Catheter Ablation in End-Stage Heart Failure with Atrial Fibrillation (CASTLE-HTx)

Authors, Journal, Affiliations, Type, DOI

• Authors: Christian Sohns, Henrik Fox, Nassir F. Marrouche, Harry J.G.M. Crijns, Angelika Costard-Jaeckle, Leonard Bergau, Gerhard Hindricks, Nikolaos Dagres, Samuel Sossalla, Rene Schramm, Thomas Fink, Mustapha El Hamriti, Maximilian Moersdorf, Vanessa Sciacca, Frank Konietschke, Volker Rudolph, Jan Gummert, Jan G.P. Tijssen, Philipp Sommer, for the CASTLE-HTx Investigators
• Journal: New England Journal of Medicine (N Engl J Med 2023;389:1380–1389)
• Affiliations: Herz- und Diabeteszentrum Nordrhein-Westfalen, Ruhr-Universität Bochum, Bad Oeynhausen, Germany (primary); additional collaborators from Germany, USA (Tulane), and the Netherlands (Maastricht, Amsterdam)
• Type of study: Single-center, open-label, investigator-initiated, superiority randomized controlled trial; funded by Else Kröner-Fresenius-Stiftung (unrestricted educational grant); German health insurance covered ablation costs
• DOI: https://doi.org/10.1056/NEJMoa2306037

Overview

CASTLE-HTx is the first RCT to assess catheter ablation in patients with symptomatic atrial fibrillation and end-stage heart failure who were being evaluated for heart transplantation or LVAD implantation. Patients (n=194) were randomized 1:1 to catheter ablation plus guideline-directed medical therapy (GDMT) or GDMT alone. The trial was stopped early for overwhelming efficacy: after a median 18-month follow-up, the primary composite endpoint (all-cause death, LVAD implantation, or urgent heart transplantation) occurred in 8% of ablation patients vs 30% of controls (HR 0.24, P<0.001). All-cause mortality was similarly reduced (6% vs 20%, HR 0.29), and ablation produced significant improvements in LVEF and AF burden not seen with medical therapy.

Keywords

End-stage heart failure, atrial fibrillation, catheter ablation, pulmonary vein isolation, heart transplantation, left ventricular assist device, guideline-directed medical therapy, arrhythmia burden, LVEF reverse remodeling

Key Takeaways

Trial Design and Patients

• Single-center RCT at a high-volume heart transplantation referral center (~80 transplants/year; ~3,000 total since 1989)
• Enrollment: November 2020 – May 2022; 97 patients per arm
• Inclusion criteria: symptomatic AF + NYHA class ≥II + LVEF ≤35% + impaired 6-minute walk test + referral for transplant/LVAD evaluation using ISHLT listing and ESC-HFA position statement criteria
• All patients were clinically stable at randomization and had an implantable cardiac device with activated arrhythmia detection for continuous rhythm monitoring
• Planned sample size: 194 patients (80% power at two-sided α=0.05, assuming 20% annualized event rate in controls and HR 0.50)

Intervention

• Ablation group: Direct current cardioversion after transseptal puncture (if AF at start); aim = electrical isolation of all pulmonary veins + sinus rhythm restoration; ablation beyond PVs at operator discretion; median 20 days post-randomization to ablation; 51/81 had PVI alone, 30/81 had PVI + additional substrate; 9% underwent multiple procedures
• Medical therapy group: AHA/ACC/HRS and ESC guideline-directed management of AF (rate control + sinus rhythm maintenance per guideline recommendations); 16/97 (16%) crossed over to ablation
• All antiarrhythmic drugs discontinued after ablation; reinitiated at discretion if arrhythmia recurred

Primary Endpoint

• Composite: death from any cause + LVAD implantation + urgent heart transplantation
• Ablation: 8/97 (8%) vs Medical therapy: 29/97 (30%); HR 0.24 (95% CI 0.11–0.52); P<0.001
• Kaplan–Meier estimates at 1 year: 6% (ablation) vs 28% (medical); at 2 years: 9% vs 31%
• Trial stopped by DSMB ~1 year after enrollment completion (post-hoc P<0.001, Haybittle–Peto rule met)

Secondary Endpoints

• All-cause death: 6% vs 20%; HR 0.29 (95% CI 0.12–0.72)
  • Kaplan–Meier 1-year mortality: 4% vs 14%; 2-year: 6% vs 23%
  • Nearly all deaths cardiovascular; one non-cardiovascular death per group
• LVAD implantation: 1% vs 10%; cause-specific HR 0.09 (95% CI 0.01–0.70)
• Urgent heart transplantation: 1% vs 6%; cause-specific HR 0.15 (95% CI 0.02–1.25)
• LVEF improvement: Ablation: +6.7 pp at 6 months; +7.8 pp at 12 months; Medical: +1.2 pp at 6 months; +1.4 pp at 12 months (between-group difference at 12 months: 6.4 pp [95% CI 4.1–8.7])
• AF burden reduction: Ablation: −30.8 pp at 6 months; −31.4 pp at 12 months; Medical: −8.3 pp at 6 months; −8.6 pp at 12 months (between-group difference at 12 months: 22.7 pp [95% CI 13.0–32.5])
• At 12 months: 56% of ablation patients were free from primary endpoint events and not in persistent AF, vs only 9% of control patients
• Amiodarone use at 12 months: 29% (ablation) vs 57% (medical therapy)

Procedural Safety

• 4 procedure-related complications total (3 in ablation group, 1 in medical therapy group); all were vascular access site complications — no atrioesophageal fistulae, no tamponade reported
• Ablation performed by operators with >400 prior procedures each

Statistical Analysis

• Intention-to-treat primary analysis; Cox proportional-hazards for time-to-event; Kaplan–Meier curves; independent t-test for continuous secondary endpoints; SPSS v29
• No multiplicity correction: secondary endpoints as point estimates with 95% CIs only

Limitations of the Document

• Single-center study at a high-volume referral center — results may not be generalizable to lower-volume centers or centers with less experience in ablation for end-stage HF
• Early termination for efficacy — effect estimates may be inflated relative to the true population effect (stopped after median 18 months vs planned 3 years)
• Open-label design — treatment decisions for LVAD/transplantation were made at heart-team meetings without knowledge of trial arm, but unblinded treating physicians could introduce bias
• Crossover: 16% of medical-therapy patients underwent ablation; this dilutes the ITT treatment effect and likely underestimates the true benefit of ablation
• Short follow-up: Median 18 months; long-term outcomes (cardiac function trajectory, late recurrences, transplant candidacy changes) unknown
• Asymptomatic AF not studied: Whether findings extend to asymptomatic AF is unknown
• No data on specific HF aetiology (ischaemic vs non-ischaemic) in detail as a pre-specified subgroup (Table S5 referenced but not shown in text)

Key Concepts Mentioned

• concepts/Catheter-Ablation-AF — PVI strategy and role in end-stage HFrEF
• concepts/Atrial-Failure — ablation evidence in AF + advanced HF
• concepts/Electrical-Remodeling — AF burden reduction and LV reverse remodeling post-ablation
• concepts/PVC-Induced-Cardiomyopathy — analogous tachycardia-mediated LV dysfunction reversibility
• concepts/Antiarrhythmic-Drugs — amiodarone reduced to 29% post-ablation vs 57% in control arm

Key Entities Mentioned

• entities/Atrial-Fibrillation — the arrhythmia targeted
• entities/Heart-Failure — end-stage HFrEF with LVEF ≤35%

Wiki Pages Updated

• wiki/sources/CA-HF-CASTLEHTx-NEJM-2023.md — created (this file)
• wiki/concepts/Catheter-Ablation-AF.md — expanded CASTLE-HTx section with full trial data
• wiki/concepts/Atrial-Failure.md — added CASTLE-HTx as ablation evidence in end-stage disease
• wiki/sourceindex.md — new entry added
• wiki/wikiindex.md — Catheter-Ablation-AF entry updated
• log.md — new entry appended