Validation of the ACMG/AMP guidelines-based seven-category variant classification system

Abstract

BackgroundOne shortcoming of employing the American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP)-recommended five-category variant classification scheme ("pathogenic", "likely pathogenic", "uncertain significance", "likely benign" and "benign") in medical genetics lies in the schemes inherent inability to deal properly with variants that fall midway between "pathogenic" and "benign". Employing chronic pancreatitis as a disease model, and focusing on the four most studied chronic pancreatitis-related genes, we recently expanded the five-category ACMG/AMP scheme into a seven-category variant classification system. With the addition of two new classificatory categories, "predisposing" and "likely predisposing", our seven-category system promises to provide improved classification for the entire spectrum of variants in any disease-causing gene. The applicability and practical utility of our seven-category variant classification system however remains to be demonstrated in other disease/gene contexts, and this has been the aim of the current analysis. ResultsWe have sought to demonstrate the potential universality of pathological variants that could be ascribed the new variant terminology ( predisposing) by trialing it across three Mendelian disease contexts (i.e., autosomal dominant, autosomal recessive and X-linked). To this end, we firstly employed illustrative genes/variants characteristic of these three contexts. On the basis of our own knowledge and expertise, we identified a series of variants that fitted well with our "predisposing" category, including "hypomorphic" variants in the PKD1 gene and "variants of varying clinical consequence" in the CFTR gene. These examples, followed by reasonable extrapolations, enabled us to infer the widespread occurrence of "predisposing" variants in disease-causing genes. Such "predisposing" variants are likely to contribute significantly to the complexity of human genetic disease and may account not only for a considerable proportion of the unexplained cases of monogenic and oligogenic disease but also for much of the "missing heritability" characteristic of complex disease. ConclusionEmploying an evidence-based approach together with reasonable extrapolations, we demonstrate both the applicability and utility of our seven-category variant classification system for disease-causing genes. The recognition of the new "predisposing" category not only has immediate implications for variant detection and interpretation but should also have important consequences for reproductive genetic counseling.