DISCUSSION
Our present study is the first to evaluate long-term clinical outcomes after AF/AT ablation procedures guided by UHD mapping system, in patients with clinical HF, with either preserved or reduced LVEF. We showed that recurrence rates after complex ablations were not lower in HF patients than in non-HF patients. There was no difference in the mortality rates, whereas HF patients were more likely hospitalized for HF worsening during follow-up. Nevertheless, as expected, CA was associated with greater NYHA and LVEF improvements in HF patients compared to controls.
CA of complex atrial arrhythmias has been shown to be safe and able to improve prognosis of patients with HF, particularly in cases with HFrEF. Nevertheless, those patients are at high risk of recurrence, and reported rates varied from 27-73% after one procedure to 23-34% after repeated ablation on or off antiarrhythmic drug therapy.4,7–9,13,14. Data are more limited in HFpEF patients, but they also seem to benefit from CA, despite high recurrence rates.7,8,15 HF has been identified as an independent risk factor for AF/AT recurrence in several reports that compared CA outcomes in patients with or without HF.7,9,14 These lower success rates in HF could be related to different mechanisms: atrial enlargement and structural remodelling, due to chronic high filling-pressures and/or mitral regurgitation, that favours perpetuation of AF and a higher proportion of persistent AF; ischemia or the cardiomyopathy itself that can also alter atrial myocardium; but also patients’ frailty, that can prevent from long-lasting procedures’ completion. There were only few studies that directly compared outcomes of CA of AF between HF and control patients. Chen et al reported a 27% rate of AF recurrence in patients with systolic dysfunction after PVI achieved without 3D mapping, whereas patients with normal cardiac function had only 13% of recurrence (p=0.03).9 Using a standard 3D mapping system, Cha et al reported respectively 38%, 25% and 16% one-year recurrence in patients, whether they had systolic dysfunction, diastolic dysfunction, or normal cardiac function.7 Furthermore, success rates of AF ablation were previously reported lower in patients with LVEF<50% compared to patients with LVEF>50%.13Black-Maier et al. recently published outcomes of CA of AF in both HFrEF and HFpEF patients with respectively 32.6% and 33.9% of recurrence (p=NS), but lower rates of repeated ablations than in our study.15 Our present one-year recurrence rates were consistent with previous reports. Moreover, HF status was not a predictive factor of AF/AT recurrence, as well as the alteration of LVEF among HF patients. Yang et al. reported that patients with previous CA of AF would rather have AT than AF recurrence, depending on the degree of atrial remodelling. Patients with more dilated atria and lower left atrial bipolar voltage were likely to have AF recurrence suggesting that HF patients should preferably relapse in AF.16 It is important to note that ATs were the most frequent arrhythmia in our study population in both groups. AF and AT share the same precipitants and are often studied as interchangeable diagnosis, even an important proportion of patients with AF would not experience AT and vice-versa. In the particular setting of HF patients, the overlap between AF and AT is important and should justify to study these atrial arrhythmias together.17 In our study, the high proportion of AT was consistent with the high proportion of repeated procedures as previous ablation can lead to additional atrial scar and subsequent complex ATs. In our study, the type of arrhythmia was not predictive of recurrence. The UHD mapping system we used was already reported to be able to improve comprehension and ablation success of complex post-AF ablation ATs.18,19
In our study, we identified mitral regurgitation, hypertrophic cardiomyopathy and persistent AF as predictive factors of recurrences. A multicentre registry also highlighted higher recurrence rates in HF patients with persistent AF compared to controls, whereas the results of CA for paroxysmal AF were not different between HF patients and controls.14 Data about CA in this setting are scarce, but a systematic review has already reported higher recurrence rates in patients with hypertrophic cardiomyopathy compared to controls and with mitral regurgitation that amplifies and aggravates the atrial remodelling.20
The other important result of our present work is the beneficial effect of CA on both NYHA class and LVEF in patients with HF. In patients with HFrEF, CA of persistent AF, using a 3D mapping system, was already reported to improve LVEF, peak oxygen consumption, and Minnesota living with HF questionnaire score.4 Another study reported better improvement in 6 minutes walking distance, Minnesota score and LVEF, after PVI compared to atrioventricular node ablation combined with cardiac resynchronization.3 Moreover, the CASTLE-AF trial, reported a decrease in a composite endpoint of mortality and hospitalization for HF in the CA group compared to control group, in patients with HFrEF.6 Another study conducted in patients with HFpEF, showed that only patients who maintained sinus rhythm had improvement in echographic parameters.8 We also reported here, that AF/AT recurrence was associated with lower improvement in functional status and LVEF. Black-Maier et al reported similar effect of CA on AF recurrence and NYHA improvement in both HFrEF and HFpEF groups, as we did, but they did not include AT in their analysis. They also showed a trend towards a greater improvement in NYHA class in patients with HFpEF that nearly reached statistical significance.15 Likewise, our patients with HFrEF did not improve their NYHA class more than HFpEF patients, despite a significant larger increase in LVEF.
Despite longer procedures with longer mapping duration in the HF group, there was no difference in fluid intake and no acute HF, even in patients with reduced LVEF. Our present complications rates were similar with those previously reported in HF patients or with this particular mapping system. The TRUE-HD study, the largest prospective study assessing outcomes of this novel UHD mapping system, reported a 4% complications rate. 21 We acknowledge that our patients did not present severely depressed LVEF, as the median LVEF was 45% in the HF group. Nevertheless, they had associated comorbidities and were more fragile than the non-HF patients. Acute HF events were not rare in patients with HFpEF (3.8%) and HFrEF (6.2%) after CA of AF in a previous report using the same LVEF cut-off values.15