Shared and distinguishing phenotypic characteristics of filamin c and desmoplakin cardiomyopathy

Abstract

Abstract Background Rare truncating variants in the genes Filamin C (FLNC) and Desmoplakin (DSP) are increasingly recognised causes of Dilated Cardiomyopathy and Arrhythmogenic Cardiomyopathy (1,2). A direct comparison of both genotypes is needed to aid in subclassification and genotype-specific risk modelling. Purpose We sought to identify shared and distinguishing clinical characteristics and imaging phenotypes between FLNC and DSP cardiomyopathy. Methods 41 patients with ‘pathogenic’ or ‘likely pathogenic’ FLNC or DSP variants by American College of Medical Genetics (ACMG) criteria were identified from a tertiary centre cardiovascular genetics laboratory. Clinical, genetic and imaging data were compiled. Continuous variables are presented as median [IQR] or mean±SD and analysed by nonparametric Mann Whitney U or t-test. Categorical variables were analysed by Chi-Squared or Fisher’s exact test. Results FLNC and DSP cohorts included a similar proportion of patients with truncating variants (91% vs 100%) and female sex (12/21, 57% vs 13/20, 65%). FLNC patients were younger (median age 30 [13-51] vs 57 [41-65] years, p<0.0008) and more likely to present with sustained ventricular arrhythmia (5/21 24% vs 3/20, 15%) (Table 1). Acute myocarditis episodes were exclusively linked to DSP cardiomyopathy (4/20, 20%) with no association to FLNC (0/21, 0%, p<0.05). FLNC cardiomyopathy patients had normal left ventricular (LV) systolic function (LVEF 57±9% SD vs 46±13% DSP, p<0.03) and normal left ventricular cavity side (mean indexed LV end diastolic volume (EDV) 85±12 vs 112±26ml/m2 DSP, p<0.02) (Table 2). Both FLNC and DSP cause exclusive left ventricular or biventricular late gadolinium enhancement (75% Vs 82%) on cardiac magnetic resonance imaging (MRI). Fibrosis was predominantly in a subepicardial distribution in both cohorts (66% FLNC vs 57% DSP). In FLNC, circumferential fibrosis was the most frequent pattern of distribution (5/9, 55%), whereas in DSP inferior and lateral wall fibrosis was the predominant feature (8/14, 57%). There were no cases of isolated RV involvement and no cases fulfilled ‘definite’ ARVC taskforce diagnostic criteria. Over a follow up of median 45months, there was no significant difference in the proportion of FLNC and DSP patients experiencing a composite end point of LVAD implantation, heart transplant, aborted sudden cardiac death or mortality (7/21, 33% FLNC Vs 7/20, 29% DSP p>0.99). Conclusion Normal systolic function with marked circumferential fibrosis is the predominant feature of FLNC cardiomyopathy whereas mild LV dilatation, mild systolic impairment and inferolateral fibrosis are more frequent in DSP cardiomyopathy. Myocarditis episodes are a distinguishing feature of DSP cardiomyopathy, suggesting a different underlying molecular mechanism driven by inflammation. The overall clinical trajectory of FLNC and DSP cardiomyopathy patients are comparable.