Catheter ablation for persistent atrial fibrillation after acute decompensated heart failure Attack: Earlier is Better?

Abstract

Background

Acute decompensated heart failure (ADHF) is often accompanied by persistent atrial fibrillation (AF). However, the optimal timing for RFCA in patients with persistent AF and ADHF is still uncertain.

Objectives

The aim of this observational cohort study is to investigate the safety and efficacy of early RFCA in patients with persistent AF after ADHF attack.

Methods

Patients with persistent AF and ADHF who underwent early RFCA as soon as the ADHF symptoms were initially controlled (Early group, n = 63) and those who received elective procedures after a transitional period (Elective group, n = 67) were investigated. After 1:1 propensity score matching, 50 matched pairs were analyzed.

Results

The overall procedural complication rates were similar (Early group: 6.0 %, n = 3; Elective group: 6.0 %, n = 3; P = 1.000). Patients in the early group had significantly less HF rehospitalization than the elective group during the 1-year post-procedure follow-up period (Mantel-Cox test: P = 0.036; HR: 0.369; 95 %CI: 0.145–0.938), though AF recurrence showed no difference (Mantel-Cox test: P = 0.645; HR: 1.204; 95 %CI: 0.547–2.648). A 90-day rehospitalization rate was significantly higher in the transitional period in the elective group, compared with patients who already received early RFCA (Elective group: 13, 26.0 %; Early group: 2, 4.0 %; P = 0.002).

Conclusions

Early RFCA therapy for persistent AF after ADHF attack was safe and effective. Patients who received early RFCA therapy had significantly less HF rehospitalization in the 1-year post-procedure follow-up period. On the other hand, the elective procedure was accompanied by a higher risk of HF rehospitalization during the waiting period.

1. Introduction

Acute decompensated heart failure (ADHF) is a syndrome caused by a short-term progressive cardiac dysfunction and characterized by exacerbating fatigue and dyspnea, orthopnea, peripheral edema, polyserous cavity effusion and/or pulmonary edema caused by fluid accumulation, overload, and redistribution, requiring urgent hospitalization [1], [2]. ADHF is the most common form of acute heart failure (AHF), frequently accompanied by tachycardia and persistent atrial fibrillation (AF) [3]. Radiofrequency catheter ablation (RFCA), a non-pharmacological treatment that manipulates cardiac rhythm, has been proven to be beneficial in patients with persistent AF and HF, even in patients end-stage HF [4], [5], [6]. Furthermore, in studies like EAST-AFNET, early rhythm control resulted in better prognosis in patients with AF [7], [8].

However, manifestations and responses are different in patients with ADHF. The initial 90 days after ADHF are considered a vulnerable period. The rehospitalization rate is 18.2 % within 30 days and 31.2 % within 90 days, as reported by Khan et al., and high mortality is observed during this period, which might be attributed to inadequate decongestion and substantial subclinical hemodynamic abnormalities after hospital discharge [2], [9], [10], [11]. The efficacy and safety of any non-pharmacological procedures used to manage ADHF are of concern. The optimal time for the safe use of RFCA in patients with persistent AF and ADHF is still to be explored since the elective RFCA after anti-HF medication and temporary rhythm control by extracorporeal cardioversion and antiarrhythmic drugs was effective in symptomatic HF with AF [12].

In this cohort study, we aimed to compare the safety and efficacy of the early RFCA in patients with persistent AF after ADHF attack.

2. Method

2.1. Study population

One hundred and fifty-nine inpatients between November 2020 and November 2022 diagnosed with persistent AF (continuous AF for > 7 days, including long-standing persistent AF) and complicated with ADHF received RFCA therapy in Xinhua Hospital.

All the patients received continuous ECG monitoring with Bedside ECG monitor or portable device since admission, and routinely underwent 24 h-Holter assessment before procedure. All the enrolled patients presented a 100 % AF burden under 24 h-Holter, and persistent AF rhythm was confirmed with ECG monitoring. Sixty-five of them received immediate RFCA therapy during hospitalization, when the ADHF symptoms were initially controlled and eligible to tolerate the procedure. Excluding 2 patients who did not attend post-procedure examination in the follow-up visit, the remaining 63 patients were fully tracked and categorized in the early group.

Another 89 patients from the whole group were discharged after controlling ADHF symptoms initially and readmitted for further RFCA treatment after a transitional period. Among them, 70 patients met the following criteria: (1) patients with no less than 90 days of transitional period; (2) patients who had standardized anti-HF medication and rhythm control therapy during the transitional period; (3) HF rehospitalization during the transitional period was acceptable, but the RFCA procedure and the most recent HF rehospitalization were within no less than 90 days; (4) patients who had no ADHF and stable hemodynamic status when admitted for the elective procedure. After 3 patients were excluded for the absence of post-procedure follow-up visit, 67 patients were fully tracked and included in the elective group.

Patients meeting all the following four criteria were considered ADHF patients: (1) patients who had exacerbating fatigue and dyspnea, or orthopnea within recent 14 days；(2) patients with peripheral edema (on physical examination), polyserous cavity effusion or pulmonary edema (on physical examination or imaging); (3) NT-proBNP of ≥ 660 pg/mL and（4）patients with NYHA III or IV heart failure classification [1], [13], [14], [15]. This study was approved by the ethics board of Xinhua Hospital and complied with the Declaration of Helsinki Principles. The study design is illustrated in Fig. 1.

Fig. 1.

Study design. AF = atrial fibrillation; AFL = atrial flutter; ADHF = acute decompensated heart failure; RFCA = radiofrequency catheter ablation.

2.2. Treatment

2.2.1. Medical therapies for ADHF management

The patients were systemically managed immediately after admission. Supplemental oxygen and ventilatory support, continuous cardiac monitoring, prompt diuretic therapy, early vasodilator therapy, and urine output monitoring with urethral catheter placement were combined with optimized Guideline-directed medical therapies (GDMT) for HF [1].

2.2.2. Radiofrequency catheter ablation protocol

The CARTO navigation system (Biosense Webster Inc., Diamond Bar, CA, USA) was used for atrial reconstruction and guidance of AF ablation. The procedures were uniformly conducted under conscious sedation. Three catheters were introduced: (1) a steerable coronary sinus (CS) catheter (2–5-2 mm, Dynamic XT; Boston Scientific, Marlborough, MA, USA); (2) the THERMOCOOL SMARTTOUCH SF (STSF) catheter (Biosense Webster, Inc.) with a 56-hole irrigation tip; and (3) a multipolar mapping catheter (Pentaray; Biosense Webster Inc.).

All patients received a tailored catheter ablation, which followed “stepwise strategy”, starting with PVI, followed by CFAEs ablation, and LA linear ablation of roof and mitral isthmus [16]. As for other ablation lines, they were decided when cardiac mapping presents specific AFL/reentrance that are dependent to these structures. We selectively conducted linear ablation in persistent AF patients, based on cardiac mapping and atrial reconstruction results. Ablation index was employed to guide lesion quality, targeting values of 500 to 550 in the anterior wall, 350 to 400 in the posterior wall, 450 to 500 in the cavotricuspid isthmus and roofline, and 550 to 600 in the mitral isthmus. The procedural end point was successful PVI and bidirectional block of the ablation lines, validated through differential pacing techniques and activation mapping after restoring sinus rhythm.

2.3. Post-procedure management and follow-up

Amiodarone was continuously administered during the 3-month post-procedure blanking period, and propafenone was considered a substitution in case of contradictions on amiodarone use. All anti-arrhythmic medications were stopped (except beta-blockers for anti-HF treatment) after the blanking period. Optimized GDMT was dynamically adjusted at every visit [1]. Standard anticoagulation therapies were delivered [3]. No procedures were repeated during the 3-month blanking period.

All patients were followed up to one-year post-procedure. For patients in the elective group, HF rehospitalization events during the blanking period were also followed. Severe procedural complications including cardiac tamponade, pericardial effusion not requiring pericardiocentesis, device embolism, stroke/TIA/systemic embolism, air embolism, major bleeding and death were recorded. Patients were advised to outpatient follow-up at 3, 6, 9 and 12 months after procedure. A whole set of scheduled examinations including 12-lead electrocardiogram, 24-hour Holter, transthoracic echocardiogram, and serum NT-proBNP was routinely ordered in the visit at the 3-month visit. Immediate clinical visit was recommended if patients had a recurrence of chest tightness, palpitation, deteriorating HF symptoms, or abnormal findings on their own personal portable heart rhythm monitor device. Follow-up by telephone or home visits were also applied as supplement.

2.4. Statistical analysis

Considering the non-randomized design of our study, propensity score matching (PSM) was performed via a logistic regression model. A rigorous nearest-neighbor matching algorithm without replacement was attempted, using a caliper width of ≤ 0.1. Standardized biases were calculated before and after PSM, among which a value of < 0.2 was considered the indicator of adequate bias reduction.

Continuous variables are shown as the mean ± standard deviation (SD) or median with interquartile range (25th–75th percentile), and categorical variables are shown as counts and percentages. Two-tailed Student’s t-tests, Pearson’s chi-square, and Fisher’s exact tests were used for the comparison of variables as appropriate. Multiple paired t-tests were conducted using Bonferroni-Dunn correction. Kaplan-Meier analysis was performed for arrhythmia recurrence (after a 3-month blanking period), HF rehospitalization (since discharge), and other endpoint events. A P-value of < 0.05 was considered statistically significant. Statistics were performed with SPSS 23.0 (IBM, Armonk, NY, USA) and GraphPad Prism 10 (GraphPad Software, Boston, MA, USA).

3. Results

3.1. Patient demographics

A total of 159 patients who had ADHF and AF between November 2020 and November 2022 received RFCA. Twenty-nine patients were either not eligible for inclusion or failed to follow up. The remaining 130 patients were categorized into the early group (n = 63) and the elective group (n = 67) based on the treatment they received, and 50 matched pairs were selected by PS matching (Fig. 1). Basic clinical covariates, including age, gender, BMI, CHA2DS2-VASc score, comorbidities (hypertension, diabetes mellites, coronary artery disease, chronic obstructive pulmonary disease, stroke/transient ischemic attack, chronic kidney disease, liver insufficiency, hypertrophic cardiomyopathy, dilated cardiomyopathy and valvular heart disease), heart failure manifestation, and classification and evaluation were used for matching. A standardized bias of < 0.20 was observed in all covariates after matching. Patients’ basic characteristics are summarized in Table 1.

Table 1.

Basic characteristics.

Characteristics	Before Matching		After Matching		Standardized Bias
	Early group (n = 63)	Elective group (n = 67)	Early group (n = 50)	Elective group (n = 50)	Before Matching	After Matching
Age, y	71.2 ± 10.3	70.8 ± 8.6	71.5 ± 10.9	71.6 ± 9.2	0.04	0.01
Female	28 (44.4)	21 (31.3)	20 (40.0)	19 (38.0)	0.27	0.04
BMI, kg/m2	24.3 ± 3.3	24.6 ± 3.8	24.4 ± 3.4	24.6 ± 4.3	0.09	0.07
CHA2DS2-VASc score	4.1 ± 1.4	4.0 ± 1.4	4.1 ± 1.5	4.0 ± 1.3	0.10	0.08
Comorbidities
Hypertension	47 (74.6)	43 (64.2)	36 (72.0)	32 (64.0)	0.23	0.17
Diabetes mellites	17 (27.0)	20 (29.9)	14 (28.0)	14 (28.0)	0.06	0.00
Coronary artery disease	32 (50.8)	24 (35.8)	25 (50.0)	21 (42.0)	0.30	0.16
COPD	2 (3.2)	2 (3.0)	1 (2.0)	2 (4.0)	0.01	0.12
Stroke/TIA	9 (14.3)	13 (19.4)	6 (12.0)	7 (14.0)	0.14	0.06
CKD stage 5	1 (1.6)	3 (4.5)	1 (2.0)	1 (2.0)	0.17	0.00
Liver insufficiency	0	4 (6.0)	0	0	0.35	0.00
HCM	2 (3.2)	7 (10.4)	2 (4.0)	2 (4.0)	0.29	0.00
DCM	0	0	0	0	0.00	0.00
Valvular Heart Disease	0	0	0	0	0.00	0.00
Heart failure manifestations, at admission
NYHA class IV	22 (34.9)	24 (35.8)	20 (40.0)	19 (38.0)	0.02	0.04
Exacerbation of fatigue and dyspnea	62 (98.4)	65 (97.0)	49 (98.0)	49 (98.0)	0.09	0.00
Orthopnea	39 (61.9)	46 (68.7)	34 (68.0)	35 (70.0)	0.14	0.04
Peripheral edema	40 (63.5)	43 (64.2)	32 (64.0)	29 (58.0)	0.01	0.12
Polyserous cavity effusion or pulmonary edema	38 (60.3)	34 (50.7)	28 (56.0)	26 (52.0)	0.19	0.08
Heart failure classification and evaluation, at admission
HFpEF	41 (65.1)	41 (61.2)	32 (64.0)	32 (64.0)	0.08	0.00
HFrEF	7 (11.1)	13 (19.4)	6 (12.0)	8 (16.0)	0.23	0.12
NT-proBNP, pg/mL	2622.5 ± 2165.5	3121.6 ± 5047.8	2628.4 ± 2231.9	2469.4 ± 2359.2	0.13	0.07
LVEF, %	52.9 ± 9.9	51.2 ± 10.6	52.5 ± 10.0	51.9 ± 10.4	0.17	0.06
LAD, mm	45.6 ± 5.5	47.3 ± 4.9	46.0 ± 5.7	46.3 ± 4.9	0.32	0.07
Long-standing persistent AF	11 (17.5)	10 (14.9)	10 (20.0)	8 (16.0)	0.06	0.10
AF history, months a	45.9 ± 61.3	31.5 ± 51.2	47.5 ± 60.1	43.6 ± 51.4	0.25	0.07
Having AF catheter ablation history	11 (17.5)	9 (13.4)	9 (18.0)	8 (16.0)	0.11	0.05

Values are given as the mean ± SD, median [25th percentile, 75th percentile], or n (%) as appropriate. BMI = body mass index; CHA2DS2-VASc = Congestive Heart Failure, Hypertension, Age ≥ 75 [Doubled], Diabetes Mellitus, prior stroke or transient ischemic attack [Doubled], Vascular Disease, Age 65–74, Female; COPD = Chronic obstructive pulmonary disease; TIA = Transient ischemic attack; CKD = Chronic kidney disease; HCM = Hypertrophic cardiomyopathy; DCM = Dilated Cardiomyopathy; NYHA = New York Heart Association; HFpEF = Heart failure with preserved ejection fraction; HFrEF = Heart failure with reduced ejection fraction; NT-proBNP = N-terminal-proB-type natriuretic peptide; LVEF = Left ventricular ejection fraction; LAD = Left atrial diameter.

^aRefers to the time span since the first diagnosis of AF, regardless of paroxysmal/persistent AF. Since some of the AF histories were provided based on recollection of patients rather than written medical records, especially in those who had a long and complicated AF history, this data is for reference only.

3.2. Treatment characteristics and procedural complications

All patients received GDMT for the HF control, based on clinical types and symptoms of HF. There was no difference in the medication between the two groups (Table 2). Cardiac function was re-assessed immediately before the procedure to confirm the HF control and determine the ability of the patients in to tolerate the ablation therapy. No difference was observed in the pre-procedural characteristics of the two groups (Table 2).

Table 2.

Pre-procedural medication and assessment.

Characteristics	Early group (n = 50)	Elective group (n = 50)	P-Value
Anti-HF Drug, before procedure
ACEi, ARB, or ARNi	33 (66.0)	33 (66.0)	1.000
Beta blocker	31 (62.0)	36 (72.0)	0.288
MRA	40 (80.0)	43 (86.0)	0.424
SGLT2i	6 (12.0)	6 (12.0)	1.000
Loop diuretics	43 (86.0)	44 (88.0)	0.766
Digitalis	19 (38.0)	19 (38.0)	1.000
rhBNP	7 (14.0)	6 (12.0)	0.766
Antiarrhythmic drug, before procedure a
Propafenone	0	0	/
Other Class I b	0	0	/
Amiodarone	8 (16.0)	9 (18.0)	0.790
Other Class III c	0	0	/
Class IV d	12 (24.0)	10 (20.0)	0.629
Pre-procedural Assessment
NYHA class IV	0	0	/
NT-proBNP, pg/mL	1003.0 ± 573.9	1090.8 ± 432.0	0.389
LVEF, %	53.4 ± 8.8	52.8 ± 8.7	0.745
LAD, mm	45.4 ± 5.1	45.5 ± 5.8	0.905

ACEi = angiotensin-converting enzyme inhibitor; ARB = angiotensin II receptor blocker; ARNi = angiotensin receptor-neprilysin inhibitor; MRA = mineralocorticoid receptor antagonist; SGLT2i = sodium-glucose co-transporter 2 inhibitor; rhBNP = reformed human B-type natriuretic peptide.

^aOnly refers to the use of antiarrhythmic drugs between admission and procedure.

^bIncluding mexiletine, and moricizine; ^c Including sotalol and dronedarone. ^d Including verapamil and diltiazem.

Tailored catheter ablation was adopted in all enrolled patients, and procedural characteristics of the early and elective groups are summarized and compared in Table 2. A total of 3 patients did not receive PVI in the procedure. All these patients had history of receiving PVI, and the previous isolation of pulmonary vein was confirmed effective in cardiac mapping. Thus, we only did linear ablation for them.

In the early group, the average time between the onset of ADHF and the start of RFCA was 10.6 ± 2.5 days while 158.3 ± 84.5 days in the elective group. However, ablation protocols were similar between the two groups (Table 3). The overall procedural complication rate was comparable in the two groups (Early group: 8.0 %; Elective group: 8.0 %, P = 1.000). All complication rates were comparable between the two groups (Table 3). No embolism due to the device, thromboembolic events, major bleeding, or death was observed in both groups.

Table 3.

Procedural characteristics and post-procedural complications.

Procedural characteristics	Early group (n = 50)	Elective group (n = 50)	P-value
Blanking period, days	10.6 ± 2.5	158.3 ± 84.5	0.000
PVI	49 (98.0)	48 (96.0)	0.558
Mitral isthmus line	43 (86.0)	41 (82.0)	0.585
LA roof line	34 (68.0)	36 (72.0)	0.663
Anterior septal line	24 (48.0)	21 (42.0)	0.546
Cavo-tricuspid line	18 (36.0)	18 (36.0)	1.000
LA posterior and/or inferior lines	8 (16.0)	6 (12.0)	0.564
CFAE ablation	26 (52.0)	24 (48.0)	0.689
Left atrial appendage closure	21 (42.0)	26 (52.0)	0.316
Intracardiac cardioversion	21 (42.0)	22 (44.0)	0.840
Intraprocedural sinus rhythm restoration	27 (54.0)	26 (52.0)	0.841
Total procedure time, min	182.7 ± 47.2	173.2 ± 43.5	0.298
Fluoroscopy time, s	776.4 ± 353.2	757.9 ± 333.1	0.788
Electrophysiological Diagnose
cAF	34 (68.0)	35 (70.0)	0.829
AFL	16 (32.0)	15 (30.0)	0.829
Severe complications
Total	4 (8.0)	4 (8.0)	1.000
HF exacerbation after procedure, resulting in delayed dischargea	4 (8.0)	4 (8.0)	1.000
Cardiac tamponade	2 (4.0)	1 (2.0)	0.558
Pericardial effusion not requiring pericardiocentesis b	1 (2.0)	2 (4.0)	0.558
Device embolism	0	0	/
Stroke/TIA/systemic embolism	0	0	/
Air embolism	0	0	/
Major bleeding	0	0	/
Death	0	0	/
Post-procedure prescription (at discharge)
ACEi/ARB/ARNi	32 (64.0)	34 (68.0)	0.672
Beta Blocker	26 (52.0)	23 (46.0)	0.548
SGLT2i	12 (24.0)	9 (18.0)	0.461
MRA	12 (24.0)	15 (30.0)	0.499
Loop Diuretics	15 (30.0)	20 (40.0)	0.294
Digitalis	2 (4.0)	1 (2.0)	0.558
Propafenone	1 (2.0)	0	0.315
Other Class I	0	0	/
Amiodarone	49 (98.0)	50 (100.0)	0.315
Other Class III	0	0	/
Class IV	2 (4.0)	2 (4.0)	1.000

Blanking period refers to the gap between the onset of ADHF and the start of RFCA. PVI = Pulmonary vein isolation; LA = Left atrial; CFAE = complex fractionated atrial electrogram; cAF = continuous atrial fibrillation; AFL = atrial flutter.

^aIdentified by delayed discharge with intensified use of diuretics.

^bDefined as mild to moderate pericardial effusion, which means size of the effusion on echocardiography was 3–20 mm.

3.3. Follow-up events

During the follow-up period of 12 months, one patient (2.0 %) died of ADHF recurrence in the elective group, 12 patients (24.0 %) were hospitalized for worsening symptoms of heart failure, and 12 patients (24.0 %) had recurrent AF/AT. In comparison, no patient in the early group died, rehospitalization for HF occurred in 5 patients (10.0 %), and AF/AT recurred in 14 patients (28.0 %). The survival analyses were illustrated in Fig. 2a and 2b. The early group had significantly less rehospitalization due to HF than the elective group (Mantel-Cox test: P = 0.036; HR: 0.369; 95 %CI: 0.145–0.938), however, AF recurrence was similar between the two groups (Mantel-Cox test: P = 0.645; HR: 1.204; 95 %CI: 0.547–2.648).

Fig. 2.

Post-procedure follow-up and full-time monitoring in patients of the early and elective group. a. Cumulative incidence of the composite of heart failure rehospitalization. b. Cumulative incidence of the composite of atrial fibrillation recurrence. c. Comparison of 90-day rehospitalization rate between the two groups.

Considering the first hospital discharge, the 90-day HF rehospitalization rate was significantly higher in the elective group (13, 26.0 %vs. 2, 4.0 %, P = 0.002) when the patients were in the transitional medication period and waiting for their elective procedure, while patients in the early group already received RFCA (Fig. 2c).

3.4. Cardiac function evaluations

Baseline and 3-month post-procedural cardiac function evaluation indexes are shown for comparison in Fig. 3a-f, respectively, based on the LVEF classification for all patients involved. In patients with HFpEF from both early and elective groups showed marked decrease in LAD (Paired t-test: Early group: 45.4 mm vs. 42.2 mm, D = 3.3 mm, S.B. = 0.803, P < 0.001; Elective group: 45.6 mm vs. 41.5 mm, D = 4.1, S.B. = 0.788, P < 0.001) (Fig. 3b). However, no significant improvement was detected in the LVEF (Paired t-test: Early group: 58.9 % vs. 60.4 %, D = 1.4, S.B. = 1.524, P = 0.344; Elective group: 58.9 % vs. 61.1 %, D = 2.2, S.B. = 1.015, P = 0.070), or NT-proBNP (Paired t-test: Early group: 2307 vs. 1583 pg/mL, D = 724.5, S.B. = 619.2, P = 0.251; Elective group: 2061 vs. 1075 pg/mL, D = 986.5, S.B. = 582.0, P = 0.190) (Fig. 3a, 3c).

Fig. 3.

Post-procedural evaluation and post-procedure follow-up in patients who had different LVEF classifications. a-c. Comparison of baseline and post-procedure LVEF, LAD, and serum NT-proBNP levels in HFpEF patients. d-f. Comparison of baseline and post-procedure LVEF, LAD, and serum NT-proBNP level in HFmrEF/HFrEF patients. g-h. Cumulative incidence of the composite of heart failure rehospitalization and atrial fibrillation recurrence in HFpEF patients. i-j. Cumulative incidence of the composite of heart failure rehospitalization and atrial fibrillation recurrence in HFmrEF/HFrEF patients. HFmrEF = Heart failure with mildly reduced ejection fraction.

In patients with HFmrEF/HFrEF, significant improvement in the LVEF was found in both groups (Paired t-test: Early group: 41.0 % vs. 55.1 %, D = 14.1, S.B. = 2.233, P < 0.001; Elective group: 39.5 % vs. 54.2 %, D = 14.7, S.B. = 1.997, P < 0.001) (Fig. 3d). An apparent reduction in the LAD was also observed in both groups (Paired t-test: Early group: 47.0 mm vs. 43.8 mm, D = 3.1, S.B. = 1.006, P < 0.001; Elective group: 47.6 mm vs. 42.5 mm, D = 5.1, S.B. = 0.881, P < 0.001) (Fig. 3e). A marked drop in serum NT-proBNP level was found in the early group (Paired t-test: Early group: 3200 vs. 1537 pg/mL, D = 1663, S.B. = 651.0, P = 0.041; Elective group: 3195 mm vs. 3677 mm, D = 482.5, S.B. = 1856, P = 0.798) (Fig. 3f).

Survival analysis on HF rehospitalization and AF recurrence in different LVEF classification subgroups are shown in Fig. 3g-j. Their basic characteristics, procedural characteristics, and post-procedural complications are presented for comparison in Table 4, Table 5. Patients with HFmrEF/HFrEF showed significantly less HF rehospitalization with early RFCA, compared with the elective group (Mantel-Cox test: P = 0.027; HR: 0.241; 95 %CI: 0.068–0.849) (Fig. 3i).

Table 4.

Baseline Characteristics in different LVEF subgroups.

Baseline Characteristics	HFpEF			HFmrEF + HFrEF
	Early group (n = 32)	Elective group (n = 32)	p Value	Early group (n = 18)	Elective group (n = 18)	p Value
Age, y	74.4 ± 9.9	74.1 ± 8.1	0.891	66.3 ± 11.2	67.2 ± 9.8	0.813
Female	16 (50.0)	12 (37.5)	0.313	4 (22.2)	7 (38.9)	0.278
BMI, kg/m2	24.4 ± 3.5	24.9 ± 4.5	0.572	24.4 ± 3.4	24.1 ± 4.1	0.820
CHA2DS2-VASc score	4.3 ± 1.3	4.1 ± 1.1	0.472	3.7 ± 1.8	3.8 ± 1.7	0.924
Comorbidities
Hypertension	24 (75.0)	22 (68.8)	0.578	12 (66.7)	10 (55.6)	0.494
Diabetes mellites	12 (37.5)	8 (25.0)	0.281	2 (11.1)	6 (33.3)	0.109
Coronary artery disease	19 (59.4)	16 (50.0)	0.451	6 (33.3)	5 (27.8)	0.717
COPD	0	1 (3.1)	0.313	1 (5.6)	1 (5.6)	1.000
Stroke/TIA	3 (9.4)	3 (9.4)	1.000	3 (16.7)	4 (22.2)	0.674
CKD stage 5	1 (3.1)	1 (3.1)	1.000	0	0	/
Liver insufficiency	0	0	/	0	0	/
HCM	2 (6.3)	1 (3.1)	0.554	0	1 (5.6)	0.310
DCM	0	0	/	0	0	/
Valvular Heart Disease	0	0	/	0	0	/
Heart failure manifestations at admission
NYHA class IV	9 (28.1)	13 (40.6)	0.292	11 (61.1)	6 (33.3)	0.095
Exacerbating fatigue and dyspnea	31 (96.9)	31 (96.9)	1.000	18 (100.0)	18 (100.0)	1.000
Orthopnea	19 (59.4)	23 (71.9)	0.292	15 (83.3)	12 (66.7)	0.248
Peripheral edema	21 (65.6)	22 (68.8)	0.790	11 (61.1)	7 (38.9)	0.182
Polyserous cavity effusion or pulmonary oedema	17 (53.1)	14 (43.8)	0.453	11 (61.1)	12 (66.7)	0.729
Heart failure evaluation at admission
NT-proBNP, pg/mL	2307.2 ± 1172.5	2061.4 ± 2106.1	0.615	3199.5 ± 2895.2	3194.6 ± 2722.2	0.996
LVEF, %	58.9 ± 4.5	58.9 ± 4.2	0.966	41.0 ± 6.1	39.5 ± 5.5	0.443
LAD, mm	45.4 ± 5.6	45.6 ± 4.4	0.880	47.0 ± 6.0	47.6 ± 5.6	0.751
Anti-heart failure drug before procedure
ACEi, ARB or ARNi	21 (65.6)	18 (56.3)	0.442	12 (66.7)	15 (83.3)	0.248
Beta blocker	21 (65.6)	22 (68.8)	0.790	10 (55.6)	14 (77.8)	0.157
MRA	26 (81.3)	28 (87.5)	0.491	14 (77.8)	15 (83.3)	0.674
SGLT2i	2 (6.3)	3 (9.4)	0.641	4 (22.2)	3 (16.7)	0.674
Loop duretics	27 (84.4)	29 (90.6)	0.450	16 (88.9)	15 (83.3)	0.630
Digitalis	11 (34.4)	11 (34.4)	1.000	8 (44.4)	8 (44.4)	1.000
rhBNP	5 (15.6)	2 (6.3)	0.230	2 (11.1)	4 (22.2)	0.371
Antiarrhythmic drug before procedure
Propafenone	0	0	/	0	0	/
Other Class I	0	0	/	0	0	/
Amiodarone	5 (15.6)	6 (18.8)	0.740	3 (16.7)	3 (16.7)	1.000
Other Class III	0	0	/	0	0	/
Class IV	9 (28.1)	5 (15.6)	0.226	3 (16.7)	5 (27.8)	0.423
Pre-procedural Assessment
NYHA class IV	0	0	/	0	0	/
NT-proBNP, pg/mL	872.3 ± 399.8	1108.7 ± 408.5	0.023	1235.2 ± 754.1	1059.0 ± 481.6	0.409
LVEF, %	57.3 ± 6.4	57.7 ± 5.0	0.746	46.5 ± 8.3	44.1 ± 6.9	0.347
LAD, mm	45.0 ± 5.1	44.8 ± 5.3	0.924	46.1 ± 5.3	46.7 ± 6.5	0.768
Long-standing persistent AF	6 (18.8)	5 (15.6)	0.740	4 (22.2)	3 (16.7)	0.673
AF history, months	41.9 ± 49.9	49.2 ± 58.9	0.594	57.3 ± 76.9	33.6 ± 35.9	0.244
Having AF catheter ablation history	5 (15.6)	6 (18.8)	0.740	4 (22.2)	2 (11.1)	0.371

Table 5.

Procedural Characteristics in different LVEF subgroups.

Procedural Characteristics	HFpEF			HFmrEF+HFrEF
	Early group (n=32)	Elective group (n=32)	p Value	Early group (n=18)	Elective group (n=18)	p Value
Blanking period, days	10.2±2.7	170.8±100.2	0.000	11.2±2.2	136.0±38.2	0.000
PVI	32 (100.0)	30 (93.8)	0.151	17 (94.4)	18 (100.0)	0.310
Mitral isthmus line	25 (78.1)	25 (78.1)	1.000	18 (100.0)	16 (88.9)	0.146
LA roof line	20 (62.5)	21 (65.6)	0.794	14 (77.8)	15 (83.3)	0.674
Anterior septal line	17 (53.1)	11 (34.4)	0.131	7 (38.9)	10 (55.6)	0.317
Cavo-tricuspid line	11 (34.4)	13 (40.6)	0.606	7 (38.9)	5 (27.8)	0.480
LA posterior and/or inferior lines	5 (15.6)	4 (12.5)	0.719	3 (16.7)	2 (11.1)	0.630
CFAE ablation	15 (46.9)	14 (43.8)	0.801	11 (61.1)	10 (55.6)	0.735
Left atrial appendage closure	17 (53.1)	17 (53.1)	1.000	4 (22.2)	9 (50.0)	0.083
Intracardiac cardioversion	14 (43.8)	12 (37.5)	0.611	7 (38.9)	10 (55.6)	0.317
Intraprocedural sinus rhythm restoration	17 (53.1)	19 (59.4)	0.614	10 (55.6)	7 (38.9)	0.317
Total procedure time, min	179.6±46.2	169.7±44.0	0.380	188.2±49.7	179.5±43.1	0.580
Fluoroscopy time, s	789.3±378.8	636.6±234.1	0.057	753.6±311.7	973.6±378.2	0.065
Electrophysiological Diagnose
cAF	21 (65.6)	22 (68.8)	0.790	13 (72.2)	13 (72.2)	1.000
AFL	11 (34.4)	10 (31.2)	0.790	5 (27.8)	5 (27.8)	1.000
Severe complications
Total	3 (9.4)	2 (6.3)	0.641	1 (5.6)	2 (11.1)	0.546
HF exacerbation after procedure, resulting in delayed discharge	3 (9.4)	2 (6.3)	0.641	1 (5.6)	2 (11.1)	0.546
Cardiac tamponade	2 (6.3)	1 (3.1)	0.554	0	0	/
Pericardial effusion not requiring pericardiocentesis	1 (3.1)	1 (3.1)	1.000	0	1 (5.6)	0.310
Device embolism	0	0	/	0	0	/
Stroke/TIA/systemic embolism	0	0	/	0	0	/
Air embolism	0	0	/	0	0	/
Major bleeding	0	0	/	0	0	/
Death	0	0	/	0	0	/
Post-procedure prescription (at discharge)
ACEi/ARB/ARNi	18 (56.3)	18 (56.3)	1.000	14 (77.8)	16 (88.9)	0.371
Beta Blocker	16 (50.0)	12 (37.5)	0.313	10 (55.6)	11 (61.1)	0.735
SGLT2i	6 (18.8)	6 (18.8)	1.000	6 (33.3)	3 (16.7)	0.248
MRA	7 (21.9)	14 (43.8)	0.062	5 (27.8)	1 (5.6)	0.074
Loop Diuretics	10 (31.3)	15 (46.9)	0.200	5 (27.8)	5 (27.8)	1.000
Digitalis	1 (3.1)	0	0.313	1 (5.6)	1 (5.6)	1.000
Propafenone	1 (3.1)	0	0.313	0	0	/
Other Class I	0	0	/	0	0	/
Amiodarone	31 (71.9)	32 (100.0)	0.313	18 (100.0)	18 (100.0)	1.000
Other Class III	0	0	/	0	0	/
Class IV	1 (3.1)	0	0.313	1 (5.6)	2 (11.1)	0.546

4. Discussion

4.1. Main findings

This study investigated patients with persistent AF complicated with ADHF who received an early RFCA after the initial control of ADHF symptoms. The main findings were as follows. (1) Early RFCA showed acceptable safety and was non-inferior to elective procedure; (2) Patients who received early RFCA therapy had significantly less HF rehospitalization in the 1-year post-procedure follow-up period; And (3) The rehospitalization rate was significantly higher in the transitional period in the elective group, compared with patients who already received early RFCA.

4.2. Safety and efficacy of early RFCA in patients with persistent AF and ADHF

As a severe and terminal event in the long-lasting course of heart failure, ADHF is associated with substantial mortality and morbidity. In the ARIC study, mortality within 1 year of the first hospitalization for ADHF was 28.3 %–35 %, whereas ADHF rehospitalization ratio within 1 year was 36.8 %–45.2 %, and the linkage between within 12-months consecutive ADHF attack and increasing mortality rate was also revealed [17], [18]. Furthermore, AF often coexists with HF, and they could cause and exacerbate each other through a variety of mechanisms, thus AF is associated with higher mortality and morbidity in HF patients [19], [20], [21], [22].

As a conventional cardiac rhythm control strategy, electrical cardioversion (ECV) was highly recommended in contemporary guidelines for managing HF with AF (1). However, the relatively high AF recurrence rate (approximately 50 % in the first two weeks and 65 % in the first year) is a limitation of using ECV in patients with AF and ADHF [23].

The long-term efficacy of RFCA in reducing mortality in patients with AF complicated with ADHF was demonstrated in several studies. This may be attributed to a better SR maintenance of RFCA over drug therapy, and further the interruption of “AF-HF circle”, resulting in less HF worsening [4], [12], [24], [25]. However, ADHF is unstable for 3 months after hospital discharge, this raised question on the optimal timing of the RFCA procedure [9], [10].

In this study, we investigated the procedural complications,1-year prognosis, and cardiac functional impact of the early RFCA strategy in AF patients with ADHF. The early RFCA intervention was proven to be safe and effective. Moreover, patients who received early RFCA showed significantly lower HF rehospitalization rates during the vulnerable period of 90 days, compared with patients who were in the transitional period waiting for the elective ablation procedure. These results suggest that early RFCA therapy after ADHF attack is effective and safe, and superior to the elective procedure in patients with persistent AF and ADHF.

4.3. How did the patients with persistent AF and ADHF benefit from early RFCA

We have to face the unneglectable fact that the 1-year AF recurrence was similar across the two groups. Therefore, the lower HF recurrence rate in the early group may not be simply attributed to a better long-term rhythm control. Hence, we need to explain the potential mechanism of what made early RFCA beneficial in AF patients who just suffered from ADHF attack.

Recent studies proved that patients with AF and HF could benefit from early catheter ablation within 12 or 6 months of their first diagnosis of AF [26], [27]. However, the setting was different in our study. The majority (80 %) of our patients had an affirmative AF history of > 12 months at enrollment, and 17 % patients had received AF-RFCA before. Hence, we do not think patients in our cohort benefited from ‘early ablation’ defined according to AF history, but a better maintenance of SR in the unstable period after ADHF attack, which could lead to the vicious cycle of ADHF recurrence if not effectively managed [28].

The ‘early ablation’, or in other words, ‘immediate ablation’ we proposed was relative to the ADHF event. It could even be defined as ‘late ablation’ from the aspect of the whole AF course. Although it is supposed that catheter ablation should be delivered in the early stage, we might dare go a little further to propose that a better rhythm control is still meaningful in AF patients who already had advanced HF, with the support of other existing evidence [12], [6]. Moreover, in clinical practice, numerous patients with AF are not diagnosed until they are hospitalized for worsening HF symptoms, and thus receiving an early RFCA therapy at the initial admission is also a potentially economical choice.

We propose the potential mechanisms of symptomatic improvement in our patients could be attributed to two major aspects: LA reverse remodeling and tachycardia induced cardiomyopathy. In a study involving 502 patients with non-paroxysmal AF and undergoing catheter ablation, improvement of LAD, LVEF and NYHA classification was observed in both HFpEF and HFmrEF/HFrEF patients[29]. However, E/e’ ratio remain unchanged in HFpEF patients, suggesting that LA reverse remodeling rather than LV diastolic function is responsible for the symptomatic improvement. Significant decrease in LAD was also witnessed in our study, which indicates the reverse remodeling of left atrium after RFCA in patients with ADHF.

On the other hand, rhythm control with catheter ablation could also help reverse the potential tachycardia induced cardiomyopathy, although the effect might be limited under the circumstance of a long arrhythmic history. Previous study showed that the median time from onset of arrhythmia to cardiomyopathy and development of HF was 4.2 years, while rhythm control, on the other hand, was found meaningful in the improvement in LVEF and resolution of HF [30]. In our study, the average AF history of the patients was approximately 40 months, and significant improvement was also observed in HFmrEF/HFrEF patients, which was coherent with previous findings.

To sum up, we propose RFCA to be performed as early as possible in patients diagnosed with persistent AF/AFL complicated with ADHF when they can to tolerate the procedure after initial medication (Fig. 4).

Fig. 4.

Central illustration: clinical outcome of different treatment. RFCA helps breaking the vicious cycle of ADHF recurrence in patients with AF and ADHF. Earlier the procedure is done, better the prognosis will be.

5. Limitations

As an observational study, this study lacked a randomized control group of patients who received drug therapy only for the whole course, and therefore, the outcome didn’t provide strong evidence of the superiority of early RFCA in patients with persistent AF and ADHF. Due to this, this study could only serve as a proof-of-concept study, but not a clinical implication claim. The number of patients with ADHF was relatively small, and the 1-year follow-up period was relatively short. Therefore, the results should be further testified in a larger population with a longer observation period and more detailed imaging evidence. Although PSM matching was applied to control the selection bias, it is possible for certain unmeasured confounders to influence the findings. Future prospective randomized clinical trials are needed to eliminate selection bias. Patients who underwent combined RFCA and LAAC procedures were also enrolled, whose impact on cardiac function was not explored, making it difficult being balance as a procedural characteristic. All procedures in this study were performed by experienced operators to guarantee procedural safety, time efficiency, and effectiveness, but a longer learning curve might be required for patients with severe HF.

6. Conclusions

Early RFCA intervention after ADHF attack is safe and effective in patients with persistent AF and ADHF. Patients who received early RFCA had significantly less HF rehospitalization in the 1-year post-procedure follow-up period. On the other hand, the elective procedure was accompanied by a higher risk of HF rehospitalization during the waiting period.

Funding support and author disclosures

This work was supported by National Science Foundation of China (No.82130009, 82070515, 82370518, 82002079), Shanghai City Committee of Science and Technology Research Projects (No.23Y21900700), Shanghai Leading Talent Plan 2020. The authors have reported that they have no relationships relevant to the content of this paper to disclose.

Acknowledgements and funding sources

This work was supported by National Natural Science Foundation of China (No.82130009, 82070515, 82370518, 82002079), Shanghai Sailing Program (No.22YF1427400). The authors have reported that they have no relationships relevant to the content of this paper to disclose.

CRediT authorship contribution statement

Qian-ji Che: Writing – review & editing, Writing – original draft, Resources, Project administration, Investigation, Formal analysis, Data curation, Conceptualization. Jun-hao Qiu: Formal analysis, Data curation, Conceptualization. Jian Sun: Writing – review & editing, Data curation, Conceptualization. Mu Chen: Project administration, Investigation. Wei Li: Supervision, Resources, Investigation. Qun-Shan Wang: Supervision, Resources, Investigation. Peng-Pai Zhang: Supervision, Resources, Investigation. Yu-li Yang: Investigation, Funding acquisition, Data curation. Rui Zhang: Writing – review & editing, Writing – original draft, Supervision, Resources, Project administration, Methodology, Investigation, Conceptualization. Yi-Gang Li: Writing – review & editing, Supervision, Project administration, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

The following are the Supplementary data to this article:

Data availability

Data will be made available on request.