Phenotypic And Genetic Profile Of Scn5a Variants In Idiopathic Dilated Cardiomyopathy

Abstract

Background Genetic variants in SCN5A, a gene encoding α-subunit of the cardiac sodium channel, cause a broad spectrum of cardiovascular phenotypes without clearly established genotype-phenotype correlations. As classified by ClinGen, SCN5A has a definitive relationship with dilated cardiomyopathy (DCM). Objective We sought to define the genetic and phenotypic presentation of SCN5A variants in DCM probands enrolled in the DCM Precision Medicine Study. Methods Medical records and exome sequence data were collected from DCM probands. Variants meeting benign or likely benign criteria (those with an allele frequency >0.1% in all non-founder populations) were filtered out. Remaining variants in 35 DCM-associated genes were analyzed using variant interpretation guidelines refined for DCM. Results Of 1133 probands, 47 (4%) subject of diverse backgrounds (47% non-Hispanic African American; 47% non-Hispanic Caucasian; 6% Hispanic) carried 40 SCN5A variants classified as variants of uncertain significance (39) or likely pathogenic (LP) (1; p.Arg222Gln) for DCM. The median [IQR (interquartile range)] age at DCM diagnosis was 43.0 [37; 54] years. The LV end-diastolic dimension by echo at diagnosis was 65.4 [61.0; 73.0] mm and ejection fraction was 20.0 [15.0; 25.5] %. Over half (51%) had conduction disturbance and/or arrhythmia phenotypes by twelve-lead ECG, including intraventricular conduction disturbance (n=14; 29.8%), left bundle branch block (8; 17.0%), right bundle branch block (2; 4.3%), atrial fibrillation (2; 4.3%), atrioventricular block (2; 4.3%), and/or prolonged QTc intervals (3; 6.4%). Unique variant combinations were also observed in the case of dual SCN5A variants (4) and additional LP/P variants (6), including truncating variants BAG3 (1) and TTN A-band (5). Nearly all variants (39) were missense, with one inframe deletion identified (p.Glu1072del). Only 4 variants were previously reported in DCM cases (p.Thr220Ile, p.Arg222Gln, p.Ala1102Thr, and p.Gln1832Glu), with others only reported in primary arrhythmic syndromes (24), and or never reported (12). Half of variants were rare (alternate allele frequency <0.05%); the other half were between 0.05% and 0.10% population frequency. Conclusions Variants potentially relevant for DCM in SCN5A were identified in 4% of DCM probands. Over half of variants had population frequencies higher than expected for monogenic disease (between 0.05% and 0.10%), yet a high prevalence of conduction disturbances was also observed. Our data suggest that while some SCN5A variants act under a monogenic mechanism in DCM, other SCN5A variants may also play a role as disease modifiers. Genetic variants in SCN5A, a gene encoding α-subunit of the cardiac sodium channel, cause a broad spectrum of cardiovascular phenotypes without clearly established genotype-phenotype correlations. As classified by ClinGen, SCN5A has a definitive relationship with dilated cardiomyopathy (DCM). We sought to define the genetic and phenotypic presentation of SCN5A variants in DCM probands enrolled in the DCM Precision Medicine Study. Medical records and exome sequence data were collected from DCM probands. Variants meeting benign or likely benign criteria (those with an allele frequency >0.1% in all non-founder populations) were filtered out. Remaining variants in 35 DCM-associated genes were analyzed using variant interpretation guidelines refined for DCM. Of 1133 probands, 47 (4%) subject of diverse backgrounds (47% non-Hispanic African American; 47% non-Hispanic Caucasian; 6% Hispanic) carried 40 SCN5A variants classified as variants of uncertain significance (39) or likely pathogenic (LP) (1; p.Arg222Gln) for DCM. The median [IQR (interquartile range)] age at DCM diagnosis was 43.0 [37; 54] years. The LV end-diastolic dimension by echo at diagnosis was 65.4 [61.0; 73.0] mm and ejection fraction was 20.0 [15.0; 25.5] %. Over half (51%) had conduction disturbance and/or arrhythmia phenotypes by twelve-lead ECG, including intraventricular conduction disturbance (n=14; 29.8%), left bundle branch block (8; 17.0%), right bundle branch block (2; 4.3%), atrial fibrillation (2; 4.3%), atrioventricular block (2; 4.3%), and/or prolonged QTc intervals (3; 6.4%). Unique variant combinations were also observed in the case of dual SCN5A variants (4) and additional LP/P variants (6), including truncating variants BAG3 (1) and TTN A-band (5). Nearly all variants (39) were missense, with one inframe deletion identified (p.Glu1072del). Only 4 variants were previously reported in DCM cases (p.Thr220Ile, p.Arg222Gln, p.Ala1102Thr, and p.Gln1832Glu), with others only reported in primary arrhythmic syndromes (24), and or never reported (12). Half of variants were rare (alternate allele frequency <0.05%); the other half were between 0.05% and 0.10% population frequency. Variants potentially relevant for DCM in SCN5A were identified in 4% of DCM probands. Over half of variants had population frequencies higher than expected for monogenic disease (between 0.05% and 0.10%), yet a high prevalence of conduction disturbances was also observed. Our data suggest that while some SCN5A variants act under a monogenic mechanism in DCM, other SCN5A variants may also play a role as disease modifiers.