Making Associations With Dilated Cardiomyopathy Using MRI

Pixabay License | Source: Michal Jarmoluk , No changes made.

Under the condition of dilated cardiomyopathy (DCM), the heart becomes enlarged and cannot pump blood effectively, which can proceed to heart failure. It is the most common reason for heart transplantation. There are both extrinsic and genetic factors that lead to DCM. The most common genetic causes of DCM are truncating variants in TTN (TTNtv), which are present in 15–20% of DCM cases. Rare variants are found in cardiomyopathy-related genes in 40% of DCM diagnoses. Interestingly, some rare variants associated with DCM do not manifest in disease.

DCM has clear morphological phenotype (features) of changes to the left ventricle. This means one mode of effective analysis is by cardiac imaging. It is unclear whether common genetic variants related to the heart’s structure and function can identify as-yet unaffected individuals at risk for DCM. Krishna G. Aragam of Harvard Medical School, Boston, and colleagues investigated common variant associations with cardiac structure and function in the UK Biobank, using automated measurements from 36,041 participants in order to ascertain the relationship between common genetic variants, cardiac imaging phenotypes, and risk for the development of DCM. The results were published in the journal Nature Communications.

Patients with congestive heart failure (CHF), coronary artery disease (CAD), or DCM at the time of enrollment were excluded. For the remaining participants seven cardiac MRI derived phenotypes were available: left ventricular end-diastolic volume (LVEDV), left ventricular end-systolic volume (LVESV), stroke volume (SV), the body-surface-area (BSA) indexed versions of each of these traits (LVEDVi, LVESVi, and SVi), and left ventricular ejection fraction (LVEF).

A series of genome-wide associations (GWAS) were made to identify common genetic variants associated with the seven cardiac MRI phenotypes. Fifty-seven loci were associated with at least one cardiac MRI phenotype at genome-wide significance threshold BOLTLMM (P < 5 × 10−8). In total, 22 loci associated with LVEDV, 14 with LVEDVi, 32 with LVESV, 28 with LVESVi, 22 with LVEF, 12 with SV, and eight with SVi, with a total of 45 previously unreported.


The results were validated in two additional biobanks; MESA and BioBank Japan. MESA demonstrated 88% agreement. In BioBank Japan there was 85% concordance. The association between polygenic scores and incident DCM (388 cases) was assessed in the remaining 358,556 individuals in the UK Biobank without cardiac MRI data, after excluding those with cardiac disease at baseline, and relatives. 

The polygenic scores for LVEDV, LVESV, LVEF, LVEDVi, and LVESVi were significantly associated with incident DCM after adjusting for confounding. The 28-SNP LVESVi polygenic score was associated most strongly with incident DCM (hazard ratio [HR] = 1.58 per SD increase in the score, P = 6.4 × 10−18 by Cox regression).

Further studies are needed to ascertain the weight of contributions between common variants, rare variants, and environmental factors in the pathogenesis of DCM. There is no universally agreed-upon GWAS P value threshold that accounts for larger sample sizes and rarer minor allele frequencies, so some could consider a more stringent threshold appropriate.

“In conclusion, we uncover common variants at 45 genetic loci not previously associated with left ventricular structure and function as measured by cardiac MRI, and reveal a robust link to DCM,” stated the authors.