Abstract 3289: Implementing evolving clinical standards in a variant interpretation knowledgebase

Abstract

Abstract The rapid expansion of clinical sequencing and targeted therapies has driven the accelerated evolution of clinical guidelines and regulatory standards surrounding sequence variants. Organizations such as the NCCN, FDA, ACMG, etc., continue to incorporate sequence variants into a variety of clinical contexts. The pace of these changes has led to several disparities between guidelines and resources used for the interpretation of clinically identified variants. 1) Despite the inherent lag in creating such standards through expert consensus, this evolution outpaces their incorporation into many resources used for clinical sequencing analysis. 2) Guidelines are often not readily accessible in programmatic, structured ways to promote rapid updates, including many guidelines only available in human-readable text formats. 3) These guidelines apply to various clinical contexts (drug response, prognosis, etc.), each requiring their own schema changes for their proper incorporation. 4) Levels of specificity provided by these guidelines are not conducive to matching specific genomic coordinates (e.g., FGFR2 fusion) or maintaining evidence provenance. We recently developed the Clinical Interpretations for Variants in Cancer (CIViC; civicdb.org) crowd-sourced, open-access knowledgebase, providing a structured framework for evaluating genomic variants of various types (e.g., fusions, SNVs) for their therapeutic, prognostic, predisposing or diagnostic utility. Since then, the CIViC resource has continued to evolve and has tackled the challenges above. This crowd-sourced model of both data entry and application development leverages a wide network of users with various expertise to adapt the resource at a pace beyond what a single organization could achieve. We have introduced a new entity to CIViC that aggregates individual evidence items into a single clinical assertion, allowing guideline incorporation at varying levels of specificity. For example, classification of germline variants by ACMG guidelines assigns individual evidence to specific codes (e.g., PVS1, PP1) that are aggregated to classify a variant as benign or pathogenic. The CIViC assertion provides the variant classification and supporting evidence that can be readily re-evaluated by any user, not just the initial submitter, and reviews and tracks any changes. An assertion can be created for highly specific variants (e.g., EGFR T790M), which can be submitted to ClinVar or more generic variants (ALK fusions) unsupported by many resources. Assertions support searchable, structured, and versioned annotations including FDA companion tests and drug approvals, germline ACMG categories, somatic AMP/ASCO variant levels, NCCN and WHO guidelines. The CIViC knowledgebase provides rapid integration of the latest guidelines and regulatory standards in a freely accessible resource that is flexible enough to evolve with this rapidly changing field. Citation Format: Kilannin Krysiak, Arpad M. Danos, Alex H. Wagner, Susanna Kiwala, Joshua F. McMichael, Adam C. Coffman, Erica K. Barnell, Yan-Yang Feng, Benjamin J. Ainscough, Cody A. Ramirez, Malachi Griffith, Obi L. Griffith. Implementing evolving clinical standards in a variant interpretation knowledgebase [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3289.