Annual Meeting Abstracts: View

  • (Abst. 3.329), 2016
  • An informatics infrastructure for KCNQ2 Encephalopathy research including a patient registry, database, curation platform, and website
  • Authors: Nishtha Joshi, Baylor College of Medicine, Houston, Texas; Maurizio Taglialatela, University of Molise, Campobasso, Italy; Sarah Weckhuysen, VIB, Antwerp, Belgium; Gerry Nesbitt, CLIRINX Research Informatics, Dublin, Ireland; and Edward Cooper, Baylor College of Medicine, Houston
  • Content:

    Rationale: KCNQ2/3 variants lead to a spectrum of early onset epilepsies that includes self-limiting forms transmitted in an autosomal dominant pattern (BFNE), and KCNQ2/3 epileptic encephalopathy (EE), which includes a range of phenotypes with more persistent seizures and developmental impairment. Factors affecting outcome are incompletely understood, as is the natural history at all points along the severity spectrum. Objectives of KCNQ2/3 variant curation include improving understanding of genotype-phenotype relationships, enabling clinical research needed for developing better outcome measures, and identifying subgroups that may require different therapeutic approaches. Methods: Under an approved IRB protocol, a custom informatics system has been established using Clirinx software incorporating a patient registry, curation platform and a website. Patients diagnosed with KCNQ2/3 variants enter the database via four pathways: (A) publication by others; (B) family self-registration; (C) physician referral; (D) clinical genetics lab disclosure (e.g., ClinVar). Data collected include clinical and family history, test results, and therapeutic responses, as available. Each entry is reviewed by a multi-institutional, multidisciplinary curation panel. Curation includes standardized abstraction of key clinical and laboratory data that ends in a "variant summary" used for classification. Variant scoring criteria for pathogenicity and severity assessment are based on American College of Medical Genetics and Genomics guidelines (Richards, Genet Med. 2015;17:405-24), but are customized based on gene-specific knowledge. A "point and click" online pathogenicity/severity calculator has been developed, in order to record scores and optimize reliability of variant category assignment (e.g., non-pathogenic, likely non-pathogenic, uncertain significance, likely pathogenic/self-limiting, likely pathogenic/epileptic encephalopathy, pathogenic/self-limiting, or pathogenic/epileptic encephalopathy). Results, including parts of the evidence summary and the pathogenicity and severity classification, are reported on a locus specific website (www.RIKEE.org) and NCBI's www.ClinVar.com. Results: The patient census has grown rapidly. Patients and pedigrees total 362 as of June 2016. After curation, 77 BFNE variants (88 pedigrees) and 50 EE variants (114 individuals) have been classified as likely pathogenic or pathogenic. Some previously published variants have been classified as non-pathogenic, likely non-pathogenic, or uncertain pathogenicity. Approximately 150 more patients or pedigrees have recently been entered and are in curation. Nearly all these remaining patients have clinical EE and de novo KCNQ2 variants. Correlations have been made between phenotype and variant characteristics, including type (e.g., missense vs. stop-gained), location in the quaternary structure, and in vitro functional effects. A growing percentage of reported cases are recurrences previously seen in an unrelated KCNQ2 encephalopathy patient. Conclusions: We have developed a comprehensive KCNQ2/3 patient registry, database, curation platform, and public website. Systematic classification may prove useful clinically, making individual diagnosis more rapid and secure, and making relationships between genotype, pathophysiology, and phenotype clearer. This may help guide development of targeted treatments for patients with KCNQ2 encephalopathy. Our work provides a model for addressing challenges in investigation of rare disease through broad multidisciplinary collaboration. Funding: We gratefully acknowledge: NINDS NS49119; AES EF (Infrastructure Grant); The Jack Pribaz Foundation, KCNQ2 Cure Alliance