Extracellular volume analysis on cardiac computed tomography is useful to predict prognosis after cardiac resynchronization therapy

Abstract Background Left-ventricular (LV) extracellular volume (ECV) (LVECV) is often evaluated on magnetic resonance imaging (MRI), and the value is well correlated with biopsy-identified myocardial fibrosis. Recently, mechanical development has been achieved, and ECV analysis on cardiac computed tomography (CCT) has been eligible using the latest software, and the ECV on CCT is well correlated with the ECV on MRI. Cardiac resynchronization therapy (CRT) has been usually performed in cases with severe heart failure, but almost 30 percent of them are non-responders. Therefore, the indicators to predict non-responders to cardiac resynchronization therapy (CRT) have been sought in advance. Purpose The purpose of this study is to reveal the utility of LVECV on CCT for the prediction of patient prognosis in the candidates of cardiac resynchronization therapy (CRT) who have low LV function. Methods We analyzed 53 patients (30 males, 65 ± 14 years old) with heart failure who underwent CCT, including late phase scan, using 320-row detector CT (Aquilion One or Aquilion One ViSion Edition, Cannon Medical Systems) or 256-row detector CT (Revolution CT Apex, GE Healthcare), before the implantation of a CRT device from December 2008 to January 2024. The primary outcomes were defined as cardiac death, hospitalization due to HF, and fatal arrhythmia after CRT implantation. CRT responders were defined as the patients whose LV endo-systolic volume was decreased by 15% or more on transthoracic echocardiography. Results Twenty-two patients (42%) had primary outcomes (10 cardiac death, 7 fatal arrhythmia, 5 hospitalization due to HF). LVECV was lower in CRT responder (n = 23) than the others (n = 25) (35 ± 6.3% vs 41 ± 6.8%, P = 0.006). LVECV was higher in the patients with primary outcomes (n = 22) than the others (n = 31) (35 ± 5.0% vs 41 ± 7.8%, P = 0.019). Based on the receiver-operating characteristics curve analysis, the best cut-off of LVECV for predicting primary outcomes was 44.03% (B). In the survival analysis using the Cox proportional hazard model, LVECV on CCT was a significant predictor of primary outcomes (HR, 1.07; 95% CI, 1.13-1.01; P = 0.012). On Kaplan-Meier analysis, the higher LVECV group (≥ 44.03%) had a significantly greater number of primary outcomes than the others (P = 0.010) (C). Conclusion LVECV on CCT may be useful for predicting responders of CRT and prognosis after CRT implantation.ROC curve and Kaplan-Meier analysis