AB001. The path to genomic medicine

Abstract: There are numerous obstacles to genomic medicine. These include the large number of rare and novel genomic variants per individual. The American College of Medical Genetics and Genomics (ACMG) has recommended that all pathogenic variants in 56 gene-disease pairs that are identified incidentally in a genomic test be offered to the patient (Green et al., 2013, PMID: 23788249). We considered an expanded list of 112 actionable gene-disease pairs, ones where medical intervention is possible to prevent or detect disease early. We estimate the rate of these incidental findings (IFs) in European and African Ancestry groups. However, we found high discordance between classifications of expert reviewers. We have reported both inconsistency across labs in variant classification and a bias towards overcalling pathogenicity (Amendola et al., 2015, PMID: 25637381). Thus, there is a need to standardize classification of genomic variants in medical sequencing. To date genomics laboratories have used non-standard classification systems. The ACMG published guidelines for variant classification for Mendelian disorders designed to increase consistency among labs (Richards et al., 2015, PMID: 25741868). The Clinical Sequencing Exploratory Research (CSER) Consortium evaluated the use of these rules by nine of the CLIA laboratories supporting CSER projects, considering 99 germline variants. The results were examined to evaluate intra-laboratory differences between variant classifications using the labs own criteria vs. adopting ACMG criteria and inter-laboratory differences using either the lab’s own system or the ACMG guidelines. Agreement among labs did not differ whether using the laboratory specific vs. ACMG criteria (P=0.9); i.e., the ACMG criteria did not yield more consistent variant classification in this exercise. We further analyzed sources of disagreement in the use of the ACMG criteria and identified causes of variance in classifications. In addition to providing useful analyses of how variant classifications approaches vary among laboratories, these data should allow clarification and refinement of the ACMG criteria that may increase consistency in variant classification.

Keywords: Genomic medicine; genomic variants; variant classifications