PO-02-225 LATE GADOLINIUM ENHANCEMENT AT CARDIAC MAGNETIC RESONANCE YIELDS HIGH NUMBER OF FALSE POSITIVE DIAGNOSES IN ARRHYTHMOGENIC RIGHT VENTRICULAR CARDIOMYOPATHY

Abstract

New diagnostic criteria have been proposed for arrhythmogenic right ventricular cardiomyopathy (ARVC). These proposed “Padua criteria” include late gadolinium enhancement (LGE) on cardiac magnetic resonance for the left ventricle (LV) and right ventricle (RV). Yet, the potential of LGE to distinguish ARVC from differential diagnoses remains unknown. Assess the diagnostic value of LGE to distinguish ARVC from its differentials. We included 130 subjects (57% male, 46±17 years) clinically diagnosed with ARVC as per 2010 Task Force Criteria (n=32, 84% PKP2 carriers), familial dilated cardiomyopathy as per World Health Organization criteria (n=25), myocarditis as per 2013 ESC position statement (n=13), sarcoidosis as per 2014 HRS expert consensus statement (n=20), biopsy-proven cardiac amyloidosis (n=19) and RV outflow tract-ventricular tachycardia (n=21). Presence and distribution of LGE was evaluated by an experienced radiologist using a 17-segment LV and 7-segment RV model. Presence of LGE was determined as per Padua criteria. Diagnostic value of RV-LGE and LV-LGE as per Padua criteria in ARVC vs. non-ARVC patients was evaluated with specificity, false positive rate (FPR) and accuracy. Distributions of LGE per subgroup are shown in the Figure. The RV-LGE criterion was equally often fulfilled in ARVC (n=16, 50%) vs. non-ARVC (n=43, 44%) patients, p=0.546. ARVC patients most often had RV-LGE in the RV inlet region (n=14/32, 44%). Consequently, the RV-LGE criterion yielded specificity 56%, FPR 44% and accuracy 55% for right-sided disease. In contrast, the LV-LGE criterion was significantly more often fulfilled in non-ARVC (n=71, 72%) vs. ARVC (n=11, 34%) patients, p<0.001. ARVC patients most often had LV-LGE in the inferolateral (n=10/11, 91%) wall, whereas it was more global in non-ARVC groups (median segments n=2 [IQR 2-6] ARVC vs. n=9 [IQR 4-15] non-ARVC, p<0.001). As such, the LV-LGE criterion yielded specificity 28%, FPR 72% and accuracy 29% for left-sided disease. LV-LGE is observed in one-third of ARVC patients, most often in the inferolateral region. RV-LGE is observed in half of ARVC patients, typically in the RV inlet region. Notably, LGE is often present in ARVC differentials. While LGE may be helpful to determine focal fibrosis, it may result in a large number of false positive diagnoses when included as a diagnostic criterion. Knowledge of LGE patterns can be useful for diagnostic classification.