Abstracts

Rationale: Copy number variants (CNVs) are well recognized as an important cause of disease. In epilepsy, CNVs have been implicated in several well-defined cohorts including epileptic encephalopathies and genetic generalized epilepsy with intellectual disability (ID). Little is known about the role of CNVs in adults with epilepsy and ID. This study aims to evaluate the prevalence of pathogenic CNVs and identify possible candidate genes in adults with epilepsy and ID of unclear etiology. Methods: Patients were recruited prospectively from Toronto Western Hospital epilepsy clinic from 2012-2014. Inclusion criteria were (i) epilepsy of childhood or adolescent onset and ongoing seizures in adulthood; (ii) ID of any degree and (iii) complete medical record. Patients with structural brain abnormalities or evidence of metabolic etiology were excluded. Clinical array-CGH was performed following standard protocols using CytoScan HD SNP Array (Affymetrix). CNVs were classified as benign, variants of unknown significance (VUS) or pathogenic and/or likely pathogenic based on their size, gene content and presence in control databases as per ACMG guidelines. CNVs classed as VUS were further analyzed and reclassified as likely pathogenic if the CNV contained candidate genes that were (i) expressed in the central nervous system; (ii) phenotypically relevant to the patient and (iii) predicted to be intolerant to variation as indicated by a high Residual Variation Intolerance Score (RVIS). Results: 149 patients were included and comprised 142 unrelated subjects, 1 pair of identical twins and 2 families with siblings of similar clinical presentation. Pathogenic or likely pathogenic CNVs were identified in 26/149 (17.4%) patients. 4 patients had 2 independent pathogenic CNVs and 2 patients had a pathogenic CNV and a second CNV classed as a VUS. Affected siblings presented with the same CNVs. De novo CNVs were present in 12/26 patients (46.1%) and 7/26 (26.9%) had inherited CNVs from unbalanced segregation of balanced parental rearrangements or maternal X-linked CNVs. 13 unrelated patients (14/26, 54%) presented with known genetic disorders including 1p36 deletion syndrome, Angelman syndrome, velocardiofacial syndrome, epilepsy and mental retardation limited to females (EFMR), MECP2 duplication syndrome and 16p13.11 hotspot deletion. 10 novel CNVs were found in 12 patients: (i) 16p13.2 duplication (dup) in Lennox-Gastaut syndrome with candidate gene ABAT; (ii) 8p23.3-p23.2 deletion (del) in epilepsy aphasia spectrum; (iii) 8q21 dup and (iv) Xq13 del in patients with GGE and ID and (v) 2p16.1-p15 dup, (vi) 6p25.3-p25.1 dup, (vii) 9p24.3-p23 del, (viii) 10q11.22 dup, (ix) 12p13.33-13.2 dup and (x) 13q34 del were identified in patients with focal epilepsy and ID. Conclusions: The high yield (17.4%) of pathogenic or likely pathogenic CNVs in patients with epilepsy and ID of undefined etiology reinforces the role of CNVs in this patient group and the need for routine SNP array studies in the clinic. CNVs can be used to identify novel candidate epilepsy genes although identification of additional cases with mutations in these genes and functional analyses are required to confirm their role in epilepsy. Funding: Ontario Brain Institute