Abstracts

Rationale: Epileptic encephalopathy (EE) is one of the most severe forms of infantile/childhood epilepsy and is often associated with cognitive, sensory, or/and motor dysfunction. EE is a genetically heterogeneous condition and discovery of the underlying genetic causes have been increasingly recognized with next generation sequencing (NGS). The clinical phenotypes of EE have a wide clinical spectrum and overlapping features, which make selection of one gene for molecular evaluation difficult. Targeted NGS has shown diagnostic yields ranging from 10 to 48.5%. We therefore designed a clinically available NGS panel of 25 targeted genes for screening patients with EE.Methods: Pilot testing of a new NGS-based EE panel that includes 25 genes. After Covaris fragmentation, Agilent probes are used to capture 25 genes (360 kb of DNA). Sequencing is performed on a HiSeq 2500 to obtain 300-1000X average coverage of these genes. At this depth we are reliably able to call low level mosaicism and deletions/duplications. This gene panel covers 17 autosomal dominant or sporadic/de novo genes (CHD2, FOXG1, GABRA1, GABRG2, KCNQ2, KCNQ3, KCNT1, MEF2C, PTEN, SCN1A, SCN2A, SCN8A, SLC2A1, STXBP1, SYNGAP1, TCF4, UBE3A); 3 autosomal recessive genes (ALDH7A1, PNPO, PNKP); and 5 X-linked genes (SLC9A6, MECP2, CDKL5, ARX, PCDH19). Four patients with a diagnosis of EE were tested. Electronic charts were reviewed for clinical features, diagnostic workup, and molecular genetic testing. Outcomes of interest include the diagnostic yield and the rate of variants of uncertain significance (VUS).Results: Two patients had pathogenic or likely pathogenic gene mutations. One patient had VUS, and one patient had a negative result. All patients had negative metabolic and SNP testing. Patient #1 had seizure onset at 3-year old, global delay, and movement disorder who had a negative test. Patient #2 had seizure onset at 2-day old and mild motor delay who had a heterozygous STXBP1 mutation at c.1652G>A (p.R551H). This mutation has not been previously reported; however a patient with a mutation of the same residue to a different amino acid (p.R551C) has been reported in a patient with EE, supporting our interpretation of this variant as likely pathogenic. Patient #3 had seizure onset at 3-month old, dysmorphic features, and global delay who had pathogenic truncating mutation in CDKL5 (c.1076delG, p.G359Vfs*9). Patient #4 had seizure onset at 6-month old and autism who had heterozygous in frame deletion with SYNGAP1 gene (p.Lys337del). In frame deletions have not been reported as pathogenic in this gene, however the variant is not present in any of the large population databases (dbSNP, EVS, or EXAC). Parental testing is pending, so this variant remains of VUS.Conclusions: EPIPX appears to be an effective tool for molecular diagnosis of EE, with a diagnostic yield of 2 out of 4 patients in our pilot study. We plan to expand our current EPIPX panel to include more genes going forward.