Abstracts

Rationale: Genetic testing in patients with DEEs is necessary to identify the genetic basis for a patient’s symptoms. Multigene panels offer the potential for increased diagnosis of DEEs particularly with broadening of phenotypic spectrums. For academic researchers and drug developers trying to identify new therapies, the size of a potential DEE diagnosis is crucial but prevalence estimates for many DEEs are not available. Our objective was to use claims and genetic testing data to model genetic testing trends and develop estimates for US prevalence for DEEs associated with precision medicine treatment approaches.

Methods: Analysis with claims (Jan 2017 - Nov 2021) and genetic test data (Jan 2020 - Dec 2021) was performed for patients under age 21 years with a history of seizures and/or developmental delay (“Potential DEE Patients”). Analogs were developed for 5 well-characterized pediatric DEE/neurodevelopmental conditions: Dravet, CDKL5 Deficiency, Angelman, Rett, and TSC. Projection factors were developed lusing both data sets to estimate prevalence in the US for 17 genes linked to DEEs without an ICD-10. Adjustments were made for data capture, anticipated reclassification, and age of onset.

Results: Claims data captured 348k genetic tests for Potential DEE Patients (2017 - 2021). An estimated 20k, not captured in claims, received a sponsored multigene panel in 2020 or 2021. Genetic testing for Potential DEE Patients was flat 2017 - 2019, then declined 25% in 2020 and 35% in 2021 v. 2019, likely due to the pandemic (Figure 1). Age distributions remained similar over the 5 years (median age of 5 years). Sponsored testing panels grew to 13% of genetic tests by 2021, but these tests mainly cannibalized those previously covered by insurance._x000D_
Based on claims, nearly 50% of genetic tests ordered in 2021 for Potential DEE Patients were by Neurologists or Geneticists. Among these, 476 providers accounted for 75% of tests. Sponsored multigene panels not captured in claims show a similar concentration among providers. A total of 22.4K patients had claims for an analog condition in 2020 or 2021. Genetic test data included 3.2k patients with a variant for an analog or a DEE; 29% of these were pathogenic/likely pathogenic (P/LP) and the remainder were classified as VUS. Projected prevalence for the 17 DEEs selected ranged from 436 for FOXG1 to 5,783 for SLC2A1 (Figure 2). 4 of the DEEs analyzed had a projected prevalence of 4k or more patients. Since diagnosis rates are likely low (< 20%) for DEEs not linked to a well-characterized clinical diagnosis, the overall prevalence of each of these conditions could be 20k patients or more.