The role of right ventricular late gadolinium enhancement scoring by cardiac magnetic resonance imaging in cardiac sarcoidosis

Abstract

Abstract Background The right ventricular (RV) function is a significant predictor of the prognosis and exercises tolerance in patients with left ventricular dysfunction. The previous study reported that RV late gadolinium enhancements (LGE) as assessed by cardiac magnetic resonance (CMR) imaging were present in 16% of pulmonary sarcoidosis. Although RV LGE is associated with a poor prognosis, a semi-quantitative assessment of RV LGE was not been established so far. Moreover, the frequency of RV dysfunction (RVD) in cardiac sarcoidosis (CS) and predictors of RVD are unclear. Purpose Our aim of this study is to establish RV LGE semi-quantitative scores and to clarify the rate and predictors of RVD in CS. Methods Forty-four consecutive patients with CS who underwent CMR imaging (median age, 63 [IQR 54–71] years old; 10 males) were studied. They were diagnosed as definitive based on the Heart Rhythm Society expert consensus statement. Two patients were excluded due to no long axial view. Thus, a total of 42 patients were included in this study. RV LGE was semi-quantitatively evaluated using the original score developed by ours. RV was classified into RV outflow tract (OT), basal, middle and apex. RVOT and apex are assigned 1 segment for each, and basal and middle are assigned 4 segments include anterior, septal, inferior and lateral (Figure1). In total, 10 segments were evaluated. RV LGE scores were analysed independently by the two operators. The RV ejection fraction was assessed in a long axial view without an RVOT (20 phases, 12 slices). We divided the cohort into the non-RVD group (RVEF ≥45%) and the RVD group (RVEF <45%) and compared the RV scores between the two groups. Results The RVD was present in 31% of the entire cohort. The presence of RV LGE was significantly higher in the RVD group than the non-RVD group (83% vs. 43%, p=0.037). RV LGE score in the RVD group was significantly higher than that in the non-RVD group (2.83±1.95 vs. 0.83±1.21, p=0.001). The septal lesions in basal and middle levels accounted for 54% in all segments. RVEF showed a significant inverse correlation with RV LGE score (r=−0.57, p<0.0001). In the ROC curve for RVD, the RV LGE score of more than 2 and equal could discriminate RVD from non-RVD with AUC of 0.813 (Sensitivity = 0.77, Specificity = 0.76, p=0.003). Conclusion RVD was observed in 31% of patients with definitive CS. The septal lesions were the most frequent in the RV LGE site, which suggests that the longitudinal abnormal motion of RV is a significant predictor of the RVD in patients with CS. RV LGE score is a valuable method to assess RVD. RV LGE score Funding Acknowledgement Type of funding source: None