Managing Atrial Fibrillation in Heart Failure

Armen Hareyan's picture

When the number of authors on a paper approach the number of patients in one arm of a prospective, randomized trial... heads up.

I have just finished reading an article that appeared yesterday in the New England Journal of Medicine (Khan MN et al., "Pulmonary Vein Isolation for Atrial Fibrillation in Patients With Heart Failure" N Engl J Med 2008; 359: 1778-1785) regarding the PABA-CHF Trial (Pulmonary Vein Antrum Isolation versus AV Node Ablation with Biventricular Pacing for Treatment of Atrial Fibrillation in Patients With Congestive Heart Failure). It was an ambitious trial that compared atrial fibrillation ablation to biventricular pacing with AV nodal ablation for the treatment of heart failure. At least 29 investigators from 17 different centers randomized a mere 81 heart failure patients with weak heart muscles (ejection fractions <40%) into two treatment groups: 41 for atrial fibrillation ablation and 40 for AV nodal ablation and implantation of a biventricular pacemaker. The pre-specified endpoint was a composite one: improvement on a 6-minute hall walk, improvement on the Minnesota Living with Heart Failure Score, and improvement in ejection fraction. After showing remarkable results favoring ablation in terms of these three measures, the authors concluded that "pulmonary vein isolation was superior to atrioventricular-node ablation with biventricular pacing with heart failure who had drug-refractory atrial fibrillation."

Those are powerful words, especially when they are published in the New England Journal of Medicine.

Given that atrial fibrillation is the most common arrhythmia in man and occurs in probably 10% of the population over 80 years of age and heart failure, too, is an incredibly common disorder, we have to ask ourselves why so few patients were enrolled in this important trial? Was it tough to find appropriate patients willing to submit to the trial? Why were so many centers involved? Might the results have been skewed by sampling error or selection bias?


These are not small questions, especially for a trial studying such an economically and therapeutically important topic.

There are important distinctions between the two procedures studied that were conveniently ignored by the authors. First, atrial fibrillation often requires the use of an anesthesiologist and an impressive array of personnel and equipment (3D mapping systems, ablation equipment, etc) to perform. The average procedure time in many centers averages 4-5 hours and is labor-intensive.

Further, earlier studies reporting on the world-wide experience of atrial fibrillation ablation in a much larger group of over 8000 patients demonstrated that 27.3% of patients required more than one procedure to render patients "cured" of their atrial fibrillation and only 52% of patients were rendered free of antiarrhythmic drugs. Further, there was a 6% incidence of major complications from atrial fibrillation ablation when performed in a larger cohort of presumably healthier patients.

Not to say that pacing and AV junctional ablation isn't potentially expensive, too. It is. But usually only one procedure is required. Further, the use of a biventricular pacemaker in these patients with ejection fraction under 40%, rather than a defibrillator, is puzzling since several trials have demonstrated a survival advantage for biventricular defibrillator therapy over biventricular pacing therapy in similar heart failure patients (SCD-HeFT, COMPANION). Were these patient's not really as sick as the authors suggest? Further, was 6 months' follow-up really enough? Why were there no deaths in such a sick patient population in either arm of the trial? Are the data reported in the new trial representative of a more representative population of patients with atrial fibrillation in heart failure?

I worry about studies like this that fail to drill down on such important issues and are published front and center in journals with the impact factor of the New England Journal. While this trial might have profound implications that urge us to delve further into catheter ablation as a means to treat heart failure, to claim that catheter ablation of atrial fibrillation is "superior" to biventricular pacing and AV nodal ablation with such limited data and only 6-months of follow-up is overreaching and might miss important morbidity and mortality endpoints.

Reported by Dr. Wes from