPs in 458 inborn appropriate for gestational age neonates with severe IVH and 866 infants without IVH, from US and Scandinavian cohorts. 134 No individual SNP reached genome-wide sig- nificance; however, a 10-SNP haplotype ranging from the intergenic region of GM140 and CACNA1E to the intron region of CACNA1E had a P value of 7.16 × 10 − 10 . CACNA1E (calcium channel, voltage-dependent, R type, α 1E subunit) is mutated in a Mendelian form of hemiplegic migraine, a known vascular phenotype, as well as various epilepsies. 151 None of the prior candidate individual SNPs reached genome-wide association significance. Like many similar endeavors, the summary of the report of results of this GWAS for severe IVH concluded, “Because common variants have small-to-moderate effects, a large-scale neo- natal genomic medicine network must be developed with the infrastructural capacity to host an accessible database of sequence variants, their phenotypic associations and environmental risk factors.” 134 GS studies have identified several candidate genetic variants linked to IVH, including one recent study dem- onstrating an association between 13 SNPs and IVH, five of which may also be linked to neuronal and neurovas- cular development. 152 The German Neonatal Network performed a multicenter study of preterm infants and identified SNPs in APOE2 and APOE4 , the gene products of which are involved in vitamin K transport, that increase the risk of IVH in preterm infants. 153 Furthermore, a study of over 100 preterm infants diagnosed with IVH showed that five SNPs in FN1 were detected in 47.2% of the cohort, demonstrating the strong link between fibronec- tin function and IVH. 154 BRONCHOPULMONARY DYSPLASIA Of the morbidities of extreme prematurity, BPD is the most common. 155,156 While severe ROP, NEC, and severe IVH have phenotypes that are usually defined by imaging or direct visualization, BPD is usually defined by some level of respiratory support, making the anatomic and physiologic details of the phenotype that might be influenced by genetic variations somewhat challenging. 157,158 In part because of relatively small sample sizes and the limitations related to phenotype definitions that are not anatomically or physi- ologically defined, the five groups that have used an agnos- tic GWAS approach to try to identify variants or groups of 10 –log 10 Position (by chromosome) 0 2 4 6 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ( P) Fig. 168.4 Necrotizing enterocolitis (NEC) genome-wide association studies (GWAS) result. “Manhattan Plot” of over 7 million genotyped and imputed single nucleotide polymorphisms tested in the GWAS for surgical NEC versus controls. Data shown are − log P values on the Y axis versus chromosome locations along the X axis. The region on chromosome 8 at the location 8q23.3 shows the strongest association with the incidence of NEC. 146
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 CHAPTER 168 — GENETIC VARIANTS AND NEONATAL DISEASE 2165 variants in genes and pathways associated with epidemiolog- ically defined BPD have met with limited success. 131–133,135,136 Hadchouel and colleagues studied 418 premature neonates (gestational age below 28 weeks, 22% with BPD, and a repli- cation cohort of 213 Finnish neonates, 26% with BPD) and identified SNPs in the SPOCK2 gene associated with BPD in the discovery cohorts (of African and Caucasian ancestry) and the replication cohort. 131 Wang and colleagues studied over 2000 very low-birth-weight infants born in California and did not identify genomic loci or pathways associated with moderate-to-severe BPD with genome-wide signifi- cance. 132 In the study of Ambalavanan and colleagues, which included 751 infants with birth weights ranging between 401 and 1000 g (428 diagnosed with BPD or who died), no SNPs achieved genome-wide significance. Pathways of lung development and repair, and novel molecules and pathways (adenosine deaminase, targets of microRNA or miR-219) were associated with BPD. Exosomal miRNAs have also been linked to BPD. Recent evidence suggests that decreased expression of one biomarker, miR 876-3p, may potentially allow for the prediction of BPD in low-birth-weight infants. 159 Pathways associated with mild BPD were different from the pathways associated with severe BPD. In addition, the vari- ants/pathways associated with BPD varied by ancestry. 133,160 In the fourth study, GWAS was performed in a Finnish dis- covery cohort of 60 cases with moderate-to-severe BPD and 114 controls. The SNP flanking the C-reactive protein gene had the strongest association with BPD ( P = 3.4 × 10 − 6 ), and in multivariate logistic regression, this SNP was associated with BPD in two replication cohorts (one Finnish and one comprised of European and African individuals). 135 In the most recent GWAS study, which tested 9 million genotypes and imputed variants from 387 preterm infants who partici- pated in the Trial of Late Surfactant (TolSURF) study for associations with BPD, no individual SNPs were associated with BPD with genome-wide significance. Ultimately, the five GWAS studies failed to identify a specific gene or variant in a gene to be specifically and convincingly associated with BPD; however, the studies, individually and collectively, are limited by small sample sizes and high levels of phenotype variation. In addition to GWAS studies, epigenetic signatures are now being considered as potential contributors to BPD. Recent epigenome-wide association studies have shown that alterations in methylation in several genes, including CTSH and RASGRF1 , may be associated with BPD. 161 Investigators have also applied ES to BPD, with hopes that by much more extensive sequencing, which includes rare variants, that the collective influence of these rare vari- ants may be detected. 162–164 Carrera and colleagues studied 26 unrelated infants with severe BPD from Italian NICUs. Identified variants were classified as likely to have high, mod- erate, or low impact based on predicted protein effects and whether or not they were novel variants in genes with occur- rence in more than one subject. As expected, each subject had approximately 200,000 identified variants, with about 10% of all variants identified from all subjects as possibly having moderate or high impact. In each sample, approxi- mately 100 variants were identified that were hypothesized by structural predictive analysis to have an impact on pro- tein structure. Two subjects had novel missense mutations in ABCA3 , which is a gene responsible for transporting phospholipids to lamellar bodies in type II alveolar cells. Rare and novel variants in NOS2 and toll-receptor genes and C-reactive protein were also identified in this cohort with BPD. 162 Li and colleagues used newborn screening samples from the state of California for exome sequencing of 50 twin pairs, including 51 infants with BPD. The accumulated vari- ants in 258 genes in the subjects with BPD had a significantly higher haploinsufficiency score (haploinsufficiency can be defined as the situation when one copy of a gene is either deleted or has a loss-of-function variant, and dosage of the gene product is reduced enough to affect function 165 ). Variants clustered in genes in pathways involved in embry- onic epithelial development, organization of collagen, and Wnt-signaling were increased among the infants with BPD compared with non-BPD infants. This group also looked at tissue expression in human tissue from patients with BPD, and in lung tissue from the hyperoxic mouse model of BPD, and detected increased expression in the exome-identified BPD candidate genes in hyperoxia-exposed animals. 163 Most recently, Hamvas and colleagues studied 146 subjects (85 with BPD and 61 unaffected) enrolled in the Prematurity and Respiratory Outcomes Program (PROP) with ES. This group tested for associations between disease status and individual common (minor allele frequency > 0.05) and rare variants, in affected or unaffected subjects. Three hun- dred forty-five genes with extremely rare, nonsynonymous variants, that is, variants that would lead to a change in the amino acid sequence of the gene product, were identified only in the BPD-affected subjects. This study, the largest to date, replicated 28 genes with extremely rare variants in patients with BPD that were previously associated with BPD in the California cohort reported by Li and colleagues. 164 Additional ES studies have also linked several genes to BPD, including BIVM and KMT2A , in addition to providing more evidence for an association between SPOCK2 altera- tions and this disease. 164,166,167 In summary, there is some progress in identifying genetic risk factors for BPD, includ- ing associations between BPD and epigenetic changes. The complexities of the phenotype of BPD, combined with the relatively small size of the study populations to date, con- tinue to be limiting factors in the validation of the relatively new association findings resulting from newer molecular methodologies. CONCLUSION Twenty years ago, the ability to clinically utilize the genetic sequence of a NICU patient with a complex disorder, to have the informatics and genomic analysis capacity to deter- mine within 48 hours whether the patient likely possesses a pathogenic genetic variant that may direct care, and to have access to a growing database of genotypes and associ- ated phenotypes to help guide our knowledge, were only possible in our imaginations. Since then, the application of genomic technology and methodologies in the NICU has become more of a reality, particularly with the emerging reports of utilization of ES and GS with rapid turnaround times relevant to care in the NICU. With the growing avail- ability of these types of tests, NICU clinicians will grapple with challenges related to the follow-up of unexpected find- ings and VUS. Clinicians must remain aware of the very real possibility of modifications to the results of these tests, as
Order your copy of Polin and Fox Fetal and Neonatal Physiology, 7th Edition at elsevierhealth.com/9780443128233 2166 SECTION 25 — PATHOPHYSIOLOGY OF NEONATAL DISEASES more information may arise to either validate a variant’s pathogenicity or shift a particular variant’s classification toward benign or uncertain. Because of the complexity and potential for new findings in the future, it is imperative that NICU providers caring for these infants and ordering these tests connect with providers who have the expertise and abil- ity to stay current with accumulating gene-disease associa- tion data and the ability to clinically follow and manage the patients and families long term. The increasing knowledge of the genome, as well as multiple other “-omics” related sciences such as transcriptomics, proteomics, and metabo- lomics, as well as studies of the microbiome, are elucidating mechanisms of the more common complex diseases of pre- maturity like BPD, 168 as well as preterm birth itself. 169 While the impact of this latter work is felt less at the clinical level currently, it continues to affirm studies regarding mecha- nisms of disease and disease prevention and therapy, and will undoubtedly inform future care in NICU. A complete list of references can be found online at Ebooks.health.elsevier.com SELECTED REFERENCES 4. Clark MM, Hildreth A, Batalov S, et al. Diagnosis of genetic dis- eases in seriously ill children by rapid whole-genome sequencing and automated phenotyping and interpretation. Sci Transl Med . 2019;11(489):eaat6177. 5. Freed AS, Clowes Candadai SV, Sikes MC, et al. The impact of rapid exome sequencing on medical management of critically ill children [published online ahead of print, 2020 Jun 15]. J Pediatr . 2020;S0022–3476(20):30721–30726. 12. Michel MC, Colaizy TT, Klein JM, Segar JL, Bell EF. Causes and circumstances of death in a neonatal unit over 20 years. Pediatr Res . 2018;83(4):829–833. 14. Clark MM, Stark Z, Farnaes L, et al. Meta-analysis of the diagnostic and clinical utility of genome and exome sequencing and chromo- somal microarray in children with suspected genetic diseases. NPJ Genom Med . 2018;3:16. 18. Lalonde E, Rentas S, Lin F, Dulik MC, Skraban CM, Spinner NB. Genomic diagnosis for pediatric disorders: revolution and evolu- tion. Front Pediatr . 2020;8:373. 19. Bick D, Jones M, Taylor SL, Taft RJ, Belmont J. Case for genome sequencing in infants and children with rare, undiagnosed or genetic diseases. J Med Genet . 2019;56(12):783–791. 24. Farnaes L, Hildreth A, Sweeney NM, et al. Rapid whole-genome sequencing decreases infant morbidity and cost of hospitalization. NPJ Genom Med . 2018;3:10. 25. Ceyhan-Birsoy O, Murry JB, Machini K, et al. Interpretation of genomic sequencing results in healthy and Ill newborns: results from the babyseq PROJECT. Am J Hum Genet . 2019;104(1):76–93. 27. Gyngell C, Newson AJ, Wilkinson D, Stark Z, Savulescu J. Rapid challenges: ethics and genomic neonatal intensive care. Pediatrics . 2019;143(suppl 1):S14–S21. 28. Smith EE, du Souich C, Dragojlovic N, CAUSES Study RAPIDOMICS Study Elliott AM. Genetic counseling considerations with rapid genome-wide sequencing in a neonatal intensive care unit. J Genet Couns . 2019;28(2):263–272. 36. Petrikin JE, Cakici JA, Clark MM, et al. The NSIGHT1-randomized controlled trial: rapid whole-genome sequencing for accelerated etiologic diagnosis in critically ill infants. NPJ Genom Med . 2018;3:6. 39. Lionel AC, Costain G, Monfared N, et al. Improved diagnostic yield compared with targeted gene sequencing panels suggests a role for whole-genome sequencing as a first-tier genetic test. Genet Med . 2018;20(4):435–443. 43. Saunders CJ, Miller NA, Soden SE, et al. Rapid whole-genome sequencing for genetic disease diagnosis in neonatal intensive care units. Sci Transl Med . 2012;4(154): 154ra135. 60. Landrum MJ, Kattman BL. ClinVar at five years: delivering on the promise. Hum Mutat . 2018;39(11):1623–1630. 79. Landrum MJ, Chitipiralla S, Brown GR, et al. ClinVar: improve- ments to accessing data. Nucleic Acids . 2020;48(D1):D835–D844. 86. Lantos JD. Ethical and psychosocial issues in whole genome sequencing (WGS) for newborns. Pediatrics . 2019;143(suppl 1):S1–S5. 114. Bhandari V, Gruen JR. What is the basis for a genetic approach in neonatal disorders? Semin Perinatol . 2015;39(8):568–573. 125. Hirschhorn JN, Daly MJ. Genome-wide association studies for com- mon diseases and complex traits. Nat Rev Genet . 2005;6(2):95–108. 127. Ioannidis JPA, Khoury MJ. Evidence-based medicine and big genomic data. Hum Mol Genet . 2018;27(R1):R2–R7. 131. Hadchouel A, Durrmeyer X, Bouzigon E, et al. Identification of SPOCK2 as a susceptibility gene for bronchopulmonary dysplasia. Am J Respir Crit Care Med . 2011;184(10):1164–1170. 132. Wang H, St Julien KR, Stevenson DK, et al. A genome-wide asso- ciation study (GWAS) for bronchopulmonary dysplasia. Pediatrics . 2013;132(2):290–297. 133. Ambalavanan N, Cotten CM, Page GP, et al. Integrated genomic analyses in bronchopulmonary dysplasia. J Pediatr . 2015;166(3):531– 537.e13. 134. Ment LR, Ådén U, Bauer CR, et al. Genes and environment in neonatal intraventricular hemorrhage. Semin Perinatol . 2015;39(8): 592–603. 135. Mahlman M, Karjalainen MK, Huusko JM, et al. Genome-wide asso- ciation study of bronchopulmonary dysplasia: a potential role for variants near the CRP gene. Sci Rep . 2017;7(1):9271. 136. Torgerson DG, Ballard PL, Keller RL, et al. Ancestry and genetic associations with bronchopulmonary dysplasia in preterm infants. Am J Physiol Lung Cell Mol Physiol . 2018;315(5):L858–L869. 138. Ioannidis JP, Ntzani EE, Trikalinos TA. Racial’ differences in genetic effects for complex diseases. Nat Genet . 2004;36(12):1312–1318. 139. Popejoy AB, Fullerton SM. Genomics is failing on diversity. Nature . 2016;538(7624):161–164. 146. Jilling T, Ambalavanan N, Cotten CM, et al. Surgical necrotizing enterocolitis in extremely premature neonates is associated with genetic variations in an intergenic region of chromosome 8. Pediatr Res . 2018;83(5):943–953. 162. Carrera P, Di Resta C, Volonteri C, et al. Exome sequencing and pathway analysis for identification of genetic variability relevant for bronchopulmonary dysplasia (BPD) in preterm newborns: a pilot study. Clin Chim Acta . 2015;451(Pt A):39–45. 163. Li J, Yu KH, Oehlert J, et al. Exome sequencing of neonatal blood spots and the identification of genes implicated in bronchopulmo- nary dysplasia. Am J Respir Crit Care Med . 2015;192(5):589–596. 164. Hamvas A, Feng R, Bi Y, et al. Exome sequencing identifies gene variants and networks associated with extreme respiratory out- comes following preterm birth. BMC Genet . 2018;19(1):94. 168. Lal CV, Bhandari V, Ambalavanan N. Genomics, microbiomics, proteomics, and metabolomics in bronchopulmonary dysplasia. Semin Perinatol . 2018;42(7):425–431.