Abstract 12381: Clinical Implications of SCN5A and SCN10A Loss of Function Variants in 169,610 Exomes Representing the General Population

Abstract

Background: Genetic variation in the genes SCN5A and SCN10A has previously been associated with cardiac disease and conduction velocity through genome-wide association studies (GWAS) and rare variant analyses. The two neighboring genes constitute the much-investigated SCN5A-SCN10A locus, and it has been challenging to determine the region’s function as GWAS often identifies variants in non-coding regions. The cardiac function of SCN5A is well established whereas the role of SCN10A in cardiac conduction and disease remains disputed. The study focuses on SCN10A while SCN5A was included for comparative analyses. Methods and Results: We accessed the UK Biobank database, which contains data on >500,000 individuals. The recent exome release provides whole-exome sequencing data on 169,610 unrelated European individuals while resting ECGs are available on 44,987 individuals. We identified 668 individuals harboring LOF-variants (leading to early stop codon, frameshift or splice site disruption) in SCN10A and SCN5A with a MAF <1%. Among the 668 individuals, 543 individuals carried SCN10A variants, whereas 125 individuals held a variant in SCN5A . There were 88 and 37 different variants in SCN10A and SCN5A , respectively. From these individuals, we obtained 69 ECGs in sinus rhythm and 55 of these were recorded on individuals with a SCN10A LOF-variant and 14 ECGs were recorded on individuals with a SCN5A LOF-variant. Linear regression analyses established an association between SCN5A LOF-variants and prolongation of the PR-interval as carrying a variant corresponds to an average PR prolongation of 25.1 milliseconds (P = 0.001). There were no statistically significant alterations among other ECG parameters. Furthermore, we could not determine correlations of ECG parameters for SCN10A LOF-variants. A burden test on SCN10A and SCN5A LOF-variants could not establish a relationship with atrial fibrillation (P = 0.26 and P = 0.41, respectively) and heart failure (P = 0.92 and P = 0.17, respectively). Remaining possible disease associations could not be properly established due to low number of cases. Conclusion: These findings confirm the significance of SCN5A in cardiac conduction while questioning the function of SCN10A in human cardiac conduction and disease.