Scroll down to participate in the poll

Gerasimos Filippatos (Athens, Greence) and Frank Ruschitzka (Zurich, Switzerland) chaired this Grand Debate about challenges in management of patients with heart failure (HF) and atrial fibrillation (AF). Frank Ruschitzka introduced the topic very briefly. The CABANA trial was presented last year, but this resulted in even more questions about what to do with devices, in who they should be used and in whom not. Ruschitzka asked the audience who was in favor of the statement ‘all patients with HF and AF should receive catheter ablation’, but only a few raised their hand, whereas far more did when was asked who was in favor of the opposite. Before introducing the debaters, Ruschitzka said that we should appreciate the warrior more than the sword, followed by a quote of Joseph Joubert (1754-1824): ‘The aim of argument, or of discussion, should not be victory, but progress’.

The first debater Douglas Packer (Rochester, MN, USA) had accepted to defend the PRO argument of the statement ‘All patients with HF and AF should receive catheter ablation’, but now heavily questioned whether it was possible to defend such a statement. Who had come up with this statement? And all patients? He said his opponent, prof Cleland, was absolutely correct and he conceded, and was at the point of stopping right there. He supported this by showing data of the DIAMANT trial, demonstrating that there was no difference in mortality up to 60 months in patients with AF and chronic HF between treatment for rate control and rhythm control (with the notion that the number of patients per group was very different). But then a study by Hsu et al. reported that ablation resulted in improvement of left ventricular ejection fraction (LVEF) and other markers of LV function in AF patients with congestive HF. And this improvement was observed in 42% of treated patients.

The CAMERA-MRI study showed large improvements in LA volume and LV stroke volume in those who underwent ablation vs those who received rate control in patients with AF and systolic dysfunction. In response to these results, Packer said that Cleland might be right. The CAMTAF trial was a randomized controlled trial in which patients underwent ablation or received medical treatment. It was a small trial with only 55 patients with a duration of only 6 months and little event data, but patients who underwent ablation had a significantly higher EF than those who received medical treatment. The AATAC trial was also a randomized controlled trial and demonstrated improved outcome of time to re-occurrence of AF in those who underwent ablation vs. the control group. Now Packer concluded that Cleland might be off a bit.

He continued by presenting the design and results of the CASTLE-AF trial. This investigator-initiated, prospective, multicenter, randomized controlled trial enrolled 397 patients, who underwent ablation (n=179) or conventional treatment (n=184) and were followed for 60 months. Patients in the ablation group had a 38% reduced risk for the primary outcome vs. the control group and all-cause mortality was reduced by 47%. He summarized the results thus far; control of AF in HF patients results in improvement of the AF burden, QoL, incidence of ventricular arrhythmias, LVEF, hospitalization and mortality, and noted that, therefore, Cleland is probably wrong.

Then he presented data from the CABANA trial, more specifically, outcomes of those with HF. Of 2204 AF patients, 778 (35.3%) patients had HF. The majority (79.0%) had EF ≥50%, 11.7% had EF ≥40-49% and 9.3% had EF <40%. The composite outcome of all-cause mortality, disabling stroke, serious bleeding or cardiac arrest was lower in those who underwent ablation vs. those who received drug therapy, with a 36% reduced risk. Outcome of all-cause mortality alone was also reduced in the ablation group vs. control group and risk was reduced by 43%. Furthermore, AF recurrence was reduced in the ablation group vs. those who received drug therapy.

He concluded that these data demonstrated a clinically important improvement in outcomes for HF patients who underwent ablation compared to those receiving drug therapy. He ended his presentation by saying that Cleland had it flat out backwards.

John Cleland (Glasgow, UK) defended the contra argument. He started his presentation by giving numbers of the prevalence (~25.000/million people) and incidence (>4.000/million people) of HF patients in the US nowadays. This would mean that in a small regional center of 2.5 million people, there would be ~60.000 prevalent cases of AF and ~10.000 incident cases. In a lab that could be operational 24/7 the whole year and do one case every 90 minutes, this would translate into the requirement of 2 labs to just do incident cases; it would not be feasible to perform ablation in all patients with AF and HF!

He continued by critically reviewing treatment goals. One of the treatment goals is improving symptoms, this can also be done with other treatments. We should think about the many possible tricks, before using expensive devices. One other treatment goal is the prevention of stroke, but stroke can also be prevented with anticoagulation, and after ablation anticoagulation is not stopped as it is unknown whether it is safe to stop anticoagulation.

One treatment goal of ablation is the enabling of treatments that only work for those with sinus rhythm, such as cardiac resynchronization therapy (CRT) and betablockers. The COMPANION trial demonstrated that benefits of CRT were lost in patients with AF. For betablockers, a meta-analysis showed that treatment with betablockers resulting in a heart rate <60 bpm was associated with improved outcomes in those in sinus rhythm, but this effect was less clear in those with AF. Cleland continued with dividing AF in HF patients into persistent/permanent AF and paroxysmal AF. Patients with persistent/permanent AF have a secure and stable rhythm and this does not precipitate hospitalization for AF, whereas paroxysmal AF is much more troublesome and might be a credible target for ablation. He emphasized that we should not make things worse by giving patients too much treatment, because then we might do harm and it might be too late to consider ablation. Things could get worse by aggressive rate control, toxic drugs, RV pacing or long periods in AF.

One should also consider the location of ablation, in the pulmonary vein or atrio-ventricular node, both with different advantages and disadvantages. The COMET trial showed that you can harm with right ventricular pacing, especially in the presence of a betablocker, as it resulted in worse outcomes in HF patients compared to non-pacing. Some studies have examined the effect of AV node ablation and showed no success in outcomes of mortality or improvement in EF. The CASTLE-AF trial demonstrated benefit for those who received ablation vs. those treated with medical therapy. But Cleland pointed out that these were differences between the 2 groups at baseline. Use of diuretics was higher in the medical therapy group, suggesting that the ablation group was less sick at the start. In addition, 10% withdrew from the study and another 10% were lost to follow-up. If you look at when mortality curves start to separate (after 36 months), there were less than 100 patients per group. So with this low number of patients per group, Cleland concludes that this was a tiny study.

He also discussed the CABANA trial and noticed the higher percentage of patients with amiodarone in the drug treatment group vs. the ablation group. Only 16% of patients had HF and 5% had LVEF ≤35%. 11% were lost to follow-up. The primary outcome of death, CVA, bleeding, VT/VF was not different between groups. With the example of Procrustes, who either chopped off legs of sleeping men with too long legs or stretched the legs of those with too short legs, he explained that we should tailor treatment. Finally, he listed some of the problems with ablation in HF patients, such as procedural complications (rare, but possibly devastating, such as cerebral infarction), relapse, atrial damage/fibrosis or possibly sudden death. He concluded that ablation is good for patients, but not for all, not for most, and probably not for many. Sometimes we must learn to interfere less.

In the rebuttal Packer said that most of the information shown by Cleland was ablation data and not about HF, the topic of this debate. For instance, Cleland spoke about CABANA and not CABANA HF. However, Packer agreed with the PRO argument in that not all patients with AF and HF should be ablated, not even most patients. Cleland, in his defense, said that unpublished work should be treated with suspicion and he cannot comment on work that has not been peer-reviewed. He ended by saying that in patients who have had AF for a substantial amount of time, unlinked to ADHF event, AF should not be aggressively managed, and the case for ablation in these patients is yet to be made.