An updated quantitative model to classify missense variants in the TP53 gene: a novel multifactorial strategy

Multigene panel testing has led to an increase in the number of variants of uncertain significance identified in the TP53 gene, associated with Li-Fraumeni syndrome. We previously developed a quantitative model for predicting pathogenicity of P53 missense variants based on the combination of calibrated bioinformatic information and somatic to germline ratio. Here, we extended this quantitative model for the classification of P53 predicted missense variants by adding new pieces of evidence (personal and family history parameters, loss-of-function results, population allele frequency, healthy individual status by age 60, and breast tumor pathology). We also annotated which missense variants might have an effect on splicing based on bioinformatic predictions. This updated model plus annotation led to the classification of 805 variants into a clinically relevant class, which correlated well with existing ClinVar classifications, and resolved a large number of conflicting and uncertain classifications. We propose this model as a reliable approach to TP53 germline variant classification and emphasize its use in contributing to optimise TP53-specific ACMG/AMP guidelines. This article is protected by copyright. All rights reserved.