Predictive value of HMGB1 for atrial brillation recurrence after catheter ablation in paroxysmal atrial brillation patients

Background: To evaluate the clinical ecacy of serum high-mobility group box-1 (HMGB1) released from the left atrium to predict atrial brillation (AF) recurrence in paroxysmal AF (pAF) patients after catheter ablation (CA) at 1-year follow-up. Methods: We included 72 pAF patients who underwent CA. To determine the expression levels of HMGB1, left atrial blood samples were collected from the patients prior to CA and after the procedure through the transseptal sheath placed into the left atrium. Patients were followed-up for AF recurrence after 1 year of CA. Results: A total of 19 (26%) patients of the 72 experienced AF recurrence. No signicant differences were noted in the clinical baseline data between the AF recurrence and AF nonrecurrence groups. The level of postoperative HMGB1 (HMGB1post) was higher in the AF recurrence group than in the AF nonrecurrence group (298.51µg/L vs. 278.17 µg/L; P = 0.03). However, no differences were noted in the levels of other biomarkers such as preoperative high-sensitivity C-reactive protein (hs-CRPpre), postoperative hs-CRP (hs-CRPpost), and preoperative HMGB1 (HMGB1pre) between the two groups. Multiple logistic regression analysis revealed that a higher level of serum HMGB1post from the left atrium (HMGB1post: ≥ 279.35 µg/L) was an independent predictor of AF recurrence (odds ratio [OR]: 5.29 [1.17–23.92], P = 0.04). Receiver operating characteristic (ROC) analysis revealed that HMGB1post had a moderate predictive power for AF recurrence (area under the curve [AUC]: 0.68; sensitivity: 72%; specicity: 68%). The 1-year AF-free survival was signicantly lower in patients with a high serum HMGB1post level than in those with a low HMGB1post level (hazard ratio [HR]: 3.81 [1.49–9.75],


Background
Atrial brillation (AF), a frequent type of supraventricular arrhythmia, is the most common atrial arrhythmia among adults worldwide, and its prevalence has been increasing (1). AF is a progressive disease, and its early stage is paroxysmal AF (pAF); more than 50% of the pAF patients progress to persistent AF or die within 10 years (2). Thus, early intervention in pAF is of great signi cance.
Currently, catheter ablation (CA) is recommended for patients with symptomatic drug refractory pAF (3,4), especially because CA techniques for AF have signi cantly evolved over the past decade (5). However, a substantial AF recurrence is common during follow-up, which overshadows the potential bene ts of the procedure (6). To stratify the risk of AF recurrence in pAF patients undergoing CA, several risk factors such as Age (> 75 years), AF duration, Congestive heart failure, Hypertension, Diabetes, previous Stroke/transient ischemic attack-vascular disease, CHA2DS2-VASc score, left atrial size, and presence of structural heart disease are considered (7)(8)(9). However, all these factors show great population heterogeneity for AF recurrence prediction.
Recently, in ammatory processes were reported to play important roles in AF recurrence (10). A study found that the level of high-sensitivity C-reactive protein (hs-CRP) at baseline has an independent prognostic value in predicting AF recurrence (11). However, the study population included both persistent AF and pAF patients. Moreover, hs-CRP, a nonspeci c acute-phase reactant synthesized by hepatocytes, is easily affected by other comorbidities (12). High mobility group protein 1 (HMGB1) is both a nuclear factor and secreted protein. The release of HMGB1 by necrotic cells can trigger in ammation (13). The expression levels of HMGB1 were signi cantly higher in AF patients and in the left atrium than in normal patients and in the periphery, respectively (14). A review concluded that HMGB1 may be released by the left atrium, and it may mediate in ammatory processes after CA and in uence the occurrence and development of AF (15).
Although these studies have presented the association between HMGB1 and AF, the possibility of HMGB1 released from the left atrium predicting AF recurrence remains unclear. Therefore, this study investigated the predictive value of serum HMGB1 released from the left atrium for AF recurrence.

Study population
This study included 72 symptomatic patients with pAF who underwent CA at the First Hospital of the China Medical University between August 2017 and August 2018. All patients were older than 18 years, suitable for CA, and had voluntarily participated in the study. AF was diagnosed with a 12-lead electrocardiogram (ECG) or Holter-ECG. The diagnostic standard was in accordance with the pAF de nition of the American Heart Association (16). Exclusion criteria were presence of persistent AF, severe heart failure, severe valvular disease, thyroid dysfunction, hepatic and renal dysfunction, LA thrombosis, malignant tumor, and autoimmune and in ammatory diseases or systemic disease and a 6-month history of surgery. Written informed consent was obtained from all patients before enrollment, and the study complied with the Declaration of Helsinki.

Echocardiographic measurements
All patients underwent transthoracic echocardiography within 72 hours before CA. Transthoracic echocardiography was performed using the Vivid E9 ultrasound system (GE Healthcare, Waukesha, WI, USA). The diameter of the left atrium was measured in the parasternal long-axis view on the 2dimensional image. The Simpson's biplane method was used to measure left ventricular ejection fraction (LVEF) as well as left atrial volume in the apical 4-chamber and 2-chamber views. Notably, the left atrial appendage should not be included in tracing of the endocardial border. Left atrial volume index (LAVI) was calculated by dividing the estimated end-systolic left atrial volume by the body surface area. All measurements were performed in accordance with the American Society of Echocardiography guidelines (17).

Ablation protocol
Transesophageal echocardiography was used to exclude any left atrium or left atrial appendage thrombosis. Multidetector computed tomography of pulmonary vein (PV) was performed 24 hours prior to CA to assess PV anatomy. Antiarrhythmic drugs were discontinued at least 5 half-lives before CA and anticoagulation drugs were administered during and after CA. The Seldinger technique was used to obtain venous access and perform uoroscopy-guided interatrial septal puncture. Given that cryoballoon ablation (CBA) is a simpler and shorter procedure with shorter uoroscopy time than radiofrequency ablation (RAF), the rst preference of patients with obvious symptoms and frequently arrhythmia occurring was CBA [ (18)(19)(20)]. In case of RFA, an open-irrigated catheter tip was used to perform pulmonary vein isolation. In case of CBA, a second-generation cryoablation balloon was placed at the ostium of each PV. Balloon selection was based on the PV size, as established by preoperative PV computed tomography. Further details about the CA procedure can be found in the international expert consensus that has been accepted in recently (21). The ablation was considered successful if all tested pulmonary potentials with a mapping catheter were completely blocked between the atrial tissue and PV.
Blood sample collection and post-processing Transseptal puncture was performed using a transseptal sheath during CA. After access to the left atrium was achieved using the sheath, blood samples for determining the preoperative HMGB1 (HMGB1pre) level were obtained before the formal procedures of CA. Similarly, after successful CA, blood samples were immediately collected through the sheath placed in the left atrium. Thereafter, the blood samples were centrifuged at 3000 rpm for 10 minutes, and the serum samples were stored at − 80 °C until analysis. The in ammatory biomarkers, hs-CRP and HMGB1, were measured using commercial enzymelinked immunosorbent assay (Boster Biological Technology, Wuhan, China) kits according to the manufacturer's instructions.

Follow-up
After CA, all patients underwent continuous electrocardiographic monitoring until discharge and were prescribed oral anticoagulation for at least 3 months based on the CHA2DS2-VASc score. The endpoint was 1-year AF recurrence after CA. Episodes of AF or atrial tachyarrhythmia lasting more than 30 s diagnosed with a 12-lead ECG or Holter ECG within 3 months were classi ed as early AF recurrence and those after 3 months as AF recurrence (16).

Statistical analysis
The baseline data were strati ed by pAF recurrence after CA. Continuous data are expressed as means ± standard deviation or medians (interquartile range), and categorical data are expressed as counts and percentages. Intergroup comparisons were performed using unpaired Student t-tests for normally distributed variables and the non-parametric Mann-Whitney U-test for non-normally distributed variables. Categorical variables were collated using the chi-square test or Fisher's exact test. Multivariate regression analysis was performed to analyze the risk factors for the prediction of AF recurrence. The results are expressed as odds ratios (ORs) and 95% con dence intervals (CIs). Optimal cutoff values were determined by analyzing the sensitivity and speci city values derived from the receiver operating characteristic (ROC) curve data. The area under the curve (AUC) was calculated to assess the discriminatory power of meaningful biomarkers. The AF-free survival rates were determined using Kaplan-Meier curves and compared using the log-rank test. Data were analyzed using SPSS ver. 25 (IBM SPSS Statistics for Windows, Version 25.0, Armonk, NY, USA). The results of the t-test and chi-square test were validated by bootstrapping (1000 repetitions). A P value of < 0.05 was considered signi cant.

Study population
Nineteen (26%) of the 72 pAF patients experienced AF recurrence, whereas 53 (74%) patients did not experience postoperative recurrence. Characteristics of patients were strati ed by recurrence and nonrecurrence of AF. Age, sex, AF duration, CHADS-VASC score, concomitant diseases, biochemical index, and echocardiographic parameters (LAD, LVEF, and LAVI) were not signi cantly different between the AF recurrence and AF nonrecurrence groups. However, the early recurrence rates were signi cantly higher in the AF recurrence group than in the AF nonrecurrence group (P = 0.001). The results were consistent after bootstrapping (1000 times; Table 1). The values shown are mean ± SD or percentages.
Early recurrence refers the episode of symptomatic or asymptomatic documented atrial tachyarrhythmia > 30 seconds in the blanking period.

Laboratory test
The results revealed that no obvious differences existed between the two groups for the preoperative hs-CRP (hs-CRPpre) and HMGB1pre levels. The levels of hs-CRP and HMGB1 in the left atrium were signi cantly elevated after ablation. Moreover, the HMGB1post levels were obviously higher in the AF recurrence group than in the AF nonrecurrence group (298.51 µg/L vs. 278.17 µg/L; P = 0.03). However, the postoperative hs-CRP (hs-CRPpost) levels did differ between the two groups (Table 2).
Further, because two types of ablation procedures (RFA and CBA) were performed, we compared the association of HMGB1 between the two groups (RFA group and CBA group). The results revealed that in the RFA group, the HMGB1pre and HMGB1post levels did not differ between the AF recurrence and AF nonrecurrence groups ( respectively) between the AF recurrence and AF nonrecurrence groups ( Table 2).  Table 3). ROC analysis revealed that serum HMGB1post released from the left atrium (AUC: 0.68; sensitivity: 72%; speci city: 68%) had a moderate predictive power for AF recurrence and the cutoff value of serum HMGB1post was calculated to be 279.35 µg/L (Fig. 1)

Discussion
AF is the most common arrhythmia among adults and pAF is the most common type of AF (22). Currently, CA is recommended for symptomatic drug-refractory AF patients (3). Totally blocked electrical potential between the atrial tissue and PV is the cornerstone of CA, which is most often achieved by RFA or CBA. Previous studies have reported that the AF recurrence rates did not differ between pAF patients who underwent RFA and CBA (23)(24)(25). According to a nationwide Danish cohort study, with up to 1-year follow-up, the rate of AF recurrence after ablation was 45% in 2005-2006 and 31% in 2013-2014 (26). Therefore, the discovery of a biomarker that can predict AF recurrence after CA would be bene cial for clinical management. We found that the left atrial serum HMGB1post level, which was measured immediately after CA, was associated with AF recurrence in pAF patients who underwent CA.
HMGB1 is both a nuclear factor and secreted protein that can be released into the extracellular space by activated immune cells or damaged cells during infection or injury. In addition, HMGB1 could be an in ammatory mediator that participates in the pathogenesis of in ammatory diseases (13). After CA, HMGB1 expression increases, causing a continuous state of oxidative stress and in ammation process (27). Further, HMGB1 may bind to HMGB1 receptor (Toll-like receptor-2,4) located in the left atrium and mediate atrial remodeling (15). In conclusion, after CA, HMGB1 is released into serum by damaged myocardial cells of the left atrium. In addition, HMGB1 acts as a pro-in ammatory factor in atrium remodeling, making atrial arrhythmia more likely to occur. This might explain the moderate predictive power of HMGB1post released from the left atrium for AF recurrence. No signi cant difference was noted between the AF recurrence and AF nonrecurrence groups for the hs-CRPpre and hs-CRPpost levels, which may be because hs-CRP is easily affected by other comorbidities (12). This is similar to the ndings of Masson S study that in pAF patients with a history of AF and without signi cant left ventricular dysfunction or heart failure, hs-CRP is a weak predictor of AF recurrence (28).
Early recurrence of AF (ERAF) could be a transient phenomenon due to in ammation or a foretaste of late AF recurrence due to CA lesion reconnection (29). Recently, increasing number of studies have reported that the incidence rate of ERAF is highly related to late AF recurrence after CA, irrespective of RFA (30) or CBA (31). Similarly, we found that a higher ERAF rate was related to higher AF late recurrence at 1-year follow-up.
Several studies have indicated that enlarged left atrial size is a potential predictor of AF recurrence with AF patients after CA (32,33). However, in our study, no difference was noted in the left atrial size of the AF recurrence and AF nonrecurrence group patients. The mean LAVI was 33.42 mL/m2 and LVEF was 62.70%, indicating that our patients had only slight structural impairment of the left atrium. Therefore, the left atrial size did not have a predictive value in our study.
To clarify the association between HMGB1 and AF recurrence, we evaluated the clinical e cacy of HMGB1 released from the left atrium for predicting AF recurrence. In contrast to previous studies, we collected blood samples from the left atrium instead of peripheral vessels to verify the effect of HMGB1. CA damages the myocardium and hence, HMGB1 is released by the injured cardiomyocytes directly into the left atrium (13). Collection of blood samples from the left atrium was aimed to emphasize the in uence of local in ammation caused by HMGB1 on AF recurrence.
Our study had several limitations. First, our sample size was small and therefore, we adopted the bootstrapping method (1000 repetitions) to compensate for the small sample size. We intend to increase our sample size and follow-up our patients for further study. Second, two CA procedures, RFA and CBA, were performed in our study, which contributed to the disequilibrium of the study population. However, both the HMGB1pre and HMGB1post levels were not signi cantly different between the AF recurrence and AF nonrecurrence group patients who underwent RFA as well as CBA. Therefore, the effect of this asymmetry was relatively small. Third, some AF events were asymptomatic and our patients were intermittently followed-up depending on patient-reported symptoms. Thus, the AF recurrence rates might have been underestimated, which can be improved by shortening the follow-up duration.

Conclusion
Our study found that in pAF patients undergoing CA, HMGB1post released from the left atrium was associated with AF recurrence and had a moderate predictive power for AF recurrence. Therefore, we offer a potential predictor to identify patients at high risk of AF recurrence. This predictor is helpful for clinical decision-making and postoperative management of AF patients.

Declarations
Ethics approval and consent to participate The study protocol was approved by the China Medical University Ethics Committee (No: AF-SOP-07-1.1-01). All patients provided written informed consent before any study-speci c procedure was performed.

Consent for publication
Not applicable Availability of data and materials Not applicable. The study is currently ongoing on patient enrollment and data collection. Thus, no datasets were generated for analysis yet.

Funding
This work was supported by the National Natural Science Foundation of China (Project number 81871373). The funders did not participate in the design of the study, the acquisition and analysis of the data or in drafting the manuscript.