Catheter ablation of atrial fibrillation in patients with autoimmune disease: A propensity score matching study based on the China Atrial Fibrillation Registry

Abstract Background Evidence on outcomes of catheter ablation (CA) for atrial fibrillation (AF) in patients with autoimmune disease (AD) is limited. Hypothesis Patients with AD had worse outcomes after CA procedures for AF. Methods A retrospective analysis was performed in patients undergoing AF ablation between 2012 and 2021. The risk of recurrence after ablation was investigated in patients with AD and a 1:4 propensity score matched non‐AD group. Results We identified 107 patients with AD (64 ± 10 years, female 48.6%) who were matched with 428 non‐AD patients (65 ± 10 years, female 43.9%). Patients with AD exhibited more severe AF‐related symptoms. During the index procedure, a higher proportion of AD patients received nonpulmonary vein trigger ablation (18.7% vs. 8.4%, p = 0.002). Over a median follow‐up of 36.3 months, patients with AD experienced a similar risk of recurrence with the non‐AD group (41.1% vs. 36.2%, p = 0.21, hazard ratio [HR]: 1.23, 95% confidence interval [CI]: 0.86–1.76) despite a higher incidence of early recurrences (36.4% vs. 13.5%, p = 0.001). Compared with non‐AD patients, patients with connective tissue disease were at an increased risk of recurrence (46.3% vs. 36.2%, p = 0.049, HR: 1.43, 95% CI: 1.00–2.05). Multivariate Cox regression analysis showed that the duration of AF history and corticosteroid therapy were independent predictors of postablation recurrence in patients with AD. Conclusions In patients with AD, the risk of recurrence after ablation for AF during the follow‐up was comparable with non‐AD patients, but a higher risk of early recurrence was observed. Further research into the impact of AD on AF treatment is warranted.


| INTRODUCTION
Previous research had revealed that selected autoimmune diseases (AD), such as rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE), are associated with an elevated risk of developing cardiovascular disease. [1][2][3] However, the impact of AD on the incidence and outcome of arrhythmic diseases is less explored.
Recently, a large-scale population-based study revealed that the hazard ratio (HR) of developing atrial fibrillation (AF) or atrial flutter was 1.40 (95% confidence interval [CI]: 1.35-1.46) in AD patients compared to the general population. 1 Other sporadic evidence also showed that the risk of developing AF is higher in patients with RA, ankylosing spondylitis (AS), and psoriasis than in the general population. [4][5][6][7] This risk is elevated through multifaceted mechanisms.
Beyond a higher prevalence of conventional risk factors, 4 autoimmunity also contributes to the occurrence of AF in a complicated way, including direct effects on connective tissue and small vessels, cardiomyocytes, and possibly from some AD treatments. Systemic inflammatory states can evoke a proinflammatory transformation in the epicardial adipose tissue, resulting in the release of proinflammatory adipocytokines 8,9 and myocardial fibrosis in the adjacent atrium.
Eventually, atrial myopathy may be developed and lead to increased stiffness of the atrial wall and reduced mechanical atrial functions.
Catheter ablation (CA) has recently become a mainstream AF treatment. 10 However, recurrences of AF after the ablation procedure are relatively common and are influenced by factors including inflammation status, atrial fibrosis, and so forth. A recent survey showed that CA for tachyarrhythmia is often withheld in patients with AD at the physician's discretion for fear of the high risk of procedure-related complications and recurrences. 11 Up to now, the safety and efficacy of CA for AF in patients with AD have not been evaluated. In this research, we evaluated the outcome of CA for AF in patients with AD.

| Study design and study population
Patients who underwent CA for AF with a confirmed diagnosis of AD at Beijing Anzhen Hospital were retrospectively screened from the China Atrial Fibrillation Registry, which is a prospective registry study launched in 2011 (Chinese Clinical Trial Registry ChiCTR-OCH-13003729). 12 Patients with previous ablation history for AF, valvular disease, previous cardiac surgery, congenital heart disease, cardiomyopathy, advanced heart failure (HF) (New York Heart Association Class III-IV), and who were diagnosed with AD after the CA procedure were excluded. A propensity score (PS) matching approach was performed to match patients without AD (the non-AD group) and the AD group at a ratio of 4 The primary outcome of this study was the postablation recurrences, which were further classified into early recurrence (within the first 3 months), late occurrence (between 3 months and 1 year), and very late recurrence (beyond 1 year). Characteristics of CA procedures and periprocedural adverse events were documented and analyzed. All outcomes are compared in patients with and without AD, and in subgroups of AD patients including (1) with connective tissue disease (CTD, such as RA, AS, SLE) or only had organ-specific diseases (such as Hashimoto thyroiditis, psoriasis) based on the location where autoimmunity attack happens 1 ; (2) with single or multiple AD.

| Data collection
Baseline characteristics, including gender, age, AF type and duration, the severity of AF-related symptoms (European Heart Rhythm Association score [EHRA]), comorbidities, as well as the echocardiographic index including left atrial diameter (LAD), left ventricular ejection fraction (LVEF), systolic and mean pulmonary artery pressure (sPAP and mPAP) were collected. Levels of hypersensitive C-reactive protein (hs-CRP), white blood cell, and differential leukocyte counts were documented. Regular medical therapy (restricted to corticosteroids and disease-modifying antirheumatic drugs [DMARDs]) for patients with AD was recorded.

| CA procedure
General principles of ablation protocol at our center have been reported. 13,14 Three-dimensional electroanatomical mapping was performed with the CARTO 3 system (Biosense Webster) and ablation was performed with a standard 3.5 mm irrigated ablation catheter. Routine ablation strategy included pulmonary vein isolation (PVI) for paroxysmal atrial fibrillation (PAF) and a "2C3L" protocol for persistent atrial fibrillation (PsAF) 13

| Follow-up
Twelve-lead electrocardiogram(ECG) and transthoracic echocardiography were routinely performed the day after the procedure.
Twenty-four-hour dynamic ECG at 1, 2, 3, and 6 months after the procedure, as well as incident ECG during symptom onset were reviewed for any arrhythmia recurrence (defined as AF/atrial tachycardia >30 seconds). Outpatient revisits and telephone followup was carried out to collect information on postprocedural adverse events and redo procedures. All patients without contraindications were described with propafenone or amiodarone for 3 months, which would be discontinued if the patients presented with stable sinus rhythm.

| Statistical analysis
Baseline characteristics were presented as mean (standard deviation) or median (interquartile range  After PS matching, 428 AF patients (mean age 65 ± 10 years, female 43.9%, PsAF 36.7%) without AD were identified and constituted the non-AD group. Table 1 Table 1).

| CA
PVI was achieved in every involved patient. Linear ablation and CFAE ablation were conducted in a comparable proportion of patients in the AD group and the non-AD group (Table 2)

| DISCUSSION
Despite the small sample size and retrospective nature, to our knowledge, our study represents the largest observational cohort of patients with AD undergoing CA procedures for AF, which covered a larger spectrum of ADs compared with previous research. 15,16 Major findings of this research include: F I G U R E 2 Kaplan-Meier curve for atrial fibrillation (AF) recurrence after index catheter ablation for AF in subgroups. (A) In patients with paroxysmal AF; (B) in patients with persistent AF; (C) in patients with connective tissue disease (CTD) and organ-specific AD; and (D) in patients with single and multiple AD. AD, autoimmune disease; p1, patients with CTD compared to patients with organ-specific AD; p2, patients with CTD compared to non-AD patients; p3, patients with organ-specific AD compared to non-AD patients.
T A B L E 3 Univariable and multivariable logistic regression analysis for predictors of recurrence.  (1) Patients with AD tended to suffer from more severe AF-related symptoms; (2) ablation procedures in patients with AD were likely to be more sophisticated than those in non-AD patients, which required more interventions at non-PV triggers and extra linear ablation (e.g., box isolation, septal line, and anterior line); (3) during a long-term follow-up, patients with AD did not present a higher risk of atrial arrhythmia recurrence compared to non-AD patients after a single procedure. However, there is an increased risk of postablation recurrence in patients with CTD. Compared to non-AD patients, AD patients experienced more early recurrence; and (4) the duration of AF history is an independent predictor of AF recurrence in both groups, and regular usage of corticosteroids was associated with AF recurrence in patients with AD. However, a recent study came out with exactly the opposite conclusion: patients with RA were at higher risk of AF recurrence and receiving repeated ablation. 16 Our research revealed that, with a similar profile of cardiovascular risk factors, AD patients were not at an increased risk of recurrence during the long-term follow-up.
However, considering the negative influence of inflammation status, patients with AD are expected to have more complicated LA substrates. In our study, although LA substrate was not routinely mapped, a higher prevalence of non-PV triggers in patients with AD and a more frequent nonroutine linear ablation in the index procedure could potentially corroborate this hypothesis.
In our study population, the preablation hs-CRP was higher in the AD group compared to the non-AD group but was not of predictive value for recurrence like the previously reported. 17,18 The reason might be explained by the difference in the level of hs-CRP between the AD group and the non-AD group was not clinically prominent although statistically significant, which may reflect that some patients in the AD group were not in active disease status. However, some results from our study still indicated that an overall inflammatory status could negatively influence the efficacy of the CA procedure, as patients with CTD, which generally suffered from a systemic inflammatory status, 1 had a higher risk of arrhythmia recurrence compared to the patients with organ-specific AD (with a more localized inflammation status). In addition, the multivariate analysis revealed that regular oral corticosteroid therapy at the time of ablation was an independent predictor of arrhythmia recurrence in the AD group, which may indicate that an active disease status (as reflected by demanding corticosteroid therapy), can exert negative effects on the efficacy of ablation procedure, even off-set the expected protective effects on postablation recurrence by corticosteroid therapy. 19 In addition, long-term treatment with corticosteroids per se is also likely to contribute to high cardiovascular risk, 20,21 incurring an elevated risk of AF recurrence.
Another problem with patients with AD is that the LA function could already be impaired by an inflammatory status. Therefore, it is reasonable to caution against further impairment of atrial function by ablation energy. 22 In our study, the preablation sPAP and mPAP were slightly higher in the AD group than that in the non-AD group. We speculated that an inflammation status combined with a further increase in LA pressure contributed to the higher risk of early recurrence. Further prospective studies are needed to confirm and explore the mechanism of a more frequent early recurrence and establish the relationship between the degree of inflammation status, and impairment of atrial function with adverse outcomes in CA for AF in patients with AD.

| Limitations
Our study has several inherent limitations. First, the limited case number precluded further analysis of the impact of an individual type of AD on procedural outcomes. Considering the component of the AD group in our study, patients with RA and AS were better represented, while the impact of rarer AD-like SLE, systemic sclerosis on the outcome of AF ablation is still worth further investigation as only sparse patients with these diseases were involved in our study.
In addition, levels of biomarkers reflecting the disease activity were lacking in our baseline data, making us unable to establish the correlation between the disease activity of AD and the procedure outcome. Moreover, potential ascertainment bias may exist as patients with ADs generally have more opportunities to get contact with the health care system and increased chances of AF confirmation after AF ablation.

DATA AVAILABILITY STATEMENT
The datasets generated and/or analyzed during the current study are not publicly available due to the protection of patient privacy but are available from the corresponding author upon reasonable request.