CE-522-03 LONGITUDINAL PREDICTION OF VENTRICULAR ARRHYTHMIAS IN PATIENTS WITH ARRHYTHMOGENIC RIGHT VENTRICULAR CARDIOMYOPATHY

Abstract

The arrhythmogenic right ventricular cardiomyopathy (ARVC) risk calculator (https://www.arvcrisk.com) stratifies risk for ventricular arrhythmia (VA) at the time of ARVC diagnosis. However, the risk factors included in the ARVC calculator change with time, and how well it performs at follow up evaluations is unknown. To identify the temporal trends of VA risk factors and test the longitudinal performance of the ARVC calculator. This was a retrospective analysis of ARVC patients without prior VA (sustained ventricular tachycardia (VT), appropriate ICD therapy, sudden cardiac death/arrest) enrolled in the Johns Hopkins Hospital ARVC registry and Netherlands ACM registry. Risk factors for VA (age, sex, history of cardiac syncope, burden of PVC on 24-hour heart monitor, history of non-sustained VT, number of T-wave inversions on ECG, and right ventricular ejection fraction) were assessed at the time of ARVC diagnosis and upon repeat evaluations. The 5-year risk of VA was predicted longitudinally using 1) the ARVC calculator (baseline risk factors), 2) the ARVC calculator (updated risk factors), 3) time varying Cox regression. 408 patients (37±15 years old, 40% male) were followed for a median of 5.2 [2.8, 9.6] years, during which time 132 experienced a first VA (4.0% events/year). Comparison of risk factors at baseline versus at >5 years of follow up revealed decreases in 24-hour PVC Burden (3,560 vs 2,510) and prevalence of non-sustained VT on the most recent Holter (53% vs 27%), and an increase in number of T-wave inversions (3.4 vs 3.9 leads). Prevalence of right ventricular dysfunction was unchanged. When calculated using baseline risk factors, the ARVC calculator’s ability to predict 5-year VA risk worsened after 3 years of follow up (from ROC-AUC 0.82 at baseline to 0.69 by 5 years). When calculated using updated risk factors, the ARVC calculator (ROC-AUC 0.77) performed as well as time varying Cox regression (ROC-AUC 0.77) after 5 years of follow up (Figure). Risk factors for VA in ARVC change over time. Up-to-date assessment of these risk factors is warranted for VA risk stratification at evaluations beyond 3 years from the time of initial ARVC diagnosis.