Refining arrhythmic mitral valve prolapse definition by including left ventricular origin of arrhythmias: Insights from long-term cardiac monitoring

Background The definition of arrhythmic mitral valve prolapse (AMVP) and risk stratification are insufficient. The 2022 European Heart Rhythm Association (EHRA) expert consensus document on AMVP defined AMVP as either (i) a premature ventricular complex (PVC) burden of ≥5% on Holter monitoring, or (ii) the presence of ventricular tachyarrhythmias. However, this definition does not differentiate between different origins of the PVCs, potentially misclassifying patients for AMVP. Purpose In this prospective study, we aimed to assess left ventricular PVCs (LV-PVCs) as a risk factor for subsequent ventricular arrhythmia (VA) assessed by implantable cardiac monitor (ICM). We also explored the impact of including LV-PVCs in the AMVP definition to improve selection of patients at risk of VA. Methods We prospectively implanted an ICM, irrespective of clinical indication, in patients with MVP and/or mitral annular disjunction and followed them for 3 years. We assessed occurrence of severe VA (ventricular fibrillation, sustained ventricular tachycardia [VT], or non-sustained VT causing syncope) and VA (severe VA and non-sustained VT) by ICM. PVCs origins were classified on 12-lead ECG. Results We included 60 patients from 2015 to 2020 (73% female, mean age 49±15years) with no previous severe VA. Median PVC burden by 24-hour Holter monitoring was 0.3% per 24-hour (IQR 0.03-1.98) at the time of ICM implantation. LV-PVCs were present in 30 (50%) patients. One hundred VAs were detected over a mean duration of 3.0±0.5 years, 96 non-sustained VTs and 4 severe VA (1 non-sustained VT causing syncope, 1 sustained polymorphic VT, 2 cases of ventricular fibrillation). LV-PVCs were strongly associated with higher occurrence of VA after adjusting for baseline non-sustained VTs and PVC burden (multivariate IRR of 10.77 (95% CI 3.23-36.06, p<0.001). Severe VA occurred only in patients with LV-PVCs (log-rank p=0.04) (Figure 1). Replacing the PVC burden in the 2022 EHRA AMVP definition with the presence of any LV-PVCs, improved VA predictive accuracy from an AUC of 0.63 to 0.75, (p=0.06), with sensitivity increasing from 46% to 92%, and negative predictive value increasing from 69% to 91% (Figure 2). Conclusion LV-PVCs was strongly associated with subsequent VA in MVP patients and increased the predictive ability to detect arrhythmic risk. We propose to use LV-PVCs in the AMVP definition.