Multidisciplinary pathogenicity assessment of congenital hypothyroidism-associated SLC5A5 gene variants.

Abstract

Next-generation sequencing (NGS) has markedly expanded the mutational spectrum of monogenic congenital hypothyroidism (CH), but has also led to the identification of numerous variants of uncertain significance, limiting the clinical utility of genetic testing. We previously developed an SLC5A5-specific machine learning-based predictor that improved the interpretation of missense variants associated with CH. Here, we systematically evaluated the pathogenicity of all CH-associated SLC5A5 variants reported in the literature using an integrated multidisciplinary framework based on the American College of Medical Genetics and Genomics (ACMG) guidelines. Among the 63 identified SLC5A5 variants, 5% were classified as benign, 16% as likely benign, 15% as of uncertain significance, 37% as likely pathogenic, and 27% as pathogenic. The computational variant effect predictor AlphaMissense was the most appropriate tool for predicting the functional impact of SLC5A5 missense variants. Experimental functional assays assessing the functional impact of SLC5A5 missense variants on radioiodide accumulation were instrumental in reclassifying variants of uncertain significance into clinically actionable categories. Moreover, patients' phenotype revealing reduced to absent thyroid radioiodide accumulation provided valuable phenotypic evidence supporting variant classification, which highlights the diagnostic value of thyroid scintigraphy in CH-associated SLC5A5 variants. Overall, our findings highlight the value of a multidisciplinary pathogenicity evaluation based on the ACMG guidelines, integrating computational predictions, functional assays, and phenotypic data into a standardized and evidence-based framework for the accurate interpretation of SLC5A5 variants, thereby representing a significant advance in the molecular diagnosis of congenital hypothyroidism.