Authors :
Olubunmi Fariyike, BS – Stanford University School of Medicine
Minye Zhou, MSc – Harvard T. H. Chan School of Public Health, Harvard University
Yingmei Li, PhD – Stanford University School of Medicine
Nishanth Narayan, BS – Stanford University School of Medicine
Somadinauwa Ezekwu, BSN, RN – Stanford University School of Medicine
Danielle Sanchez, BS – Stanford University School of Medicine
August Huang, PhD – Harvard Medical School
Presenting Author: H. Westley Phillips, MD – Stanford University School of Medicine
Rationale:
Brain-limited, somatic variants, especially those in genes involving the mTOR pathway, have been shown to drive the histopathologic and electrophysiologic abnormalities seen in pediatric drug-resistant epilepsy (pDRE) secondary to malformations of cortical development (MCDs). Recent studies have demonstrated that these mutations exhibit gradient patterns in the brains of MCD patients, with the nidus of the seizure-onset zone containing a greater proportion of variant-positive cells than other secondary regions of seizure-onset and those regions containing a greater proportion of variant-positive cells still than uninvolved regions. Although this phenomenon has been seen clearly in MCD-associated pDRE, the presence of regional somatic mosaicism has not yet been characterized in multiple pDRE patients with diverse etiologies. Therefore, this study sought to identify the role of mosaic variants in different etiologies of epileptic disease.
Methods:
Multiple resection-derived brain tissue samples from 27 cases of pDRE (range 2-7 samples per case) were subjected to deep targeted DNA sequencing with a panel of 283 epilepsy- and malignancy-associated genes. When available, matched blood was also sequenced to exclude likely germline variants. Average sequencing depth across all samples after consensus read aggregation was 250X. For each variant called by targeted panel sequencing analysis, targeted amplicon sequencing at a minimum depth of 10,000 reads was performed to validate the variant. For variants that were validated, targeted amplicon sequencing for the validated variant was then performed in all other tissue samples of the same case. The primary SOZ was defined as the resected epileptic region(s) with the most prominent epileptic activity per corresponding electrographic data. The secondary SOZ was defined as the region(s) the adjacent regions that were still epileptic and indicated for resection but were not the primary focus of seizure onset.
Results:
Of the 27 cases sequenced, 13 had a variant called by targeted panel sequencing and validated by targeted amplicon sequencing. Of these 13, 11 (84.6%) had a primary MCD (including two patients with tuberous sclerosis), and two (15.4%) low-grade glioma. Seven cases (53.8%) exhibited regional somatic mosaicism. Notably, for the four cases with clear primary and secondary SOZs, there was a considerable difference in average VAFs between primary and secondary SOZs (average VAF 14.5% vs. 2.7%, p=0.016). The average difference between the highest and lowest VAFs in a sample was 17.6%.
Conclusions:
To date, somatic mosaicism has only been suggested to drive epileptic networks in MCDs. This study reveals that pDRE patients of multiple etiologies possess regional mosaicism consistent with this theory, while other included pDRE patients may either not exhibit regional mosaicism, exhibit regional mosaicism in uninterrogated genomic regions, or have variants with VAFs too low to be detected by current methods.
Funding: There are no sources of funding to report.