Background Up to fifty percent of idiopathic dilated cardiomyopathy (DCM) has a familial basis. Variants can occur in over 40 genes, though truncating variants in the sarcomeric gene titin account for the largest proportion (~20%). At least half of familial DCM cases are genetically orphan. We sought to study whether familial DCM was associated with distinct clinical characteristics, independently of the underlying genetic variant.

Methods 595 prospectively recruited DCM patients underwent detailed phenotyping with cardiac MRI (Siemens scanners, 1.5T) and were sequenced using a customised panel of ~100 cardiomyopathy genes on Illumina and 5500xl platforms. Variants were identified and annotated using a customised bioinformatics pipeline. Clinical information including family pedigree data, ECG, and arrhythmia status at diagnosis (presence of confirmed ventricular or atrial arrhythmias) was collected on all patients. Familial DCM was defined as DCM occurring in 2 or more 1st or 2nd degree family members. Chi squared or Fisher’s exact test was used to compare across categorical variables and t-tests or Mann-Whitney U tests across continuous variables as appropriate.

Results Overall, 16% of patients (95 out of 595) had familial DCM. Thirty individuals came from 13 families, the remaining were unrelated probands.

Patients with familial DCM had an earlier age of disease onset (49.8 years vs 58.8 years, p < 0.0001). Non-familial DCM was characterised by a male preponderance (71% vs 56%, p = 0.004).

Patients with familial DCM had less conduction disease at baseline (11% vs 36%, p < 0.0001). There was no difference in confirmed VT, NSVT or atrial fibrillation at baseline between groups.

Patients with familial DCM had a milder intermediate phenotype of DCM (left ventricular ejection fraction 45.2% vs 38.2%, p < 0.0001). Right ventricular ejection fraction was similar in both groups (39.1% familial vs 37.1% non-familial, p = 0.14). There was no difference in the presence of mid wall fibrosis detected on late gadolinium imaging (p = 0.54).

There were 44 potentially disease-causing variants in DCM genes in the familial DCM cohort (Table 1). Genetic testing had a yield of 44% in familial (n = 42), and 22% in non-familial DCM (n = 117). Five patients carried 2 variants. Truncating variants in titin were the most common variant (n = 17) and were over twice as common in patients with familial DCM compared to those without (18% vs 6.8%, p < 0.001). Truncating and missense variants in LMNA were ten times more frequent in familial DCM compared to non familial DCM (p < 0.001).

Conclusions Disease causing variants in TTN and LMNA were more commonly associated with familial DCM, but 56% of patients with familial DCM remain genetically unexplained. This highlights the need for further novel DCM disease gene discovery. Our data show that familial DCM is characterised by a younger age of disease onset and less severe ventricular dysfunction as compared to non-familial DCM.