Identification of Pathological FBN1 Variants Is Not Straightforward

he ability to identify DNA variants that are pathogenic forms the basis for unbiased characterization of inherited diseases.Improved access to genetic testing and the increasing clinical use of whole exome and whole genome sequencing have identified variants of uncertain pathogenicity often associated with an indefinite disease phenotype.In 2015, the American College of Medical Genetics and Genomics and the Association for Molecular Biology published guidelines to assist in pathogenicity classification of variants in Mendelian diseases.1 However, the application of the 28 criteria is complex and can still be subjective.Difficulties in determining the pathogenicity of sequenced FBN1 (fibrillin-1) variants associated with the Marfan syndrome is a good example of the general problem.In this issue of Circulation: Genomic and Precision Medicine, Muiño-Mosquera et al 2 propose a modification of the American College of Medical Genetics and Genomics and the Association for Molecular Biology guidelines for ascertainment of the significance of FBN1 mutations in patients with clinical signs of Marfan syndrome using nonsubjective criteria and show that their modified criteria result in changes in diagnostic determinations of some patients thought to have a clinical diagnosis of Marfan syndrome.The American College of Medical Genetics and Genomics and the Association for Molecular Biology criteria use population data, the correlation between genotype and phenotype, literature and locus-specific databases, computational and predictive data, functional data, segregation data, and allelic data.The modified guidelines developed by a working party for classification of FBN1 variants refined 15 of the 28 criteria.These included lowering the threshold allele frequency indicative of a benign variant because of the rarity of Marfan syndrome, defining the phenotype more stringently and downgrading a mutation if there was a likely pathogenic mutation in another gene associated with a related phenotype, assessing only publications with strong clinical or functional evidence, adjustment of the criteria for computational evidence, assessing functional data based on the known role of specific amino acids and domains in the function of the protein and quantifying the impact of segregation in multiple family members.Appropriate resources and specifications for FBN1 classification are provided for each category.The authors then compared classification status between previous methods and these adapted American College of Medical Genetics and Genomics and the Association for Molecular Biology criteria using 715 FBN1 in-house consecutively identified variants from their MFS database spanning ≈25 years.There was an overall agreement of 86% between the methods.The number of variants classified as pathogenic was reduced, leading to the number of variants of uncertain significance (VUS) rising from 14.5% to 24.2%.In addition, 14 variants (2%) were downgraded to likely benign, and one was upgraded from VUS to likely EDITORIAL Identification of Pathological FBN1 Variants Is Not Straightforward