Abstracts

Rationale:
Over the past several years genetic testing for pediatric epilepsy has become increasingly available, due in part to low-cost self-pay and industry-funded testing options. It is unknown how improved accessibility has impacted ordering practices, testing yields, and incidence of variants of uncertain significance (VUS).
Method:
A retrospective chart review was performed at Children’s Hospital Colorado from February 2016 through February 2020. Patients were included if they had an epilepsy gene panel sent for seizures with onset between 0 and 18 years of age. Data were collected on demographics, payer sources, time of seizure onset, seizure types, MRI results, developmental delay, and epilepsy panel results. MRI results were classified as normal/acquired abnormality, developmental abnormality, or other/unclear. Developmental delay included any degree of delay, learning disabilities, and/or autism. Disease-causing results were classified as results with pathogenic or likely pathogenic variants (one for autosomal dominant or X-linked conditions, two for autosomal recessive conditions). Analyses included descriptive statistics with percentages, means, and medians. Comparisons between groups were done via odds ratios and t-tests.
Results:
A total of 761 epilepsy gene panels were sent over the study period. The average number of panels sent per month increased by 292% from 2016 to 2020. This corresponded to an increase in industry-funded testing, which accounted for 68% of panels sent in 2020. The time from seizure onset to panel result decreased steadily over the study period from a median of 2.9 years in 2016 to 0.7 years in 2020. Despite the increase in testing, the percentage of panels yielding a disease-causing result remained stable each year at 11-13%, resulting in a greater absolute number of positive diagnoses per annum. A total of 90 disease-causing results were identified, 75% of which provided guidance in management (either through precision medicine implications or medical management guidelines). Children were more likely to have a disease-causing result if they were < 3 years old at seizure onset (OR 4.4, p < .001), had developmental delay (OR 2.2, p = .002), or had a developmentally abnormal MRI (OR 3.8, p < .001). Forty-seven patients (6%) were found to be disease carriers. A total of 1417 VUS were identified, equating to 15.7 VUS per disease-causing result. The average number of VUS per panel remained stable over time at 1.8. Non-Hispanic white patients had a lower average number of VUS than patients of all other races/ethnicities (1.7 vs 2.1, p < .001) with all others having 1.4 times the likelihood of having 2 or more VUS than non-Hispanic white patients (p = .029).
Conclusion:
A rapid expansion in the volume of genetic testing has occurred over the last four years at our institution, with a decrease in the time from seizure onset to testing result. This has resulted in an increase in the total number of disease-causing results, most of which have implications for management. There has also been an increase in total VUS with higher rates in minority populations, highlighting the need for comprehensive support in VUS resolution.
Funding:
:None