CASTLE-AF

Published

Marrouche NF, et al. "Catheter Ablation for Atrial Fibrillation with Heart Failure". The New England Journal of Medicine. 2018. 378(5):417-27.
PubMed • Full text • PDF

Clinical Question

In patients with atrial fibrillation (AF) and symptomatic (NYHA II-IV) systolic heart failure (LVEF ≤ 35%), is catheter ablation superior to medical therapy (rate or rhythm control) with regard to death or hospitalization for heart failure?

Bottom Line

In patients with AF and symptomatic (NYHA II-IV) systolic heart failure (LVEF ≤ 35%), catheter ablation is associated with a 16.1% absolute reduction in death or hospitalization for heart failure when compared to medical therapy (rate or rhythm control). This difference was driven both by a 11.6% absolute reduction in death and a 15.2% absolute reduction in hospitalization for heart failure. Catheter ablation was also associated with greater improvement in LVEF and long-term maintenance of sinus rhythm.

Major Points

AF is frequently comorbid with systolic heart failure, and is associated with increased risk of stroke, hospitalization for heart failure, and death and this population. The ideal management strategy for AF in the setting of significant systolic heart failure, however, remains unclear. Multiple RCTs including the landmark AFFIRM trial (2002) have demonstrated similar outcomes with a rate control strategy (i.e., use of AV nodal blockade to attain a safe heart rate with no intent to restore sinus rhythm) versus a rhythm control strategy (i.e., use of antiarrhythmic drugs [AAD] or procedures to restore and maintain sinus rhythm) in patients with AF. Nevertheless, these initial trials included primarily older patients without structural heart disease and in particular patients with significant LV dysfunction were underrepresented. Furthermore, patients with significant LV dysfunction often fail to tolerate the degree of AV nodal blockade required to maintain adequate rate control due to negative inotropic effects or are dependent on atrial contraction to maintain adequate forward flow. Finally, the majority of these trials were performed prior to the emergence of catheter ablation for AF, which has demonstrated good efficacy in maintaining sinus rhythm and does not pose the same risk of long-term side effects as antiarrhythmic drugs (see FIRE and ICE, 2016).

The recent AATAC trial demonstrated that in patients with symptomatic (NYHA II-III) systolic heart failure (LVEF ≤40%) and persistent AF chosen for a rhythm control strategy, catheter ablation is associated with reductions in AF recurrence, hospitalization for heart failure, and death when compared to medical rhythm control. Notably, AATAC was limited by its lack of an inclusion of a rate control arm, as it remains unclear whether rhythm control (of any kind) is associated with improved outcomes when compared to pure rate control in patients with systolic heart failure. Furthermore, AATAC was not designed to assess for hard cardiovascular outcomes.

The 2018 Catheter Ablation versus Standard Conventional Therapy in Patients with Left Ventricular Dysfunction and Atrial Fibrillation (CASTLE-AF) study randomized 363 patients to catheter ablation versus medical therapy and assessed for a primary outcome of death or hospitalization for heart failure. Importantly, medical therapy could include either rate or rhythm control, although rhythm control was encouraged (70% of patients were ultimately rate controlled). At a median 37.8 months, catheter ablation was associated with a 16.1% absolute reduction in death or hospitalization for heart failure when compared to medical therapy. This difference was driven both by a 11.6% absolute reduction in death and a 15.2% absolute reduction in hospitalization for heart failure. Patients randomized to ablation underwent a mean of 1.3 ablations during the study period. 63% of patients randomized to catheter ablation were in sinus rhythm at 60 months compared to 22% of patients randomized to medical therapy. In addition, patients randomized to ablation had an improvement in LVEF of 8% at 60 months compared to 0.2% in the medical therapy group.

CASTLE-AF adds to emerging evidence suggesting that catheter ablation is not only associated with improved rhythm control in patients with AF and symptomatic systolic heart failure, but it is also associated with improved cardiovascular outcomes in this population, with lower rates of death and hospitalization for heart failure. Importantly, patients had to have failed or refused antiarrhythmic drug therapy in order to be included in the trial, limiting the generalizability of these findings to patients for whom rhythm control is never attempted or is easily obtained using anti-arrhythmic drugs. Nevertheless, the inclusion of a rate control arm in CASTLE-AF is a critical first step towards the suggestion that catheter ablation may be a reasonable first-line treatment for AF in patients with heart failure. Future studies randomizing all comers with AF and systolic heart failure to catheter ablation versus medical rate control are needed before an upfront role for catheter ablation can be fully established.

Guidelines

As of February 2018, no guidelines have been published that reflect the results of this trial.

Design

• Multicenter, open-label, randomized, controlled trial
• N=363
  • Catheter ablation (N=179)
  • Medical therapy (N=184)
• Setting: 33 sites in Europe
• Enrollment: January 2008 - January 2016
• Median follow-up: 37.8 months
• Analysis: Modified intention-to-treat (excluded patients who died or withdrew during run-in period)
• Primary Outcome: Death or hospitalization for heart failure

Population

Inclusion Criteria

• Age ≥ 18 years
• Symptomatic paroxysmal or persistent AF
  • AF episodes had to be documented in the last 3 months prior to enrollment by ECG, holter, loop recorder, or implanted device
• Failure of antiarrhythmic drug therapy or unwillingness to take antiarrhythmic drugs
• Left ventricular dysfunction with left ventricular ejection fraction ≤ 35% (measured within last 6 weeks)
• NYHA class II-IV
• Indication for implantable defibrillator therapy for primary or secondary prevention

Exclusion Criteria

• Previous ablation procedure for AF
• Contraindication to anticoagulation therapy
• Document left atrial diameter > 6cm
• Acute coronary syndrome, cardiac surgery, angioplasty, or stroke within 2 months
• Planned cardiovascular intervention
• Listed for heart transplant
• Cardiac assist device implanted
• Life expectancy < 1 year
• Uncontrolled hypertension
• Uncontrolled thyroid disease
• Requirement for dialysis due to terminal renal failure
• Mental or physical incapability to participate in the study
• Women currently pregnant or breastfeeding or not using reliable contraception
• Enrollment in another investigational drug or device study, or participation in another telemonitoring concept

Baseline Characteristics

From ablation group.

• Demographics: Age 64 years, male 87%
• Comorbidities: BMI 29.0
• CHF: NYHA I 11%, NYHA II 58%, NYHA III 29%, NYHA IV 2%, ischemic 40%, nonischemic 60%, LVEF 32.5%
• AF: paroxysmal 30%, persistent 70%, long-standing persistent 28%, LA diameter 48mm
• Devices: CRT-D 27%, ICD 73%,
• Medications: History of amiodarone use 57%

Interventions

• Patients randomized 1:1
  • Cather ablation
  • Medical therapy
• A run-in phase of 5 weeks was used to adjust the administration of medications for heart failure in accordance with latest guidelines
• After the run-in period, a baseline assessment was performed, and patients were referred for assigned treatment
• Aim of ablation was isolation of all pulmonary veins and restoration of sinus rhythm
  • Additional ablation lesions at the operator's discretion
  • All operators had to have performed at least 50 ablation procedures and were allowed to use their preferred ablation system
• Left atrial thrombus was ruled out with transesophageal echocardiography before any ablation
• After ablation, all patients received warfarin for at least 6 months
• Medical therapy was administered in accordance with guidelines, with maintenance of sinus rhythm recommended
• Patients chosen for rate control were titrated for goal heart rate 60-80 at rest and 90-115 with exertion
• For all patients, the Biotronik Home Monitoring option was activated to monitor for recurrent AF
  • Recurrent AF defined as AF lasting at least 30 seconds
• If recurrent AF was found, repeat ablation was recommended unless contraindicated clinically
• Patients attended regular follow-up at 3, 6, 12, 24, 36, 48, and 60 months at which devices were checked, adverse events documented, echocardiograms performed, and 6-minute walk tests conducted

Outcomes

Comparisons are ablation vs. medical therapy

Primary Outcomes

Death or hospitalization for heart failure

51 (28.5%) vs. 82 (44.6%); HR 0.62 (95% CI 0.43-0.87); p = 0.007; NNT = 6

Secondary Outcomes

Death

24 (13.4%) vs. 46 (25.0%); HR 0.53 (95% CI 0.32-0.86) p = 0.01; NNT = 9

Heart failure hospitalization

37 (20.7%) vs. 66 (35.9%); HR 0.56 (95% CI 0.37-0.83); p = 0.004; NNT = 7

Cardiovascular death

20 (11.2%) vs. 41 (22.3%); HR 0.49 (95% CI 0.29-0.84); p = 0.009; NNT = 9

Cardiovascular hospitalization

64 (35.8%) vs. 89 (48.4%); HR 0.72 (95% CI 0.52-0.99); p = 0.04; NNT = 8

All hospitalization

114 (63.7%) vs. 122 (66.3%); HR 0.99 (95% CI 0.77-1.28); p = 0.96

Stroke

5 (2.8%) vs. 11 (6.0%); HR 0.46 (95% CI 0.16-1.33); p = 0.15

Subgroup Analysis

• There were no significant interactions with the primary endpoint among the subgroups tested.

Adverse Events

• 3 patients in the ablation group developed a pericardial effusion
  • 1 patient required pericardiocentesis
• 3 patients had severe bleeding requiring transfusion
• 1 patient was noted to develop asymptomatic pulmonary vein stenosis

Criticisms

• Highly selected patient population from open-label design limits generalizability to all-comers with HFrEF + AF
• Mortality difference may be driven by harm from medical therapy group, with >40% having failed amiodarone and likely requiring advanced antiarrhythmics
• Limited generalizability to those with easily obtained rhythm control, who did not meet inclusion criteria for this trial.
• Potential bias in outcomes assessment from lack of blinding and open-label design, though mitigated by outcomes adjudication performed by a blinded independent committee.
• High rates of crossover from the ablation group may lead to bias, though more likely toward the null hypothesis, which would attenuate the primary outcome
• Although rate control was allowed in the medical therapy arm, rhythm control was encouraged. It thus remains unclear whether a strategy of pure rate control may be equally effective or superior to catheter ablation.

Funding

• Study sponsor Biotronik assisted with data management and quality control of the statistical analysis
• Authors with multiple ties to industry