Modeling seizures in the Human Phenotype Ontology according to contemporary ILAE concepts makes big phenotypic data tractable

Ontology accelerates few-shot learning capability of large language model: A study in extraction of drug efficacy in a rare pediatric epilepsy

OBJECTIVE

Dravet Syndrome (DS) is a developmental and epileptic encephalopathy that is characterized by severe, prolonged motor seizures and high resistance to multiple antiseizure medications (ASMs) with multiple comorbidities. Evaluating the efficacy of new drugs in DS preclinical models and mapping them to human phenotypes of DS through analysis of published literature is an important goal for improving outcomes in this rare pediatric epilepsy.

MATERIALS AND METHODS

Large language models (LLM) have demonstrated great promise in parsing published literature; however, the performance of LLMs falls short in medical applications. In this study, we investigate the effectiveness of domain ontology developed by human experts to optimize LLMs for medical text processing in a rare disease. Utilizing a benchmark dataset that describes the efficacy of 17 ASMs tested in preclinical models and DS patients, we define a new ontology-augmented phased in-context learning (PCL) approach to process 4935 full-text DS articles. We expand this analysis to 7 new drugs that demonstrate efficacy in reducing seizures to identify gaps in current knowledge for designing new experimental studies for drug discovery in DS.

RESULTS

Few-shot or in-context learning is a foundational capability of LLMs and the few-shot learning capability of the Gemini 1.0 Pro version LLM dramatically increases when we augment prompts with the DS epilepsy ontology. The DS epilepsy ontology is the largest epilepsy and seizure ontology in clinical use that was developed by DS basic scientists and clinical neurologists. The ontology-augmented PCL prompt achieves 100% accuracy in reproducing the benchmark drug efficacy dataset for 17 ASMs with only two examples for in-context learning.

CONCLUSION

The new ontology-augmented PCL approach significantly accelerates the few-shot learning capabilities of the Gemini LLM, thereby reducing the number of required examples and time needed to optimize LLMs for medical applications.

Deciphering the Natural History of SCN8A-Related Disorders

BACKGROUND AND OBJECTIVES

SCN8A-related disorders encompass a range of neurodevelopmental and epilepsy phenotypes. However, despite representing one of the most common epilepsy-associated channelopathies, its longitudinal phenotypes remain largely uncharacterized.

METHODS

In this study, we harmonized electronic medical record data from 82 individuals with SCN8A-related disorders to reconstruct the natural history of the disorder in comparison with a cohort of 2,833 individuals with known or presumed genetic epilepsies.

RESULTS

Compared with the cohort of other known or presumed genetic epilepsies, those with SCN8A-related disorders (mean age = 8.3 years, 52% female) had >10-fold odds of bilateral tonic-clonic seizures as early as at 1 year (p = 1.70 × 10-14, OR 10.56, CI 5.85-18.90). Individuals carrying gain-of-function (GOF) SCN8A variants had particularly high seizure risk at 6 months (p = 0.007/pthreshold = 4.25 × 10-4, OR 4.71, CI 1.36-21.25) and an increased risk of global developmental delay as early as at 3 months (p = 0.002/pthreshold = 4.72 × 10-5, OR 5.67, CI 1.74-20.23) when compared with the broader SCN8A cohort. Individuals with loss-of-function variants were more likely to have atypical absence seizures, most prominently at 4.25 years (p = 0.013/pthreshold = 7.08 × 10-4, OR 32.71, CI 1.44-2,193.51). Compared with the broader SCN8A cohort, individuals with the recurrent p.Arg850Gln variant were more likely to have infantile spasms at 6 months and those with variants at the p.Arg1872Trp/Gln/Leu hotspot were more likely to have neonatal seizures. Individuals with the recurrent p.Gly1475Arg variant were more likely to have active epilepsy after 5 years of age. In later childhood, focal seizures were more prominent in individuals with the recurrent p.Arg1617Gln variant while generalized-onset seizures were more prominent with the p.Asn1877Ser variant. We also established the effectiveness of sodium channel blockers in managing SCN8A epilepsy in individuals carrying GOF variants and those whose variants have not been functionally characterized, suggesting that many unstudied SCN8A variants may have GOF mechanisms.

DISCUSSION

SCN8A-related disorders distinguish themselves from other genetic epilepsies by the frequent bilateral tonic-clonic seizures in infancy, prominent early epileptic and developmental features in GOF variant carriers, and unique seizure phenotypes in those with recurrent variants. Our study provides a longitudinal perspective on SCN8A-related disorders, paving the way for future precision medicine approaches.

Updated classification of epileptic seizures: Position paper of the International League Against Epilepsy

The International League Against Epilepsy (ILAE) has updated the operational classification of epileptic seizures, building upon the framework established in 2017. This revision, informed by the implementation experience, involved a working group appointed by the ILAE Executive Committee. Comprising 37 members from all ILAE regions, the group utilized a modified Delphi process, requiring a consensus threshold of more than two thirds for any proposal. Following public comments, the Executive Committee appointed seven additional experts to the revision task force to address and incorporate the issues raised, as appropriate. The updated classification maintains four main seizure classes: Focal, Generalized, Unknown (whether focal or generalized), and Unclassified. Taxonomic rules distinguish classifiers, which are considered to reflect biological classes and directly impact clinical management, from descriptors, which indicate other important seizure characteristics. Focal seizures and those of unknown origin are further classified by the patient's state of consciousness (impaired or preserved) during the seizure, defined operationally through clinical assessment of awareness and responsiveness. If the state of consciousness is undetermined, the seizure is classified under the parent term, that is, the main seizure class (focal seizure or seizure of unknown origin). Generalized seizures are grouped into absence seizures, generalized tonic-clonic seizures, and other generalized seizures, now including recognition of negative myoclonus as a seizure type. Seizures are described in the basic version as with or without observable manifestations, whereas an expanded version utilizes the chronological sequence of seizure semiology. This updated classification comprises four main classes and 21 seizure types. Special emphasis was placed on ensuring translatability into languages beyond English. Its aim is to establish a common language for all health care professionals involved in epilepsy care, from resource-limited areas to highly specialized centers, and to provide accessible terms for patients and caregivers.

Phenotypic analysis of 11,125 trio exomes in neurodevelopmental disorders

Genomic sequencing is widely used to identify causative genetic changes in neurodevelopmental disorders, such as autism, intellectual disability, and epilepsy. Most neurodevelopmental disorders also present with diverse clinical features, and delineating the interaction between causative genetic changes and phenotypic features is a key prerequisite for developing personalized therapies. However, assessing clinical features at a scale that parallels genomic sequencing remains challenging. Here, we standardize phenotypic information across 11,125 patient-parent trios with exome sequencing data using biomedical ontologies, analyzing 674,767 phenotypic terms. We find that individuals with de novo variants in 69 out of 261 neurodevelopmental genes exhibit statistically significant clinical similarities with distinct phenotypic fingerprints. We also observe that phenotypic relatedness follows a gradient, spanning from highly similar to dissimilar phenotypes, with intra-gene similarities suggesting clinically distinct subgroups for seven neurodevelopmental genes. For most genetic etiologies, only a small subset of highly phenotypically similar individuals carried de novo variants in the same gene, highlighting the heterogeneous and complex clinical landscape of neurodevelopmental disorders. Our study provides a large-scale overview of the dynamic relationship between genotypes and phenotypes in neurodevelopmental disorders, underscoring how the inherent complexity of these conditions can be deciphered through approaches that integrate genomic and phenotypic data.

Clinical signatures of genetic epilepsies precede diagnosis in electronic medical records of 32,000 individuals

PURPOSE

An early genetic diagnosis can guide the time-sensitive treatment of individuals with genetic epilepsies. However, most genetic diagnoses occur long after disease onset. We aimed to identify early clinical features suggestive of genetic diagnoses in individuals with epilepsy through large-scale analysis of full-text electronic medical records.

METHODS

We extracted 89 million time-stamped standardized clinical annotations using Natural Language Processing from 4,572,783 clinical notes from 32,112 individuals with childhood epilepsy, including 1925 individuals with known or presumed genetic epilepsies. We applied these features to train random forest models to predict SCN1A-related disorders and any genetic diagnosis.

RESULTS

We identified 47,774 age-dependent associations of clinical features with genetic etiologies a median of 3.6 years before molecular diagnosis. Across all 710 genetic etiologies identified in our cohort, neurodevelopmental differences between 6 to 9 months increased the likelihood of a later molecular diagnosis 5-fold (P < .0001, 95% CI = 3.55-7.42). A later diagnosis of SCN1A-related disorders (area under the curve [AUC] = 0.91) or an overall positive genetic diagnosis (AUC = 0.82) could be reliably predicted using random forest models.

CONCLUSION

Clinical features predictive of genetic epilepsies precede molecular diagnoses by up to several years in conditions with known precision treatments. An earlier diagnosis facilitated by automated electronic medical records analysis has the potential for earlier targeted therapeutic strategies in the genetic epilepsies.

Advances in big data and omics: Paving the way for discovery in childhood epilepsies

The insights gained from big data and omics approaches have transformed the field of childhood genetic epilepsy. With an increasing number of individuals receiving genetic testing for seizures, we are provided with an opportunity to identify clinically relevant subgroups and extract meaningful observations from this large-scale clinical data. However, the volume of data from electronic medical records and omics (e.g., genomics, transcriptomics) is so vast that standardized methods, such as the Human Phenotype Ontology, are necessary for reliable and comprehensive characterization. Here, we explore the integration of clinical and omics data, highlighting how these approaches pave the way for discovery in childhood epilepsies.

Genome-wide identification and phenotypic characterization of seizure-associated copy number variations in 741,075 individuals

Copy number variants (CNV) are established risk factors for neurodevelopmental disorders with seizures or epilepsy. With the hypothesis that seizure disorders share genetic risk factors, we pooled CNV data from 10,590 individuals with seizure disorders, 16,109 individuals with clinically validated epilepsy, and 492,324 population controls and identified 25 genome-wide significant loci, 22 of which are novel for seizure disorders, such as deletions at 1p36.33, 1q44, 2p21-p16.3, 3q29, 8p23.3-p23.2, 9p24.3, 10q26.3, 15q11.2, 15q12-q13.1, 16p12.2, 17q21.31, duplications at 2q13, 9q34.3, 16p13.3, 17q12, 19p13.3, 20q13.33, and reciprocal CNVs at 16p11.2, and 22q11.21. Using genetic data from additional 248,751 individuals with 23 neuropsychiatric phenotypes, we explored the pleiotropy of these 25 loci. Finally, in a subset of individuals with epilepsy and detailed clinical data available, we performed phenome-wide association analyses between individual CNVs and clinical annotations categorized through the Human Phenotype Ontology (HPO). For six CNVs, we identified 19 significant associations with specific HPO terms and generated, for all CNVs, phenotype signatures across 17 clinical categories relevant for epileptologists. This is the most comprehensive investigation of CNVs in epilepsy and related seizure disorders, with potential implications for clinical practice.

Individualised human phenotype ontology gene panels improve clinical whole exome and genome sequencing analytical efficacy in a cohort of developmental and epileptic encephalopathies

BACKGROUND

The majority of genetic epilepsies remain unsolved in terms of specific genotype. Phenotype-based genomic analyses have shown potential to strengthen genomic analysis in various ways, including improving analytical efficacy.

METHODS

We have tested a standardised phenotyping method termed 'Phenomodels' for integrating deep-phenotyping information with our in-house developed clinical whole exome/genome sequencing analytical pipeline. Phenomodels includes a user-friendly epilepsy phenotyping template and an objective measure for selecting which template terms to include in individualised Human Phenotype Ontology (HPO) gene panels. In a pilot study of 38 previously solved cases of developmental and epileptic encephalopathies, we compared the sensitivity and specificity of the individualised HPO gene panels with the clinical epilepsy gene panel.

RESULTS

The Phenomodels template showed high sensitivity for capturing relevant phenotypic information, where 37/38 individuals' HPO gene panels included the causative gene. The HPO gene panels also had far fewer variants to assess than the epilepsy gene panel.

CONCLUSION

We have demonstrated a viable approach for incorporating standardised phenotype information into clinical genomic analyses, which may enable more efficient analysis.

Enriching representation learning using 53 million patient notes through human phenotype ontology embedding

The Human Phenotype Ontology (HPO) is a dictionary of >15,000 clinical phenotypic terms with defined semantic relationships, developed to standardize phenotypic analysis. Over the last decade, the HPO has been used to accelerate the implementation of precision medicine into clinical practice. In addition, recent research in representation learning, specifically in graph embedding, has led to notable progress in automated prediction via learned features. Here, we present a novel approach to phenotype representation by incorporating phenotypic frequencies based on 53 million full-text health care notes from >1.5 million individuals. We demonstrate the efficacy of our proposed phenotype embedding technique by comparing our work to existing phenotypic similarity-measuring methods. Using phenotype frequencies in our embedding technique, we are able to identify phenotypic similarities that surpass current computational models. Furthermore, our embedding technique exhibits a high degree of agreement with domain experts' judgment. By transforming complex and multidimensional phenotypes from the HPO format into vectors, our proposed method enables efficient representation of these phenotypes for downstream tasks that require deep phenotyping. This is demonstrated in a patient similarity analysis and can further be applied to disease trajectory and risk prediction.

Leveraging electronic medical record-embedded standardised electroencephalogram reporting to develop neonatal seizure prediction models: a retrospective cohort study

BACKGROUND

Accurate prediction of seizures can help to direct resource-intense continuous electroencephalogram (CEEG) monitoring to neonates at high risk of seizures. We aimed to use data from standardised EEG reports to generate seizure prediction models for vulnerable neonates.

METHODS

In this retrospective cohort study, we included neonates who underwent CEEG during the first 30 days of life at the Children's Hospital of Philadelphia (Philadelphia, PA, USA). The hypoxic ischaemic encephalopathy subgroup included only patients with CEEG data during the first 5 days of life, International Classification of Diseases, revision 10, codes for hypoxic ischaemic encephalopathy, and documented therapeutic hypothermia. In January, 2018, we implemented a novel CEEG reporting system within the electronic medical record (EMR) using common data elements that incorporated standardised terminology. All neonatal CEEG data from Jan 10, 2018, to Feb 15, 2022, were extracted from the EMR using age at the time of CEEG. We developed logistic regression, decision tree, and random forest models of neonatal seizure prediction using EEG features on day 1 to predict seizures on future days.

FINDINGS

We evaluated 1117 neonates, including 150 neonates with hypoxic ischaemic encephalopathy, with CEEG data reported using standardised templates between Jan 10, 2018, and Feb 15, 2022. Implementation of a consistent EEG reporting system that documents discrete and standardised EEG variables resulted in more than 95% reporting of key EEG features. Several EEG features were highly correlated, and patients could be clustered on the basis of specific features. However, no simple combination of features adequately predicted seizure risk. We therefore applied computational models to complement clinical identification of neonates at high risk of seizures. Random forest models incorporating background features performed with classification accuracies of up to 90% (95% CI 83-94) for all neonates and 97% (88-99) for neonates with hypoxic ischaemic encephalopathy; recall (sensitivity) of up to 97% (91-100) for all neonates and 100% (100-100) for neonates with hypoxic ischaemic encephalopathy; and precision (positive predictive value) of up to 92% (84-96) in the overall cohort and 97% (80-99) in neonates with hypoxic ischaemic encephalopathy.

INTERPRETATION

Using data extracted from the standardised EEG report on the first day of CEEG, we predict the presence or absence of neonatal seizures on subsequent days with classification performances of more than 90%. This information, incorporated into routine care, could guide decisions about the necessity of continuing EEG monitoring beyond the first day, thereby improving the allocation of limited CEEG resources. Additionally, this analysis shows the benefits of standardised clinical data collection, which can drive learning health system approaches to personalised CEEG use.

FUNDING

Children's Hospital of Philadelphia, the Hartwell Foundation, the National Institute of Neurological Disorders and Stroke, and the Wolfson Foundation.

Computational analysis of neurodevelopmental phenotypes: Harmonization empowers clinical discovery

Making a specific diagnosis in neurodevelopmental disorders is traditionally based on recognizing clinical features of a distinct syndrome, which guides testing of its possible genetic etiologies. Scalable frameworks for genomic diagnostics, however, have struggled to integrate meaningful measurements of clinical phenotypic features. While standardization has enabled generation and interpretation of genomic data for clinical diagnostics at unprecedented scale, making the equivalent breakthrough for clinical data has proven challenging. However, increasingly clinical features are being recorded using controlled dictionaries with machine readable formats such as the Human Phenotype Ontology (HPO), which greatly facilitates their use in the diagnostic space. Improving the tractability of large-scale clinical information will present new opportunities to inform genomic research and diagnostics from a clinical perspective. Here, we describe novel approaches for computational phenotyping to harmonize clinical features, improve data translation through revising domain-specific dictionaries, quantify phenotypic features, and determine clinical relatedness. We demonstrate how these concepts can be applied to longitudinal phenotypic information, which represents a critical element of developmental disorders and pediatric conditions. Finally, we expand our discussion to clinical data derived from electronic medical records, a largely untapped resource of deep clinical information with distinct strengths and weaknesses.

Current practice in diagnostic genetic testing of the epilepsies

Epilepsy genetics is a rapidly developing field, in which novel disease-associated genes, novel mechanisms associated with epilepsy, and precision medicine approaches are continuously being identified. In the past decade, advances in genomic knowledge and analysis platforms have begun to make clinical genetic testing accessible for, in principle, people of all ages with epilepsy. For this reason, the Genetics Commission of the International League Against Epilepsy (ILAE) presents this update on clinical genetic testing practice, including current techniques, indications, yield of genetic testing, recommendations for pre- and post-test counseling, and follow-up after genetic testing is completed. We acknowledge that the resources vary across different settings but highlight that genetic diagnostic testing for epilepsy should be prioritized when the likelihood of an informative finding is high. Results of genetic testing, in particular the identification of causative genetic variants, are likely to improve individual care. We emphasize the importance of genetic testing for individuals with epilepsy as we enter the era of precision therapy.

Precision medicine for genetic epilepsy on the horizon: Recent advances, present challenges, and suggestions for continued progress

The genetic basis of many epilepsies is increasingly understood, giving rise to the possibility of precision treatments tailored to specific genetic etiologies. Despite this, current medical therapy for most epilepsies remains imprecise, aimed primarily at empirical seizure reduction rather than targeting specific disease processes. Intellectual and technological leaps in diagnosis over the past 10 years have not yet translated to routine changes in clinical practice. However, the epilepsy community is poised to make impressive gains in precision therapy, with continued innovation in gene discovery, diagnostic ability, and bioinformatics; increased access to genetic testing and counseling; fuller understanding of natural histories; agility and rigor in preclinical research, including strategic use of emerging model systems; and engagement of an evolving group of stakeholders (including patient advocates, governmental resources, and clinicians and scientists in academia and industry). In each of these areas, we highlight notable examples of recent progress, new or persistent challenges, and future directions. The future of precision medicine for genetic epilepsy looks bright if key opportunities on the horizon can be pursued with strategic and coordinated effort.

Prenatal phenotyping: A community effort to enhance the Human Phenotype Ontology

Technological advances in both genome sequencing and prenatal imaging are increasing our ability to accurately recognize and diagnose Mendelian conditions prenatally. Phenotype-driven early genetic diagnosis of fetal genetic disease can help to strategize treatment options and clinical preventive measures during the perinatal period, to plan in utero therapies, and to inform parental decision-making. Fetal phenotypes of genetic diseases are often unique and at present are not well understood; more comprehensive knowledge about prenatal phenotypes and computational resources have an enormous potential to improve diagnostics and translational research. The Human Phenotype Ontology (HPO) has been widely used to support diagnostics and translational research in human genetics. To better support prenatal usage, the HPO consortium conducted a series of workshops with a group of domain experts in a variety of medical specialties, diagnostic techniques, as well as diseases and phenotypes related to prenatal medicine, including perinatal pathology, musculoskeletal anomalies, neurology, medical genetics, hydrops fetalis, craniofacial malformations, cardiology, neonatal-perinatal medicine, fetal medicine, placental pathology, prenatal imaging, and bioinformatics. We expanded the representation of prenatal phenotypes in HPO by adding 95 new phenotype terms under the Abnormality of prenatal development or birth (HP:0001197) grouping term, and revised definitions, synonyms, and disease annotations for most of the 152 terms that existed before the beginning of this effort. The expansion of prenatal phenotypes in HPO will support phenotype-driven prenatal exome and genome sequencing for precision genetic diagnostics of rare diseases to support prenatal care.

The clinical and biochemical hallmarks generally associated with GLUT1DS may be caused by defects in genes other than SLC2A1

Glucose transporter 1 deficiency syndrome (GLUT1DS) is a neurometabolic disorder caused by haploinsufficiency of the GLUT1 glucose transporter (encoded by SLC2A1) leading to defective glucose transport across the blood–brain barrier. This work describes the genetic analysis of 56 patients with clinical or biochemical GLUT1DS hallmarks. 55.4% of these patients had a pathogenic variant of SLC2A1, and 23.2% had a variant in one of 13 different genes. No pathogenic variant was identified for the remaining patients. Expression analysis of SLC2A1 indicated a reduction in SLC2A1 mRNA in patients with pathogenic variants of this gene, as well as in one patient with a pathogenic variant in SLC9A6, and in three for whom no candidate variant was identified. Thus, the clinical and biochemical hallmarks generally associated with GLUT1DS may be caused by defects in genes other than SLC2A1.

Clinical and therapeutic significance of genetic variation in the GRIN gene family encoding NMDARs

Considerable genetic variation of N-methyl-d-aspartate receptors (NMDARs) has recently become apparent, with many hundreds of de novo variants identified through widely available clinical genetic testing. Individuals with GRIN variants present with neurological conditions such as epilepsy, autism, intellectual disability (ID), movement disorders, schizophrenia and behavioral disorders. Determination of the functional consequence of genetic variation for NMDARs should lead to precision therapeutics. Furthermore, genetic animal models harboring human variants have the potential to reveal mechanisms that are shared among different neurological conditions, providing strategies that may allow treatment of individuals who are refractory to therapy. Preclinical studies in animal models and small open label trials in humans support this idea. However, additional functional data for variants and animal models corresponding to multiple individuals with the same genotype are needed to validate this approach and to lead to thoughtfully designed, randomized, placebo-controlled clinical trials, which could provide data in order to determine safety and efficacy of potential precision therapeutics.

Seizures, semiology, and syndromes: A narrative review

Clinical seizure signs continue to be of central importance to guide diagnosis, classification, treatment and prognosis. Some basic principles guide history-taking and observation in clinical epileptology. The information contained within subjective seizure descriptions can be framed within standardized vocabulary and a classification of ictal signs, seizure types, and the integrated framework of epilepsy syndromes. As illustrative examples, we discuss the historical origins and current research context of Dravet syndrome and Janz syndrome, two genetic epilepsy syndromes. In candidates for epilepsy surgery, ictal signs aid us in identifying the symptomatogenic zone and hence delineating the ictal onset zone. Here, historical reports from Victor Horsley and Hughlings Jackson provide valuable perspective on clinical reasoning. Lastly, the information contained within clinical signs and syndromes presents an indispensable data source in future efforts of large-scale genotype-phenotype correlations and machine learning methods.