Epicardial Adipose Tissue and Postoperative Atrial Fibrillation

Background Atrial fibrillation (AF) often occurs after cardiac surgery and is associated with increased risk of stroke and mortality. Prior studies support the important role of inflammation in the pathogenesis of postoperative atrial fibrillation (POAF). It is known that an increased volume and a pro-inflammatory phenotype of epicardial adipose tissue (EAT) are both associated with AF onset in non surgical context. In the present study, we aim to evaluate whether also POAF occurrence may be triggered by an increased production of inflammatory mediators from EAT. Methods The study population was composed of 105 patients, with no history of paroxysmal or permanent AF, undergoing elective cardiac surgery. After clinical evaluation, all patients performed an echocardiographic study including the measurement of EAT thickness. Serum samples and EAT biopsies were collected before surgery. Levels of 10 inflammatory cytokines were measured in serum and EAT conditioned media. After surgery, cardiac rhythm was monitored for 7 days. Results Forty-four patients (41.3%) developed POAF. As regard to cardiovascular therapy, only statin use was significantly lower in POAF patients (65.1% vs. 84.7%; p-0.032). Levels of Monocyte Chemoattractant Protein-1 (MCP-1), in both serum and EAT, were significantly higher in POAF patients (130.1 pg/ml vs. 68.7 pg/ml; p = <0.001; 322.4 pg/ml vs. 153.4 pg/ml; p = 0.028 respectively). EAT levels of IL-6 were significantly increased in POAF patients compared to those in sinus rhythm (SR) (126.3 pg/ml vs. 23 pg/ml; p = <0.005). Conclusion Higher EAT levels of IL-6 and MCP-1 are significantly associated with the occurrence of POAF. Statin therapy seems to play a role in preventing POAF. These results might pave the way for a targeted use of these drugs in the perioperative period.


INTRODUCTION
Atrial fibrillation (AF) is the most common arrhythmia occurring after cardiac surgery. The incidence of postoperative AF (POAF) ranges between 20% and 50% according to the type of surgical procedure, with higher rates after valve replacement or repair respect to isolated coronary artery bypass grafting (CABG) surgery (1)(2)(3)(4). Combined procedures (CABG and valvular surgery) present the highest incidence of this complication, reaching up to 80% (5).
POAF occurrence determines a significant increase in stroke risk, morbidity and mortality with consequent increase of hospitalization time and healthcare costs (6).
Although the pathogenesis of POAF remains uncertain, accumulating evidence suggest an important role of inflammatory mechanisms and mediators. In particular, the interleukin-6 (IL-6), recognized as a primary cytokine in the inflammatory cascade, has been identified as one of the main molecules involved in the development of POAF (4,(7)(8)(9). Interestingly, the epicardial adipose tissue (EAT), the visceral fat depot of the heart, produces numerous pro-inflammatory cytokines which can affect the myocardium through paracrine or vasocrine mechanisms (10,11). Moreover, increased EAT thickness is associated with higher levels of secreted inflammatory mediators and with the onset of atrial fibrillation (10,12).
The aim of this study was to evaluate the correlation between EAT secretory profile and POAF occurrence in patients undergoing cardiac surgery.

METHODS
Study population: the study population included 105 patients without history of paroxysmal or permanent atrial fibrillation, undergoing elective surgery for CABG or valve replacement for severe aortic stenosis at the cardiac surgery unit of University of Naples "Federico II". The presence of chronic inflammatory diseases and/or cancer represented exclusion criteria, given their association with systemic and/or visceral fat inflammation. Demographic and clinical data including drug therapies were collected from all patients. The study protocol was compliant to the ethical guidelines of the 1975 Declaration of Helsinki. All the study procedures received approval by our institution's human research committee (Protocol n. 301/19). All patients provided written informed consent before their inclusion into the study.
Transthoracic Echocardiography: before cardiac surgery, all patients underwent complete echocardiographic study, performed with a VIVID E9 (GE Healthcare) machine. In addition to the standard parameters, the maximum EAT thickness was evaluated, from the parasternal long-axis view, at end systole, between the right ventricle and the ascending aorta (13). Measurements of EAT thickness were performed offline by two independent blinded echocardiographers. The average value from three cardiac cycles was used for the statistical analysis.
Tissues and serum collection: serum samples and EAT biopsies were collected from all patients undergoing cardiac surgery before the cardiopulmonary bypass (CPB). EAT biopsy samples (average 0.1-0.5 g) were taken between the free wall of the right ventricle and the anterior surface of the ascending aorta. EAT secretomes were obtained as follows: tissues were weighted, cut into small pieces, and transferred into a 12-well plate. According to tissue weight, serum-free Dulbecco modified Eagle medium (DMEM) (1 mL medium/0.1 g tissue) was added to the well and incubated at 37 • C in a CO 2 incubator. After 24 h, medium was collected and centrifuged at 14,000g to remove debris and analyzed for cytokines content, as described below.
ECG monitoring: after surgery, heart rate and rhythm were monitored for 7 days, by continuous telemetry (ApexPro 7lead system; General Electric Medical Systems), at the cardiac intensive care unit. Atrial fibrillation has been termed as irregularly atrial rhythm without clear P waves that was confirmed by a 12-lead ECG. In this study, POAF was defined as any episode of atrial fibrillation lasting more than 5 min, with or without symptoms requiring intervention to maintain hemodynamic stability, arisen in the seven days following the cardiac surgery. POAF episodes recorded in condition of hemodynamic and volemic balance were considered for analysis. We excluded POAF episodes potentially related to a sudden fluid loss (diuretic administration, postoperative bleeding, etc. . . ), low blood oxygenation or intravenous high inotropic dose administration.
Anesthesia and surgical technique: surgical technique and perioperative management were the same for all patients according to the specific surgical procedure. Perioperative anesthesiologic management was the same in all cases: according to the institutional protocol, surgical anesthesia was obtained with continuous intravenous infusion of Propofol + Remifentanil + Cisatracurium, while fluid balance was managed paying attention to the hemodynamic conditions, in order to obtain a mean arterial pressure of at least 70 mmHg. Perioperative drugs management was carried out according to the 2017 EACTS Guidelines on perioperative medication in adult cardiac surgery (14). Before CPB, heparin was intravenously administered at a dose of 300 units/Kg in all cases; protamine need was assessed using the HMS Plus Hemostasis Management system (Medtronic, Minneapolis, MN, USA). Transesophageal echocardiography was routinely performed before the surgical incision in order to assess myocardial and cardiac valves function during and after the surgical procedure. All patients underwent surgery through a standard full sternotomy approach and hypothermic CPB. In order to optimize the surgical times and to avoid confounding factors, all tissue collections were performed before the heparin administration and the placement of extracorporeal circulation cannulas. Surgical excision of EAT was performed from the fat pad of right ventricle infundibulum near the atrioventricular groove, using only a surgical scalpel blade no. 11 (without using diathermy) in order to prevent any additional inflammatory damage. After collection, all biopsies were placed in a sterile pipe and quickly transferred to the laboratory to preserve EAT secreting activity. Extracorporeal circulation was performed through aortic and atrio-caval cannulation. All patients requiring high doses of inotropic drugs during their intensive care unit stay were excluded from the study due to the known pro-arrhythmogenic effect (Epinephrine or Norepinephrine > 0.1 µg/Kg/min or Dobutamine > 5 µg/Kg/min). At the end of surgery, all patients were moved into the cardiac surgery intensive care unit and weaned from the mechanical ventilation after at least 2 h of general postoperative monitoring. Fluid intake was regulated to achieve a central venous pressure of at least 10 mmHg according to the cardiac anatomy and myocardial function. All patients received MgSO 4 continuous intravenous infusion in a dose of 17.5 gr in the first 24 h after the surgery as anti-arrhythmia prophylaxis. Packed red cells were transfused only in presence of serum Hemoglobin lower than 8 g/dl while Fresh frozen plasma was used as plasma expander only in case of postoperative bleeding.
Statistical analysis: all statistical analyses were conducted using the statistical platform R (vers. 4.0). Standard descriptive statistics were used to describe the sample. Mean ± standard deviation (std. dev.) with range in case of numerical variables and absolute frequencies and percentages for categorical factors. Numerical variables showing highly skewed distribution were described using median, interquartile range (25-75th percentile). Accordingly, between-groups comparisons were based on the chi-square test (or Fisher exact test where appropriate), the ttest for independent samples, or the Mann-Whitney U test. To account for imbalances between the two groups, the inflammatory mediators' levels were log-transformed and the difference between groups were assessed using a linear model where age, statin use, atrial volume and E/e ′ were entered as covariates.
All tests were two-sided with a p value<0.05 denoting statistical significance. Due to exploratory nature of all the analyses, no adjustments were made for multiple comparison.

EAT and Serum Inflammatory Profile
IL-1β, IL-1ra, IL-6, IL-8, IL-13, FGF, IFN-γ, MCP-1, RANTES/CCL5 and TNF-α were detected in serum samples and EAT conditioned media obtained from all patients. No significant differences were found in serum levels of the pro-inflammatory mediators, except for MCP-1 which was significantly higher in patients who developed POAF than in those who remained in SR (130.1 pg/ml vs. 68.7 pg/ml; p = ≤0.001; Table 2). Interestingly, also EAT levels of MCP-1 were significantly higher in patients with POAF (322.4 pg/ml vs. 153.4 pg/ml, p = 0.028; Figure 1; Table 3) even after adjusting the analysis for age, statin use and atrial volume, using general linear model (p = 0.008). Moreover, EAT levels but not serum levels of IL-6 were significantly increased in patients who developed POAF compared to those in SR patients (126.3 pg/ml vs. 23 pg/ml; p = 0.005; Figure 1; Table 3). The difference remained statistically significant in adjusted analysis (p = 0.043).

DISCUSSION
POAF represents one of the most frequent complication after cardiac surgery and accounts for a significant increase of stroke risk, disability and mortality. Although many intraoperative and clinical factors seem to be involved in the development of POAF, its pathogenesis remains still unclear. Recent evidence suggest that inflammation can alter atrial conduction, facilitating multiple re-entry wavelets, thus predisposing to the development of POAF (15,16). EAT, the visceral fat depot of the heart, located between the myocardium and the visceral layer of pericardium, is mainly represented in the atrio-ventricular and inter-ventricular grooves, and along the lateral wall of the right ventricle (11). In pathological conditions, EAT produces and secretes numerous inflammatory mediators (12,17). The absence of fascial boundaries allows a tight connection with surrounding tissues and coronary arterial vessels. Numerous lines of evidence   suggest a strong association between increased EAT thickness and atrial fibrillation (11). This is one of the few studies evaluating the potential association between the levels of EAT-secreted pro-inflammatory cytokines and the development of POAF in patients undergoing elective cardiac surgery, without history of paroxysmal or persistent atrial fibrillation. Viviano et al. have previously described EAT secretome as a possible substrate for POAF indicating the role of gelsolin, involved in inflammation and ion channel regulation, in maintaining sinus rhythm in these patients (18).
Consistent with previous studies, we have found that patients who developed POAF were older than those who remained in sinus rhythm throughout the postoperative period (15,19). It is known that aging is characterized by a chronic low-grade inflammation and is associated with degenerative and inflammatory modifications in atrial anatomy, such as dilation and fibrosis, which are responsible for the alterations of the atrial electrophysiological properties (20)(21)(22). The EAT thickness increases with age and the release of proinflammatory adipocytokines from cardiac visceral fat into the systemic circulation may contribute to the inflammatory state, which in turn promotes the accumulation and inflammation of EAT (22,23). Furthermore, it is well known that extracorporeal circulation is also characterized by a systemic inflammatory response (4,24). In the present study, EAT thickness tended to be higher in POAF compared to non POAF patients, although the difference between groups did not reach the statistical significance. To note, EAT thickness mean value in POAF population was higher than that reported as cut-off value (10 mm) in a recent study from our group that validated, against cardiac magnetic resonance, the echocardiographic assessment of EAT thickness at the level of the Rindfleisch fold (13).
Significantly higher levels of MCP-1, both in serum and in EAT secretomes, were found in patients who developed POAF. The expression of the gene coding for MCP-1 can be induced by a variety of mediators, including numerous interleukins, platelet derived growth factor, and vascular endothelial growth factor (25). MCP-1 is mainly produced by monocytes and macrophages and exerts potent chemotactic and activating effects on CCR2positive leukocytes. Several studies have reported that serum MCP-1 levels are independently associated with atrial fibrillation (7,25).
IL-6 levels were significantly higher in EAT secretomes of patients who developed POAF. Numerous evidence have demonstrated elevated serum levels of IL-6 in patients who develop POAF (25)(26)(27). IL-6 is a pleiotropic cytokine with a variety of biological activities. It is produced not only by immune cells and immune accessory cells including monocytes and macrophages, but also by endothelial cells, vascular smooth-muscle cells, adipocytes and ischemic cardiomyocytes. It stimulates the synthesis of several acute-phase reaction proteins (7). Of note, in our study, EAT production of this proinflammatory cytokine was higher in POAF patients and resulted significantly associated with atrial arrhythmia. Interestingly, Mazurek et al. have previously described the upregulation of MCP-1 and IL-6 in the EAT of CAD patients (17). These previous data reinforce our opinion that these two cytokines may contribute to the EAT pro-inflammatory phenotype. Although left atrial volume is known to be an important factor in conditioning AF occurrence, no differences in this parameter were found between POAF and SR patients. Furthermore, differences in EAT levels of MCP-1 and IL-6 between POAF and SR patients remained significant even after adjusting the analysis for age, statin use prior to surgery, left atrial volume and E/e ′ .
We also excluded the other possible causes of POAF, such as electrolyte imbalance or acid-base disturbances. Interestingly, statin intake was lower in patients who developed POAF compared with those who remained in SR with a higher use of atorvastatin in SR compared to POAF patients. We hypothesize that this result could be ascribed to the known anti-inflammatory and pleiotropic effect of statins. A randomized, controlled trial (ARMYDA-3) demonstrated that treatment with 40 mg of atorvastatin daily for 7 days significantly reduces the incidence of new-onset POAF and shortens the length of hospital stay in patients undergoing cardiac surgery with CPB (28). Moreover, a recent meta-analysis from Yuan et al., including 20 randomized controlled trials of patients who underwent cardiac surgery, concluded that preoperative statin therapy might be promising for the prevention of POAF, especially for patients undergoing isolated CABG surgery (29). Contrarily, a large randomized controlled trial by Zheng et al. (30) did not demonstrate an association between rosuvastatin assumption and reduction of POAF incidence, thus indicating conflicting evidence on this issue. In this regard, our group has previously demonstrated, both in vivo and in vitro, that statin administration reduces EAT secretion of IL-6 and IL-8 levels in patients with aortic stenosis (31).

CONCLUSIONS
The present study has evaluated the association of the pro-inflammatory secretory profile of EAT with the onset of POAF in patients undergoing cardiac surgery. It is plausible that the inflammatory substrate is mainly promoted by the EAT secretion of IL-6 and MCP-1. Further studies are needed to establish whether statin therapy could play a protective role, mediated by the reduction of EAT levels of IL-6, thus paving the way for a targeted use of these drugs in the peri-operative period. Further studies will be needed to confirm these findings and investigate the role of cardiac visceral fat in the pathogenesis of POAF.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Ethical Committee of Federico II. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
LP has enrolled patients, has collected clinical/anamnestic data, has performed echocardiograms, and elaborated the manuscript. MC and PC have performed echocardiograms. GC and EP have performed epicardial adipose tissue biopsies during cardiac surgery. SC and PF have analyzed serum samples and epicardial fat biopsies for the concentration of proinflammatory interleukins. CR and IA have enrolled patients and have collected clinical/anamnestic data. DB has performed the statistical analysis. NF, DL, and VP have designed the project and have revised the manuscript. All authors read and approved the final manuscript.