Extra cellular volume by computed tomography is useful for prediction of prognosis in cardiac sarcoidosis

Abstract Background Extracellular volume (ECV) fraction on magnetic resonance imaging (MRI) has been useful in evaluating the degree of myocardial fibrosis and has been reported to help diagnose and evaluate the prognosis of cardiomyopathy. However, there are also several contraindications in the cardiac MRI, and the performance of it takes time. Computed tomography (CT) helps screen for coronary artery disease, is more versatile than cardiac MRI, and is also useful in screening for coronary artery disease. Myocardial extracellular volume (ECV) analysis on CT has been recently eligible and has shown a good correlation with ECV on MRI. We hypothesized that ECV on CT, a new fibrosis indicator, would be useful in predicting prognosis in patients with cardiac sarcoidosis (CS). Purpose The purpose of this study is to evaluate the utility of ECV analysis on CT for predicting major adverse cardiac events (MACEs) in cases with cardiac sarcoidosis (CS). Methods As a single-center study, 48 patients (18 males, 67 ± 11 years old) diagnosed with CS based on the 2016 Japanese Circulation Society criteria and who underwent cardiac CT using 320-row detector CT or 256-row detector CT, and including late-phase images from December 2008 to January 2024 were analyzed retrospectively. We evaluated the ECV of left ventricular myocardium (LVM) using commercially available software (Figure A). The primary endpoint was MACEs, a composite of all-cause death, heart failure hospitalization, and fatal ventricular arrhythmia. Results Among all 48 patients, 14 combined events (10 fatal ventricular arrhythmias, three heart failure hospitalizations, and one all-cause death) were observed during a follow-up of 56 ± 57 months. ECV of LVM on CT was significantly higher in patients with MACEs than the others (37 ± 5.6% vs 33 ± 4.6%, P = 0.0087). The best cut-off value of ECV of LVM on CT for prediction of MACEs was 34.97%, and the sensitivity and specificity of LV-ECV to predict MACEs were 71% and 76% at the best cut-off, and the area under the curve was 0.73 based on the receiver operating characteristics curve analysis (P = 0.009) (Figure B). The cases with ECV ≥ 34.97% (n = 18) had significantly higher MACEs than the cases with ECV < 34.97% based on the Kaplan-Meier analysis (P = 0.0031) (Figure C). In the multivariate Cox proportional hazard model, ECV on LVM (Hazard ratio 1.11, 95% confidence interval 1.00-1.21, p=0.0493) and history of VT ablation (Hazard ratio 27.7, 95% confidence interval 4.27-180, p=0.0011) were the significant predictors of MACEs. Conclusion Adverse events mainly caused by fatal ventricular arrhythmia are relatively common in CS. Evaluation of ECV by CT may be useful for the prediction of MACEs in cases with CS.