Abstract
BACKGROUND: Epicardial adipose tissue (EAT) contributes to inflammation and fibrosis of the neighboring myocardial tissue via paracrine signaling. In this retrospective study, we investigated the abnormal changes in the amount of EAT in male children with Duchenne muscular dystrophy (DMD) using cardiac magnetic resonance (CMR) imaging. Furthermore, we constructed and validated a nomogram including EAT-related CMR imaging parameter for predicting the occurrence of myocardial fibrosis in patients with DMD. METHODS: This study enrolled 283 patients with DMD and 57 healthy participants who underwent CMR acquisitions to measure the quantitative parameters of EAT, pericardial adipose tissue (PAT), paracardial adipose tissue, and subcutaneous adipose tissue. Late gadolinium enhancement (LGE) was performed to confirm myocardial fibrosis in patients with DMD. The DMD group consisted of 200 patients from institution 1 (the ratio of the training set and the internal validation set was 7:3) and 83 patients from four other institutions (the external validation set). Logistic and least absolute shrinkage and selection operator (LASSO) regression was used to select the optimal predictors and to develop and validate the nomogram model predicting LGE risk in the training set, internal validation set, and external validation set. RESULTS: Compared with those in healthy controls, some regional EAT thicknesses, areas, and global volumes were significantly higher in patients with DMD, and 41.7% of patients with DMD showed positive LGE. These LGE-positive patients with DMD showed significantly higher EAT volume (median 23.9 mL/m(3); P<0.001) and PAT volume (median 31.8 mL/m(3); P<0.001) compared with the LGE-negative patients with DMD. Age [odds ratio (OR) 2.0; P<0.001], body fat percentage (OR 1.3; P<0.001), and EAT volume (OR 1.4; P<0.001) were independently associated with positive LGE in the training set. The interactive dynamic nomogram showed superior prediction performance, with a high degree of the calibration, discrimination, and clinical net benefit in the training and validation of the DMD datasets. The area under the curve (AUC) values of the nomogram in the training set, internal validation set, and external validation set were 0.95 [95% confidence interval (CI): 0.91-0.98], 0.97 (95% CI: 0.92-0.99), and 0.95 (95% CI: 0.91-0.99), respectively. CONCLUSIONS: The onset of LGE-based myocardial fibrosis was associated with EAT volume in patients with DMD. Additionally, the nomogram with EAT volumes showed superior performance in patients with DMD for predicting the occurrence of myocardial fibrosis.