論文ID: CJ-20-0191
Background: The purpose of this study was to elucidate the effect of the temporal relationship between atrial fibrillation (AF) and heart failure (HF) on clinical outcomes after catheter ablation.
Methods and Results: We included 129 consecutive patients with AF and HF who underwent catheter ablation in hospital from December 2014 to September 2017. The patients were divided into 2 groups based on the temporal relationship between AF and HF. Group 1 consisted of 42 patients with AF following HF while Group 2 consisted of 87 patients with AF preceding HF or those who developed both of them simultaneously at the timing of first visit to a doctor. The primary endpoint was a composite of death and hospitalization due to HF during a 2-year follow-up. AF recurrence was more common in Group 1 (45% vs. 23%; hazard ratio [HR], 2.49; 95% confidence interval [CI], 1.25–4.94; P=0.009). Death and HF hospitalization were more frequent in Group 1 (19 [45%], 6 [7%] patients, respectively, P<0.0001). After adjustment for several covariates, patients in Group 1 were independently associated with poorer outcomes after AF ablation (HR, 8.66; 95% CI, 2.942–5.5; P<0.0001).
Conclusions: Adverse clinical outcomes of death, HF hospitalization and AF recurrence were more frequent in patients with AF following HF than in those with AF preceding HF.
Heart failure (HF) and atrial fibrillation (AF) often coexist and each complicates the course of the other.1,2 AF is independently associated with all-cause death.3 Newly developed AF in patients with preexisting HF has a worse effect on mortality than preexisting AF in HF patients.4,5
Editorial p ????
With regard to treatment, rhythm control therapy using antiarrhythmic drugs has failed to demonstrate any further benefit on mortality over rate control therapy in patients with AF and HF.6 Catheter ablation of AF has evolved over the past 20 years to become a commonly performed procedure, and has demonstrated better rhythm outcomes than conservative antiarrhythmic drug therapy.7 In addition to the many observational studies demonstrating the safety and efficacy of AF ablation in this patient population,8 several randomized controlled trials have also shown the value of maintaining sinus rhythm and its beneficial effect on mortality among HF patients.9,10 Based on these findings, an international consensus group on AF ablation suggested that it is reasonable to use similar indications for AF ablation in selected patients with HF as in patients without HF (Class IIa, level of evidence B-R).11
In the clinical setting, however, patients with both AF and HF do not always benefit from catheter ablation. Here, we hypothesized that the benefits of catheter ablation would greatly depend on causal or temporal relationships between AF and HF. For example, catheter ablation vastly improves clinical outcomes when AF causes HF, such as in tachycardia-induced cardiomyopathy, because elimination of AF leads to the elimination of HF. On the other hand, the beneficial effect of catheter ablation may be limited if AF develops as a result of preceding HF, because HF and advancement of atrial remodeling will continue even after catheter ablation has been performed.
The purpose of this study was to compare clinical outcomes after catheter ablation in patients who developed AF following HF with those in whom AF preceded HF.
This prospective observational study was undertaken at Kansai Rosai Hospital between December 2014 and September 2017. The participants were 129 consecutive patients with a history of HF requiring any therapeutic intervention and who were planned to undergo catheter ablation of AF. The diagnosis of HF was based on the Framingham Heart Failure Diagnostic Criteria. The patients were divided into 2 groups based on the temporal relationship between AF and HF, with Group 1 consisting of patients with AF following HF, and Group 2 of the remaining patients including those with AF preceding HF or who developed both of them simultaneously at the time of first visit to a doctor. We used the ambiguous definition of temporal relationship in Group 2, because AF and HF symptoms often overlap, making it difficult and arbitrary to determine the specific timing of HF development in AF patients.
Exclusion criteria included age <20 years, unstable hemodynamic condition requiring intravenous drugs, acute coronary syndrome, previous cardiac surgery, and previous catheter ablation. This study complied with the Declaration of Helsinki. Written informed consent for ablation and participation in the study was given by all patients, and the protocol was approved by the institutional review board.
Radiofrequency Catheter Ablation ProcedureElectrophysiological studies and catheter ablation were performed under intravenous sedation with dexmedetomidine. A 6-Fr decapolar electrode was inserted into the coronary sinus while a second 6-Fr decapolar electrode was placed in the right atrium. Following a transseptal puncture at the fossa ovalis, 2 long sheaths were introduced into the left atrium. A 20-pole circular catheter was placed in a pulmonary vein via the sheath. The operators then performed mapping and ablation under the guidance of an electroanatomical mapping system (Carto3®; Biosense Webster, Diamond Bar CA, USA).
A dragging technique was used to perform circumferential ablation around both ipsilateral pulmonary veins using an open-irrigated ablation catheter with a 3.5-mm tip (Thermocool SmartTouch®; Biosense Webster) via an Agilis® or SL0® sheath (St. Jude Medical, St. Paul MN, USA). Radiofrequency energy was applied for 30 s (15 s at the posterior left atrial [LA] wall near the esophagus) at each site using a maximum temperature of 42℃ and maximum power of 35 W. The irrigation rate was 17 mL/min. Operators attempted to maintain an appropriate contact force between the catheter and endocardium of between 5 and 20 g. Pulmonary vein isolation (PVI) was considered complete when both entrance and exit blocks were created. If atrial flutter was observed spontaneously or induced by atrial burst stimuli, additional ablation was performed. During sinus rhythm, isoproterenol was infused at 5, 10, and 20 µg/min at 2-min intervals and discontinued to provoke AF-trigger ectopies and AF. Attempts were made to ablate the AF-trigger ectopies that triggered AF. If AF persisted or was still inducible after these procedures, ablation targeting areas with low-voltage (bipolar peak-to-peak voltage <0.50 mV) and/or those with complex fractionated electrograms were additionally conducted at the discretion of the attending operators.
Follow-upPatients were followed up every 4–8 weeks at a dedicated arrhythmia clinic for 2 years. Clinical events of death, worsening HF requiring hospitalization, and AF recurrence were monitored.
Routine ECGs were obtained at each outpatient visit, and 24-h ambulatory Holter monitoring was performed at 6 and 12 months post-ablation. When patients experienced symptoms suggestive of an arrhythmia, a surface ECG, ambulatory ECG, and/or cardiac event recording was also obtained. Either of the following events after the initial 3 months from the ablation (known as the blanking period) was considered to indicate AF recurrence: (1) atrial tachyarrhythmia recorded on a routine or symptom-triggered ECG during an outpatient visit or (2) atrial tachyarrhythmia of at least 30 s duration on ambulatory ECG monitoring.
Echocardiography was performed before and 6 months after the ablation. Left ventricular ejection fraction (LVEF) and LA volume were measured using the Simpson method with 4- and 2-chamber views.
Antiarrhythmic drug usage was at the discretion of the attending physician.
Statistical AnalysisContinuous data are expressed as the mean±standard deviation or median (1st, and 3rd quartiles). Categorical data are presented as absolute values and percentages. Tests for significance were conducted using the unpaired t-test or nonparametric test (Mann-Whiney U-test) for continuous variables, and the chi-squared test or Fisher’s exact test for categorical variables. Univariate and multivariate Cox proportional hazards models were used to determine the clinical factors associated with the clinical events of death, HF hospitalization, and AF recurrence. Variables with a P-value ≤0.05 in the univariate models were included in the multivariate analysis. Clinical event-free survival rates were calculated using the Kaplan-Meier method. Survival curves of the 2 groups were compared with a 2-sided Mantel-Haenszel (log-rank) test. All analyses were performed using commercial software (SPSS version 26.0®, SPSS, Inc., Chicago IL, USA).
Patients in Group 1 were more likely to be male, have paroxysmal AF, and have HF of ischemic etiology (Table 1). HF was also more severe, as indicated by higher B-type natriuretic peptide levels and a high New York Heart Association functional class. Left ventricular remodeling was more advanced, with dilation, reduced contraction, and reduced relaxation. LA size was comparable between the 2 groups.
| Group 1 (n=42) |
Group 2 (n=87) |
P value | |
|---|---|---|---|
| Age (years) | 68±10 | 69±10 | 0.66 |
| Male | 32 (76%) | 48 (55%) | 0.021 |
| Body mass index (kg/m2) | 23.5±4.4 | 23.7±4.5 | 0.87 |
| CHA2DS2-VASc score | 3.3±1.3 | 3.4±1.4 | 0.61 |
| Paroxysmal AF | 23 (55%) | 15 (17%) | <0.0001 |
| Hypertension | 14 (33%) | 37 (43%) | 0.32 |
| Diabetes mellitus | 10 (24%) | 12 (14%) | 0.16 |
| Stroke or TIA | 3 (7%) | 10 (11%) | 0.33 |
| Creatinine clearance (mL/min) | 57.8±27.6 | 63.4±31.1 | 0.32 |
| B-type natriuretic peptide (pg/dL) | 468 (176, 911) | 286 (147, 433) | 0.005 |
| NYHA Class III or IV | 10 (24%) | 9 (10%) | 0.042 |
| HF etiology | <0.0001 | ||
| Ischemic | 17 (40%) | 1 (1%) | |
| Non-ischemic | 25 (60%) | 86 (99%) | |
| Medications | |||
| Amiodarone | 8 (19%) | 5 (6%) | 0.017 |
| ACEI/ARB | 20 (49%) | 31 (36%) | 0.18 |
| β-blocker | 33 (79%) | 54 (62%) | 0.073 |
| Furosemide | 26 (62%) | 39 (45%) | <0.001 |
| Spironolactone | 2 (5%) | 3 (3%) | 0.66 |
| Echocardiography | |||
| LV diastolic diameter (mm) | 56±11 | 48±6 | <0.001 |
| LV mass (g) | 210±73 | 156±44 | <0.001 |
| LVEF (%) | 41±15 | 53±14 | <0.001 |
| <40% | 22 (52%) | 17 (20%) | <0.001 |
| LA diameter (mm) | 44±8 | 44±7 | 0.84 |
| LA volume (cm3) | 80±33 | 78±36 | 0.78 |
| LA volume index (cm3/m2) | 48±20 | 49±22 | 0.95 |
| e’ (cm/s) | 6.1±2.3 | 7.4±2.4 | 0.008 |
| E/e’ | 13.1±4.9 | 11.7±4.7 | 0.14 |
| Moderate or severe MR | 16 (39%) | 19 (22%) | 0.042 |
ACEI, angiotensin-converting-enzyme inhibitor; AF, atrial fibrillation; ARB, angiotensin II receptor blocker; e’, diastolic early mitral annular velocity; E, diastolic early transmitral flow velocity; EF, ejection fraction; HF, heart failure; LA, left atrial; LV, left ventricular; MR, mitral regurgitation; NYHA, New York Heart Association.
Procedural and fluoroscopic times were similar between groups. PVI was completed in all patients, and other sites of ablation were closely similar between groups (Table 2).
| Group 1 (n=42) |
Group 2 (n=87) |
P value | |
|---|---|---|---|
| Procedure time (min) | 107±45 | 106±28 | 0.882 |
| Fluoroscopy time (min) | 23±9 | 22±8 | 0.264 |
| Site of ablation | |||
| PVI | 42 (100%) | 87 (100%) | 1.000 |
| Isolation of superior vena cava | 1 (2%) | 1 (1%) | 0.596 |
| Cavotricuspid isthmus linear ablation | 8 (19%) | 13 (15%) | 0.554 |
| Septal linear ablation | 2 (5%) | 6 (7%) | 0.638 |
| Posterior isolation | 0 (0%) | 3 (3%) | 0.223 |
| Roof linear ablation | 1 (2%) | 6 (7%) | 0.289 |
| Low-voltage area*-guided ablation | 5 (12%) | 11 (13%) | 0.905 |
| Complex fractionated atrial electrogram ablation | 2 (5%) | 4 (5%) | 0.967 |
| Adverse events within 30 days of ablation | 3 (7%) | 2 (2%) | 0.33 |
| HF worsening | 3 (7%) | 1 (1%) | |
| Intestinal thromboembolism | 0 (0%) | 1 (1%) | |
| Bleeding necessitating transfusion | 0 (0%) | 0 (0%) | |
| Cardiac tamponade | 0 (0%) | 0 (0%) | |
| Esophageal injury | 0 (0%) | 0 (0%) | |
| Cerebral infarction/transient ischemic attack | 0 (0%) | 0 (0%) | |
| Death | 0 (0%) | 0 (0%) | |
*Defined as areas with bipolar peak-to-peak voltage <0.5 mV during sinus rhythm. HF, heart failure; PVI, pulmonary vein isolation; RFCA, radiofrequency catheter ablation.
No fatal procedure-related complications occurred. Worsening HF necessitating in-hospital treatment within 1 month after the ablation procedure occurred in 3 patients (7%) from Group 1 and 1 patient (1%) from Group 2. HF worsened following early recurrence of AF after hospital discharge in 3 patients, and during hospitalization on the day after the ablation procedure in 1 patient. All 4 patients eventually recovered. Intestinal thromboembolism was observed in 1 patient from Group 2, occurring 27 days after the ablation procedure while taking optimal anticoagulation therapy. The patient underwent conservative treatment and recovered.
Long-Term OutcomesAll patients were followed up until 24 months or death, except for 1 patient who was lost to follow-up 12 months after the ablation procedure. More patients in Group 1 died than in Group 2 (8 [19%] vs. 2 [2%], P=0.002, Figure 1A). There were 5 cardiac-related deaths in Group 1 (2 HF, 2 sudden deaths suggestive of ventricular tachyarrhythmias, and 1 infective endocarditis) and 3 non-cardiac deaths (strangulation ileus, tongue cancer, and 1 unknown cause). In Group 2, there was 1 cardiac-related death (sudden death suggestive of ventricular tachyarrhythmia) and 1 non-cardiac death (liver cirrhosis). HF hospitalization was more frequent in Group 1 than in Group 2 (16 [38%] vs. 4 [5%] patients, P<0.0001, Figure 1B). Death and HF hospitalization were more frequent in Group 1 (19 [45%], 6 [7%], respectively, P<0.0001, Figure 1C).
Event-free survival curves. Kaplan-Meier curves showing freedom from death (A), HF hospitalization (B), composite of death and HF (C), and composite of death and HF stratified according to the presence and absence of AF recurrence (AFR) (D). Clinical events of death, HF, and a composite of both were more frequent in Group 1. In Group 1, patients with AFR had poorer outcomes than those without recurrence. In contrast, AFR did not affect the Kaplan-Meier curves in Group 2. AF, atrial fibrillation; HF, heart failure.
Mean number of ablation procedures was 1.2±0.5 in Group 1 and 1.2±0.4 in Group 2 (P=0.94). Antiarrhythmic drugs were comparably prescribed between the 2 groups (8 [19%] vs. 9 [10%] patients, P=0.17). AF recurrence-free survival rates after the initial and multiple procedures were significantly higher in Group 2, irrespective of antiarrhythmic drug use (Figure 2). AF recurrence was more frequently observed in Group 1 than in Group 2 even after adjustment for LA volume index (Supplementary Table 1). In Group 1, patients with AF recurrence had a lower HF hospitalization-free survival rate than those without recurrence (Figure 1D). In contrast, AF recurrence did not affect the Kaplan-Meier curves in Group 2 (Figure 1D).
(A–D) Kaplan-Meier curves for AF-recurrence-free survival. AF recurrence after initial and multiple procedures was significantly higher in Group 2, irrespective of the use of AADs. AADs, antiarrhythmic drugs; AF, atrial fibrillation.
In Group 1, the unadjusted hazard ratio for HF hospitalization and death for off-drug AF recurrence after the initial ablation procedure was 2.17 (95% confidence interval [CI], 1.16–4.07, P=0.016). The hazard ratios were 2.49 (95% CI, 1.25–4.94, P=0.009) after adjustment for the CHA2DS2-VASc score and AF type, and 4.37 (95% CI, 1.33–14.4, P<0.015) after adjustment for incorporated factors shown in Table 3 except for allocation to Group 1.
| Variable | Death or HF hospitalization | Univariate analysis | Multivariate analysis | |||
|---|---|---|---|---|---|---|
| With (n=25) |
Without (n=104) |
OR (95% CI) |
P value | OR (95% CI) |
P value | |
| Age (years) | 72±10 | 68±10 | 1.05 (0.99–1.08) | 0.17 | ||
| Male | 13 (52%) | 66 (64%) | 1.52 (0.69–3.32) | 0.30 | ||
| Body mass index (kg/m2) | 21.9±3.8 | 24.0±4.5 | 0.87 (0.78–0.97) | 0.014 | 0.86 (0.74–1.002) | 0.053 |
| Non-paroxysmal AF | 13 (52%) | 78 (75%) | 0.42 (0.19–0.91) | 0.029 | 0.71 (0.22–2.34) | 0.82 |
| NYHA Class ≥III | 9 (36%) | 10 (10%) | 4.59 (2.02–10.4) | <0.0001 | 1.76 (0.46–6.69) | 0.41 |
| LVEF (%) | 43±18 | 50±15 | 0.97 (0.94–1.00) | 0.035 | 1.02 (0.98–1.06) | 0.46 |
| Left atrial volume index (cm3/m2) | 56±22 | 46±21 | 1.02 (1.001–1.04) | 0.044 | 1.02 (0.99–1.05) | 0.067 |
| MR moderate or severe | 12 (48%) | 23 (22%) | 3.52 (1.40–8.88) | 0.006 | 1.94 (0.61–6.09) | 0.26 |
| Group 1 | 18 (72%) | 24 (23%) | 8.28 (3.30–20.8) | <0.0001 | 9.42 (2.55–34.7) | <0.0001 |
*Cox proportional hazard model incorporating factors with P<0.05 in the univariate analyses. CI, confidence interval; OR, odds ratio. Other abbreviations as in Table 1.
Clinical factors associated with the composite endpoint of death and HF hospitalization were low body mass index, paroxysmal AF, severe HF with a New York Heart Association functional class ≥III, large LA volume index, and classification to Group 1 (Table 3). Multivariate analysis incorporating these factors revealed that classification to Group 1was independently associated with the composite endpoint.
In Group 1, a subgroup of patients with AF recurrence was independently associated with the composite endpoint with a hazard ratio of 4.37 and 95% CI of 1.33–14.4 (P<0.015) after adjusting for incorporated factors shown in Table 3 except for allocation to Group 1. Preprocedural factors associated with the composite endpoint in Group 1 are shown in Supplementary Table 2. Only low body mass index independently predicted the composite endpoint with a hazard ratio of 0.79 for 1 kg/m2 increase, 95% CI of 0.64–0.99 (P=0.037).
Changes in Cardiac FunctionGroup 2 showed a greater increase in LVEF 6 months after catheter ablation (Figure 3). LV diastolic diameter did not change in either group. The LA volume index decreased after catheter ablation in Group 2, but did not change in Group 1.
Cardiac function before and 6 months after ablation. LVEF before and after the ablation procedure (A) and the difference (B) were compared between the 2 groups. LVEF significantly increased in both groups, but the increase was higher in Group 2. LVDd before and after the ablation procedure (C) and the difference (D) were compared between the 2 groups. LVDd did not change after compared with before the ablation procedure. LAV before and after the ablation procedure (E) and the difference (F) were compared between the 2 groups. LAV decreased significantly in Group 2, but not in Group 1. LAV, left atrial volume; LVDd, left ventricular diastolic dimension; LVEF, left ventricular ejection fraction.
This observational study compared the clinical outcomes after catheter ablation between patients with AF following HF and those with AF preceding HF. The main findings were: (1) catheter ablation of AF in HF patients was safe without any fatal complications; (2) the composite endpoint of death and HF hospitalization during the 2-year follow-up period occurred more frequently in Group 1 than in Group 2; (3) AF recurrence was more frequent in Group 1; and (4) the temporal relationship between AF and HF independently affected clinical outcomes even after adjustment for covariates.
To our knowledge, this is the first report to demonstrate worse clinical outcomes after AF ablation in patients with AF following HF than in those in whom AF precedes HF.
Effect of AF/HF Temporal Relationships on Rhythm Outcomes After PVIAF recurrence rate was significantly higher in Group 1, even though Group 1 included more patients with paroxysmal AF than Group 2. The poorer outcomes related to rhythm after catheter ablation indicated that patients in Group 1 had extensive extrapulmonary vein arrhythmogenic substrate. The atria in Group 1 were estimated to have been exposed to pressure overload and neurohormonal abnormality arising from reduced ventricular function over a period of time, likely resulting in arrhythmogenic substrate formation such as fibrosis. Prior studies support this hypothesis, having demonstrated that a high LA pressure is associated with extensive extrapulmonary vein low-voltage areas in patients undergoing AF ablation.12,13 Conversely, AF probably developed without preceding HF in Group 2, suggesting that PVI would be effective, presumably because of less advanced atrial remodeling. In such cases, AF can be successfully treated by PVI.
Taking the above into consideration, more extensive atrial ablation in addition to PVI may be needed to improve rhythm outcomes in Group 1 type patients. However, in this study, the attending operators did not often perform additional ablation in either of the 2 groups, possibly due to poor patient condition such as reduced cardiac function.
Paradoxically, paroxysmal type of AF was more frequent in Group 1 than in Group 2. AF episodes are supposed to have more negative hemodynamic effect in patients with preceding depressed ventricular function (Group 1) than in those without (Group 2), possibly prompting the patient with paroxysmal AF to visit hospital before there is progression to persistent AF.
Effect of AF/HF Temporal Relationships on Death and HF HospitalizationHeterogeneous causal relationships between AF and HF lead to differing prognoses after AF ablation. The 2-year composite endpoint of death and HF hospitalization was more frequent in Group 1 than in Group 2 in this study. Previous reports investigating patients with coexisting AF and HF have demonstrated poorer clinical outcomes for patients with AF following HF compared with those showing AF preceding HF.4,14 AF following HF is likely to be caused by persistent atrial pressure overload and abnormal neurohormonal conditions as a result of reduced ventricular function. Thus AF following HF would indicate HF progression and a poor clinical outcome. Indeed, in our present study, more severe HF parameters were observed in Group 1 (Table 1).
The essential pathogenesis of AF preceding HF is a purely atrial electrophysiological event. HF symptoms and reduced cardiac function are mainly attributable to rapid and irregular ventricular responses and loss of atrial contraction, known as tachycardia-induced cardiomyopathy.15 Optimal management of AF can preserve cardiac function and improve clinical outcomes in this situation.16 In addition to a worse clinical course when AF follows HF, the lower efficacy of catheter ablation may enhance the poorer clinical outcomes in this population.
The beneficial effect of AF ablation on HF prognosis depends greatly on the causal relationship between AF and HF. The limited benefit of AF ablation in Group 1 can be explained by the different etiology of HF in the 2 groups. Although the HF in Group 1 was unlikely to be caused by AF, the HF in Group 2 was most probably caused by the preceding AF.
Importantly, the adverse prognostic effect of AF recurrence was observed only in Group 1. The negative effect of recurrent AF on hemodynamics would be greater in Group 1 with poorer ventricular function than in Group 2. In addition, in Group 1, recurrent AF after PVI indicated the presence of extra-PV atrial arrhythmogenic substrate caused by elevated atrial pressure and increased neurohormonal activity and might work as an indicator of depressed ventricular function.
Reverse Remodeling of LV FunctionThe different etiologies of HF may explain the significant improvement in LVEF in Group 2. LV dysfunction arising from tachycardia-induced cardiomyopathy is known to show greater improvement after sinus conversion than most other cardiomyopathies, such as ischemic and dilated cardiomyopathies.16 Less myocardial fibrotic change is observed in tachycardia-induced myopathy compared with other cardiomyopathies and would explain the improved LV reverse remodeling.17 A negative correlation between the extent of LV gadolinium enhancement on cardiac magnetic resonance imaging and the amount of improvement in LVEF after AF ablation was clearly demonstrated in the CAMERA-MRI study.18
Clinical ImplicationsTwo randomized controlled trials, CASTLE AF10 and AATAC,9 highlighted the benefits of catheter ablation of AF among HF patients. However, in the clinical setting, it is important to recognize that the efficacy of catheter ablation varies, and indications should therefore be determined carefully by considering each patient’s situation. In particular, the temporal relationship between AF and HF warrants careful review because this is one of the most influential factors in the efficacy of ablation.
Study LimitationsSeveral limitations of this study warrant mention. First, this was a non-randomized obsevational study. Patient selection for catheter ablation and patient management may have introduced bias. Second, classification of the temporal relationship between AF and HF was occasionally difficult because the time from AF onset and subsequent HF worsening was sometimes very close, hampering differentiation of the order of development. Therefore, some patients in Group 2 seemed to present with AF and HF simultaneously and the temporal relationship was not precise. Third, AF recurrence after discharge was quantified on the basis of intermittent short-time ECG and the patients’ symptom-triggered ECG, giving rise to the possibility that asymptomatic episodes of AF might have been missed. Fourth, echocardiographic data at 6 months post-ablation were missing for some patinets. Finally, the small sample size may limit the statistical accuracy of the results. Multicenter prospective trials including a larger number of patients are warranted.
Adverse clinical outcomes of death, HF hospitalization and AF recurrence were more frequent in patients with AF following HF than in those with AF preceding HF. The indication for catheter ablation should be determined carefully, with consideration for the temporal relationship between AF and HF.
Concept, design, drafting the manuscript: A. Tsuji, M. Masuda
Data collection and revising the manuscript: A. Tsuji, M. Masuda, M. Asai, O. Iida, S. Okamoto, T. Ishihara, K. Nanto, T. Kanda, T. Tsujimura, Y. Matsuda
Final approval of the manuscript: T. Mano
None.
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-20-0191
