Development and Validation of a Prediction Model for Early Diagnosis of SCN1A-Related Epilepsies

Abstract

Background and objectives: Pathogenic variants in the neuronal sodium channel α1 subunit gene (SCN1A) are the most frequent monogenic cause of epilepsy. Phenotypes comprise a wide clinical spectrum, including severe childhood epilepsy; Dravet syndrome, characterized by drug-resistant seizures, intellectual disability, and high mortality; and the milder genetic epilepsy with febrile seizures plus (GEFS+), characterized by normal cognition. Early recognition of a child's risk for developing Dravet syndrome vs GEFS+ is key for implementing disease-modifying therapies when available before cognitive impairment emerges. Our objective was to develop and validate a prediction model using clinical and genetic biomarkers for early diagnosis of SCN1A-related epilepsies.

Methods: We performed a retrospective multicenter cohort study comprising data from patients with SCN1A-positive Dravet syndrome and patients with GEFS+ consecutively referred for genetic testing (March 2001-June 2020) including age at seizure onset and a newly developed SCN1A genetic score. A training cohort was used to develop multiple prediction models that were validated using 2 independent blinded cohorts. Primary outcome was the discriminative accuracy of the model predicting Dravet syndrome vs other GEFS+ phenotypes.

Results: A total of 1,018 participants were included. The frequency of Dravet syndrome was 616/743 (83%) in the training cohort, 147/203 (72%) in validation cohort 1, and 60/72 (83%) in validation cohort 2. A high SCN1A genetic score (133.4 [SD 78.5] vs 52.0 [SD 57.5]; p < 0.001) and young age at onset (6.0 [SD 3.0] vs 14.8 [SD 11.8] months; p < 0.001) were each associated with Dravet syndrome vs GEFS+. A combined SCN1A genetic score and seizure onset model separated Dravet syndrome from GEFS+ more effectively (area under the curve [AUC] 0.89 [95% CI 0.86-0.92]) and outperformed all other models (AUC 0.79-0.85; p < 0.001). Model performance was replicated in both validation cohorts 1 (AUC 0.94 [95% CI 0.91-0.97]) and 2 (AUC 0.92 [95% CI 0.82-1.00]).

Discussion: The prediction model allows objective estimation at disease onset whether a child will develop Dravet syndrome vs GEFS+, assisting clinicians with prognostic counseling and decisions on early institution of precision therapies (http://scn1a-prediction-model.broadinstitute.org/).

Classification of evidence: This study provides Class II evidence that a combined SCN1A genetic score and seizure onset model distinguishes Dravet syndrome from other GEFS+ phenotypes.

Figure 1. Study Overview

Study workflow. Genetic data (SCN1A genetic score) and clinical data (age at seizure onset in months) from 743 patients (training cohort) were introduced to a supervised machine learning approach to produce a prediction model. We tested the prediction model with 2 independent blinded validation cohorts (n = 275). GEFS+ = genetic epilepsy with febrile seizures plus.

Figure 2. Training Cohort Data

(A) Density plot showing the distribution of the age at seizure onset in training cohort patients with Dravet syndrome (purple area) and genetic epilepsy with febrile seizures plus (GEFS+) (gray area). (B) Density plot showing the distribution of the SCN1A genetic score in training cohort patients with Dravet syndrome (purple area) and GEFS+ (gray area). Statistical difference between the observed means was evaluated with the Wilcoxon test.

Figure 3. Training Cohort Model Performance: ROC Curve Analysis

Receiver operating characteristic (ROC) curves showing the relationship between the observed sensitivity and specificity for different models using genetic scores and seizure age at onset: SCN1A score and Onset (blue line, n = 743), onset only (orange line, n = 743), CADD (Combined Annotation Dependent Depletion) score and Onset (green line), and REVEL (Rare Exome Variant Ensemble Learner) score and Onset (purple line). Because CADD and REVEL scores are not available for all variants contained in the training cohort, the CADD and Onset and REVEL and Onset models were built with a subset of 651 and 438 training cohort patients, respectively (eTable 1, links.lww.com/WNL/B785). The 6-months seizure onset threshold model (gray line) proposed previously is shown for comparison. Area under the curve (AUC) values and 95% CIs are shown at the bottom right corner of the plot.

Figure 4. Calibration Curves per Model

Training cohort model performance. Individual calibration curves showing the relationship between the predicted risk and the observed frequency for each of the tested models. Index of prediction accuracy (IPA) is shown below each model. Color code: SCN1A score and Onset (blue line), Onset-only (orange line), CADD (Combined Annotation Dependent Depletion) score and Onset (green line), and REVEL (Rare Exome Variant Ensemble Learner) score and Onset (purple line).

Figure 5. Validation Cohort 1 and 2 Prediction Results

Patients with probability values above 50% were predicted to have Dravet syndrome and patients with values below 50% were predicted to have genetic epilepsy with febrile seizures plus (GEFS+). (A, B) Predicted values across validation cohorts 1 and 2 are shown, respectively. Each bar corresponds to a patient. The height of each bar represents the probability of that patient developing Dravet syndrome. Patients with true Dravet syndrome are shown in purple; patients with true GEFS+ are shown in gray. Dotted horizontal line denotes a 50% threshold with values above 50% predicting Dravet syndrome and values below 50% predicting GEFS+. Area under the curve (AUC) and index of prediction accuracy (IPA) 95% CIs are given.

Figure 6. Validation Cohort 1 and 2 Phenotype Distribution

Phenotype distribution with density of prediction performed on validation cohorts 1 and 2, respectively. Patients with true Dravet syndrome and patients with genetic epilepsy with febrile seizures plus (GEFS+) accumulate across their corresponding model predictions (horizontal axis). Dotted vertical line denotes a 50% threshold with values above 50% predicting Dravet syndrome and values below 50% predicting GEFS+.