P-wave duration-amplitude ratio quantifies atrial low-voltage area and predicts atrial arrhythmia recurrence following pulmonary vein isolation.

Abstract

Atrial low-voltage area (LVA) in AF patients increase the risk of atrial arrhythmia (AA) recurrence following pulmonary vein isolation (PVI). Contemporary LVA prediction scores (DR-FLASH, APPLE) do not include P-wave metrics. We aimed to evaluate the utility of P-wave duration-amplitude ratio (PWR) in quantifying LVA and predicting AA recurrence post-PVI. In 65 patients undergoing first-time PVI, 12-lead ECGs were recorded during sinus rhythm. PWR was calculated as the ratio between the longest P-wave duration and P-wave amplitude in lead I. High-resolution bi-atrial voltage maps were collected and LVAs included bipolar electrogram amplitudes <0.5mV or <1.0mV. A LVA quantification model was created using clinical variables and PWR, then validated in a separate cohort of 24 patients. Seventy-eight patients were followed for 12-months to evaluate AA recurrence. PWR strongly correlated with left atrial (LA) (<0.5mV: r=0.60; <1.0mV: r=0.68; p<0.001) and bi-atrial LVA (<0.5mV: r=0.63; <1.0mV: r=0.70; p<0.001). Addition of PWR to clinical variables improved model quantification of LA LVA at the <0.5mV (adjusted R2: 0.59 to 0.68) and <1.0mV cutpoints (adjusted R2: 0.59 to 0.74). In the validation cohort, PWR model-predicted LVA correlated strongly with measured LVA (<0.5mV: r=0.78; <1.0mV: r=0.81; p<0.001). PWR model was superior to DR-FLASH (AUC 0.90 vs. 0.78; p=0.030) and APPLE (AUC 0.90 vs. 0.67; p=0.003) at detecting LA LVA and comparable at predicting AA recurrence post-PVI (AUC=0.67 vs. 0.65 and 0.60). Our novel PWR model accurately quantifies LVA and predicts AA recurrence post-PVI. PWR model-predicted LVA may help guide patient selection for PVI.