# Withdrawing guideline-directed medical therapy after left ventricular ejection fraction recovery following atrial fibrillation ablation: a multicentre cohort study

## TL;DR

This study finds that reducing heart failure medications after atrial fibrillation ablation does not harm heart function in patients with recovered ejection fraction.

## Contribution

The study provides evidence that guideline-directed medical therapy can be safely de-escalated after left ventricular ejection fraction recovery in atrial fibrillation-induced cardiomyopathy.

## Key findings

- Mean LVEF remained comparable between groups continuing and withdrawing GDMT.
- LV end-diastolic diameter and sinus rhythm maintenance were stable after GDMT withdrawal.
- Only 2% of patients experienced relapse unrelated to GDMT withdrawal.

## Abstract

Atrial fibrillation (AF)-induced cardiomyopathy (AIC) is characterised by reversible left ventricular (LV) dysfunction after restoration of sinus rhythm (SR). The need for continued guideline-directed medical therapy (GDMT) for heart failure after LV ejection fraction (LVEF) recovery in AIC after catheter ablation (CA) is unclear.

This multicentre cohort study across 12 UK centres included adults undergoing index AF ablation (June 2019–June 2024) with LVEF <50% preablation and recovery to ≥50% at three timepoints: preablation; early postablation (≥4 weeks) and late postablation (≥3 months or ≥3 months post-GDMT withdrawal). Patients were stratified post recovery of LVEF after CA. The primary outcome was mean LVEF at late follow-up; secondary outcomes included absolute change in LVEF, LV end-diastolic diameter (LVEDD) and SR maintenance.

88 patients met inclusion enrolment criteria (61.7±10.6 years old; 91% male), of which 50 (56.8%) continued full-dose GDMT and 38 (43.2%) withdrew ≥50% of GDMT. In the GDMT-withdrawn group, mean GDMT classes decreased from 2.97±0.88 to 1.03±0.79 (p<0.001). At late follow-up, mean LVEF was comparable (56.3%±3.8% GDMT-continued vs 56.8%±5.5% GDMT-withdrawn; p=0.59), as was LVEF change (1.2% vs 0.4%; p=0.48). One relapse occurred in each group secondary to an acute coronary syndrome (2.0% vs 2.6%; p=1). LVEDD remained stable (p>0.8). SR was maintained in 82.0% vs 92.1% of patients; p=0.17.

Selective GDMT withdrawal after sustained LVEF recovery and rhythm control did not compromise LV systolic function, remodelling or rhythm maintenance. This supports the study of personalised de-escalation strategies in AIC in prospective trials.