CABANA
(2019)Objective
To determine whether a strategy of catheter ablation is more effective than conventional medical therapy (rhythm and/or rate control drugs) in reducing mortality, stroke, serious bleeding, and cardiac arrest in symptomatic patients with atrial fibrillation.
Study Summary
• Lower-than-expected event rates and a 27.5% crossover rate from drug therapy to ablation substantially diluted the treatment effect in the ITT analysis; treatment-received and per-protocol analyses showed HRs of 0.67 and 0.73-0.74 respectively, favoring ablation.
• Catheter ablation significantly reduced the composite of death or cardiovascular hospitalization (51.7% vs 58.1%; HR 0.83; P=.001) and AF recurrence (HR 0.52; P<.001), supporting a clinical benefit when patients actually received ablation.
Intervention
Catheter ablation (pulmonary vein isolation +/- ancillary ablation at operator discretion) vs antiarrhythmic drug therapy (rate and/or rhythm control per contemporaneous guidelines); all patients received anticoagulation per guidelines
Inclusion Criteria
Age >=65 years, or <65 years with >=1 stroke risk factor (hypertension, heart failure, prior stroke, diabetes, or other heart disease); >=2 episodes paroxysmal AF or >=1 episode persistent AF in prior 6 months; symptomatic and inadequately treated; suitable for catheter ablation or drug therapy
Study Design
Arms: Catheter ablation vs Drug therapy (rate and/or rhythm control)
Patients per Arm: 1108 (ablation), 1096 (drug therapy)
Outcome
• Death or cardiovascular hospitalization: 51.7% vs 58.1% (HR 0.83; 95% CI 0.74-0.93; P=.001) - significantly lower with ablation.
• AF recurrence: 49.9% vs 69.5% (HR 0.52; 95% CI 0.45-0.60; P<.001) - significantly lower with ablation.
Bottom Line
Catheter ablation did not significantly reduce the primary composite endpoint of death, disabling stroke, serious bleeding, or cardiac arrest vs drug therapy in the ITT analysis, largely due to lower-than-expected event rates and 27.5% crossover from drug therapy to ablation. Treatment-received and per-protocol analyses, as well as the significant reduction in death or CV hospitalization and AF recurrence, suggest a genuine clinical benefit of ablation when actually performed.
Major Points
- Largest randomized trial of catheter ablation vs drug therapy in AF; 2204 patients at 126 centers in 10 countries
- Median follow-up 48.5 months (approximately 4 years)
- Primary ITT result: 14% relative reduction in composite endpoint with ablation (HR 0.86) was not statistically significant (P=.30)
- Trial was underpowered due to lower-than-expected drug therapy event rates (drug group 3-year mortality ~4.1% vs projected 12%)
- Major confound: 27.5% of drug therapy patients crossed over to catheter ablation during follow-up
- Treatment-received analysis (ablation performed vs not): primary endpoint HR 0.67 (P=.006); mortality HR 0.60 (P=.005)
- Per-protocol analysis (6-month window): primary endpoint HR 0.74 (P=.056); 12-month window HR 0.73 (P=.046)
- Death or CV hospitalization significantly reduced with ablation: 51.7% vs 58.1% (HR 0.83; P=.001)
- AF recurrence significantly reduced: 49.9% vs 69.5% postblanking (HR 0.52; P<.001)
- No significant difference in serious bleeding between groups; disabling strokes were rare and directionally favored ablation
- Most common ablation adverse event: cardiac tamponade (0.8%); no atrial esophageal fistulas observed
- 19.4% of ablation patients required repeat ablation during postblanking period
- No prespecified subgroup showed a statistically significant differential treatment effect, though trends favored ablation in younger patients, those with heart failure, and minority patients
Study Design
- Study Type
- Investigator-initiated, open-label, multicenter, parallel-group randomized clinical trial
- Randomization
- Yes
- Blinding
- Open-label; no blinding of patients or investigators. Events independently adjudicated by a blinded clinical events committee using prospectively defined definitions. Hospitalization cause characterized by site investigator (unblinded). ECG recurrence reviewed by blinded ECG Core Laboratory Committee.
- Sample Size
- 2204
- Follow-up
- Through December 31, 2017; median 48.5 months (IQR 29.9-62.1 months)
- Centers
- 126
- Countries
- United States, Canada, Russia, Germany, Italy, United Kingdom, Czech Republic, China, Korea, Australia
Primary Outcome
Definition: Composite of death (all-cause), disabling stroke (Rankin Stroke Scale >=2), serious bleeding (hemodynamic compromise requiring surgery or >=3 units blood transfusion), or cardiac arrest; time to first event
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 101 events (9.2%); 4-year KM rate 8.9% | 89 events (8.0%); 4-year KM rate 7.2% | 0.86 (0.65-1.15) | .30 |
Limitations & Criticisms
- Open-label design; unblinded site adjudication of hospitalization cause introduces potential bias
- Severely underpowered: drug therapy group 3-year mortality was ~4.1% vs projected 12%, reducing ability to detect primary endpoint differences
- High crossover rate: 27.5% of drug therapy patients received catheter ablation, substantially biasing ITT toward null
- 9.2% of ablation-assigned patients did not receive ablation, further diluting ITT treatment effect
- Significance threshold was not adjusted for multiple secondary endpoint comparisons, increasing false-positive risk for secondary results
- AF recurrence data derived from only the subset using the study ECG recording system (n=1240/2204), limiting generalizability
- Ablation and drug therapies evolved over the 7-year enrollment period, potentially introducing temporal confounding
- 11% of patients received rate control drugs only (without rhythm control attempt), which may have affected drug arm outcomes
- Trial design was modified mid-study (February 2013): primary endpoint changed from all-cause mortality to composite; sample size reduced from 3000 to 2200; raises concerns about integrity of original hypothesis
- Generalizability limited: 92% White patients enrolled across predominantly Western countries
- Patients with prior ablation or failed >=2 antiarrhythmic drugs were excluded, limiting applicability to the broader AF population
Citation
Packer DL et al. Effect of Catheter Ablation vs Antiarrhythmic Drug Therapy on Mortality, Stroke, Bleeding, and Cardiac Arrest Among Patients With Atrial Fibrillation: The CABANA Randomized Clinical Trial. JAMA. 2019;321(13):1261-1274. doi:10.1001/jama.2019.0693