Effect of Dexmedetomidine on Atrial Fibrillation-Based Tachyarrhythmia After Cardiac Surgery: A Systematic Review and Meta-Analysis

Background: The incidence of tachyarrhythmia with atrial brillation as the main manifestation increases after adult cardiac surgery, which leads to an increase in adverse events. Dexmedetomidine has been widely used in the perioperative period, but the effect of dexmedetomidine on tachyarrhythmia after cardiac surgery in adults remains controversial. Objective: To evaluate the effect of perioperative use of dexmedetomidine on tachyarrhythmia with atrial brillation as the main manifestation after cardiac surgery. Methods: We searched six databases, including Embase, PubMed, Cochrane, CNKI, Wanfang, and Sinomed, for literature published up to November 2020, without restrictions on language. The primary endpoint was the number of patients with atrial brillation after cardiac surgery. The secondary endpoints included: the number of patients with supraventricular tachycardia, the number of patients with ventricular tachycardia, the number of patients with ventricular brillation, the number of patients with myocardial infarction, the number of dead patients, mechanical ventilation duration, and the length of ICU stay and hospitalization. We used Stata (Version 12.0) and Review Manager (Version 5.3) provided by Cochrane Collaboration for data analysis. If the included studies have high statistical heterogeneity (P ≤ 0.1, I 2 >50%), we will use a random-effects model. Otherwise, a xed-effects model will be used for calculation. Results: Among the 1388 studies retrieved, a total of 18 studies met our inclusion criteria (N=3171 participants). The use of dexmedetomidine reduced the incidence of atrial brillation by 17% (RR=0.83, 95% CI 0.73-0.93; Z=3.06, P=0.002), reduced the incidence of supraventricular tachycardia by about 70% (RR=0.29, 95% CI 0.11-0.77; Z=2.47, P=0.01), reduced the incidence of ventricular tachycardia by about 80% (RR=0.23, 95% CI 0.08-0.63; Z=2.85, P=0.004), but had no effect on the incidence of ventricular brillation (RR=1.02, 95% CI 0.14-7.31; Z=0.02, P=0.99 (cid:0) .There was no signicant difference in the incidence of myocardial infarction between the two groups (RR = 0.90, 95% CI 0.37-2.18; Z = 0.24, P = 0.81). There was no signicant difference in mortality between the two groups (RR = 0.86, 95% CI 0.31-2.44; Z = 0.28, P = 0.78). Dexmedetomidine group can reduce the time of patients in ICU (SMD = - 0.35, 95% CI -0.69 to -0.02; Z = 2.07, P = 0.04), but the heterogeneity between studies is high (I 2 = 93%). There was no effect on duration of mechanical ventilation (SMD = -0.10, 95% CI -0.25 to 0.06; Z = 1.18, P = 0.24) and length of hospitalization (SMD = -0.46, 95% CI -1.08 to 0.16; Z = 1.46, P = 0.14). Conclusion: Dexmedetomidine can reduce the incidence of atrial brillation, supraventricular tachycardia, and ventricular tachycardia after cardiac surgery in adults but does not affect the occurrence of ventricular brillation. infarction; MV: mechanical ventilation; RCT: randomized controlled trial; SVT: supraventricular tachycardia; VF: ventricular brillation; VT: ventricular tachycardia.


Introduction
Atrial brillation (AF) is a tachyarrhythmia characterized by the deterioration of atrial mechanical function due to uncoordinated activities of the atria. The incidence of AF increases after heart operation. It was reported that the incidence of AF is about 30% in patients with coronary artery bypass grafting, 40% in patients with heart valve repair or replacement, and 50% in patients with combined surgery [1]. The exact mechanism of AF remains unclear. Atrial traction, myocardial ischemia and hypoxia, in ammatory reaction, electrolyte disturbance, and conduction system damage during cardiac surgery all increase the incidence of AF [2]. AF may cause a series of adverse events, including discomfort, anxiety, cognitive impairment, hemodynamic deterioration, myocardial infarction, stroke, prolonged hospitalization, and increased medical expenses [3][4][5][6][7].
As we all know, dexmedetomidine (DEX) is a highly selective α2 receptor agonist, which has been widely used in the perioperative period. DEX can reduce sympathetic activity, the tension of the sinus node and atrioventricular node [8], thus slowing down patient's heart rate. However, the effect of DEX on AF after adult heart operation is still controversial. Retrospective study [9,10] found that DEX can reduce the occurrence of AF after heart operation, which was also con rmed by randomized controlled trials (RCTs) [11][12][13][14]. However, other studies [15][16][17][18][19] found that DEX had no in uence on AF incidence after heart operation. Besides, a meta-analysis [20] suggested that DEX may increase the incidence of AF after heart operation in patients with a history of AF. Given these controversies, we sorted out the relevant literature and did this systematic review and meta-analysis.

Search strategy
The Preferred Reporting Items for Systematic Reviews and Meta-analyses Protocols (PRISMA) statement was the guideline during this study's design. This study's protocol has been registered in the International Prospective Register of Systematic Reviews (PROSPERO), and the registration number is CRD42021233613. ))) AND ("dexmedetomidine" [Mesh]). The search was performed with no restrictions on language and focused on the type of RCT. We also reviewed the reference lists of the relevant articles and performed the manual search to identify additional studies. Contact the corresponding author or the rst author when in doubt or if the data is incomplete.

Inclusion criteria for study selection
(1) Participants: The adult patients (≥18 years) who underwent cardiac surgery and were transferred to intensive care unit (ICU) after surgery, without restrictions on the type of surgery and whether they underwent cardiopulmonary bypass. (2) Intervention measures: DEX was used during the perioperative period, without restrictions on the starting time, dose, loading dose, and drug duration. year, number of cases in the two groups, type of surgery, load and maintenance dose of DEX, control group drugs, follow-up time, the number of patients with AF after cardiac surgery, the number of patients with supraventricular tachycardia (SVT), the number of patients with ventricular tachycardia (VT), the number of patients with ventricular brillation (VF), the number of patients with myocardial infarction (MI), the number of reported deaths, the duration of mechanical ventilation (MV), the ICU stay and hospital stay. We pre-extracted the designed data extraction form. If the opinions of the two data extractors were inconsistent and could not be resolved through negotiation, a third researcher (MSW) assisted in the resolution. If data from any category were not reported in the preliminary study, we would contact the corresponding authors or the rst author to request the information. We used the Cochrane bias risk assessment tool to assess the methodological quality of the included studies.

Statistical analysis
We
In one of the articles [25], we contacted the author and obtained data on postoperative AF. The study [12] with the smallest sample size included 60 patients, and the study [29] with the largest sample size included 794 patients. The surgical types of 8 studies were simple coronary artery bypass grafting (CABG) [13, 15-17, 19, 21, 22, 24], 2 studies were simple valve surgery [12,26], and the type of surgery for the remaining studies were adult cardiac surgery including all types [11,18,23,25,[27][28][29]. Five studies were sponsored by pharmaceutical companies [16,17,25,28,29]. For the evaluation of double-blind in the trial's methodological quality, most of the studies were unclear or highrisk, and three studies were "open-label" [11,16,22]. The included articles' baseline characteristics were shown in Table 1, and the risk of bias was shown in Fig. 2 and 3.

Main results
Regarding the incidence of AF, a total of 18 studies with 3171 patients (1603 in the DEX group and 1568 in the control group) were included in the meta-analysis. After cardiac surgery, the total incidence of AF was 22 Fig. 6).
Regarding the incidence of VT, a total of 4 studies [13,16,17,28] with 836 patients (421 in the DEX group and 415 in the control group) were included. The total incidence of VT after cardiac surgery was 2.75%, in which the incidence of the DEX group was 0.95%, and that of the control group was 4.58%. Compared with the control group, the risk of VT in the DEX group was 23% of the control group (RR = 0.23, 95% CI 0.08-0.63; I 2 = 0%; Z = 2.85, P = 0.004; Fig. 7).
Regarding the incidence of VF, a total of 2 studies [22,28] with 388 patients (195 in the DEX group and 193 in the control group) were included. The incidence of VF was 0.52%, and there was no difference between the two groups (RR = 1.02, 95% CI 0.14-7.31; I 2 = 0%; Z = 0.02, P = 0.99; Fig. 8).
We also counted the incidence of MI after surgery. A total of 3 studies [16, 17,29] with 1169 patients (586 in the DEX group and 583 in the control group) were included. The total MI incidence was about 1.63% (1.54% in the DEX group and 1.72% in the control group). The incidence of MI in the DEX group was slightly lower than that in the control group, but there was no statistical signi cance (RR = 0.90, 95% CI 0.37-2.18; I 2 = 23%; Z = 0.24, P = 0.81; Fig. 9).
Regarding the mortality rate, a total of 7 studies [11,13,17,22,23,28,29] with 1679 patients (842 in the DEX group and 837 in the control group) were included. The overall mortality incidence was about 0.66% (0.59% in the DEX group and 0.72% in the control group). The mortality in the DEX group was lower than that in the control group, but there was no statistical signi cance (RR = 0.86, 95% CI 0.31-2.44; I 2 = 0%; Z = 0.28, P = 0.78; Fig. 10).

Discussion
Our study had achieved remarkable results on DEX's effect on the incidence of AF in adult patients after heart surgery.
The meta-analysis of 18 RCTs and 3171 participants con rmed the anti-arrhythmic effect of DEX, which is not only re ected in the effect on AF, but also in the impact on SVT and VT.
Various types of arrhythmia often occur after heart surgery, but AF and malignant ventricular arrhythmia have the most signi cant in uence on postoperative hemodynamics. Studies have con rmed that the autonomic nervous system is closely related to the occurrence of postoperative AF [30]. Stimulation of vagus nerves and sympathetic nerves can cause AF. High-intensity vagal nerve stimulation can induce the atrium and pulmonary veins' electrical remodeling, thus promoting AF [31]. Drugs that increase sympathetic nerve tension can also increase the incidence of postoperative AF, while perioperative use of beta-blockers can decrease the incidence of postoperative AF [32].
Besides, compared with other cardiac surgery, the incidence of AF after heart transplantation is relatively lower [33].
In ammation promotes the occurrence of AF as well. The relationship between AF and in ammatory response was rst proposed by Bruins et al. [34] They found that the peak of postoperative AF in patients undergoing coronary artery bypass grafting is generally 2-3 days after surgery, and C reactive protein also reached its peak at the same time. Korantzopulos et al. [35] found that the white blood cell level of patients with AF gradually decreased after cardioversion and considered that the occurrence of AF was closely related to the white blood cell count. Weymann et al. [36] found that white blood cell count was related to the recurrence of AF and emphasized that the percentage of neutrophils to lymphocytes and the ratio between them were more predictive than white blood cells for the recurrence of AF. Oxidative stress also plays an essential role in the occurrence and maintenance of AF. During cardiac surgery, injured myocardial cells produce many reactive oxygen species, which leads to atrial structural and electrical remodeling, imbalance of redox gene expression, and oxidative damage to mitochondrial DNA [37]. The mechanism of postoperative malignant ventricular arrhythmia is similar to that of AF, which is related to many factors, such as atrioventricular hypertrophy, myocardial ischemia and hypoxia, acid-base imbalance, and electrolyte disorder, pain, drainage tube stimulation, and drug effect. DEX was rst approved by U.S. Food and Drug Administration for sedation in ICU. However, with the in-depth understanding of DEX's pharmacological effects and a series of clinical observations, DEX's protective effects on essential organs such as the heart, brain, lung, and kidney have been gradually con rmed. DEX is a highly selective α2 adrenergic receptor agonist, mainly including, α2B, and α2C, which has speci c effects on α2 receptors in the brain, spinal cord, and periphery. Among them, α2A excitement can regulate autonomic nervous tension, regulate the release of endogenous catecholamines, and produce sedative, analgesic and anti-anxiety effects. The stimulation of α2B can constrict blood vessels, maintain hemodynamic stability, and produce an analgesic effect by inhibiting the transmission of pain signals and reducing substance P release. The antiarrhythmic effect of DEX may be related to regulating the autonomic nervous system, inhibition of in ammatory response, and oxidative stress response. DEX can inhibit the sympathetic nerve, excite the vagus nerve, reduce catecholamine concentration, and inhibit epinephrine's arrhythmogenic effect [13,38]. Animal experiments showed that DEX could reduce calcium in ux, inhibit early and late depolarization, and signi cantly reduce the incidence of VF in rabbits with acquired long QT syndrome [39]. Yu et al. [40] found that DEX can induce myocardial autophagy, improve in ammatory state and inhibit cardiomyocyte apoptosis in rats through the animal model of sepsis induced by cecal ligation and puncture. Besides, DEX can inhibit apoptosis induced by reactive oxygen species and thus has a myocardial protective effect on rat cardiomyocytes [41]. Given the above myocardial protective effects of DEX, our meta-analysis showed consistent results. As for the effect of DEX on VF, we did not get a positive result. The reason was no more than the following two factors. First of all, with the in-depth understanding of patients' pathophysiology with heart disease, the improvement of surgical skills, and the increase of anesthesia, cardiopulmonary bypass, and postoperative nursing team level, the incidence of VF after cardiac surgery is gradually decreasing. Secondly, our meta-analysis research's primary indicator is AF, so collecting information on the incidence of VF is imperfect.
Previous systematic reviews and meta-analysis [20,42] found that DEX did not affect AF incidence after cardiac surgery. However, Liu et al. [43] indicated that DEX could reduce the incidence of AF and VT after cardiac surgery but has no effect on VF incidence (this result was consistent with our research results). By comparison, we found that the literature as mentioned earlier has uniform inclusion criteria, and the types of surgery were CABG or valve replacement surgery, and the language of publication was English. Six of the included randomized controlled trials were different. The reason lay in the different retrieval times and comprehensiveness of the author. The inclusion indicators of our study have slightly changed. The type of surgery included all cardiac operations. The occurrence time of the analysis results was limited after the operation, and there was no restriction on the language of publication. A total of 18 studies involving 3171 patients were included. The number of patients included in the meta-analysis was twice as large as that in the previous meta-analysis, including a 5-year multicenter large-scale randomized controlled study conducted by Turan and his colleagues [29] in 2020. In conclusion, the reliability of our results is signi cantly increased.
Our study also performed a pooled analysis of MI and mortality, and found that DEX did not affect these indicators, which was consistent with the results of other meta-analyses [43,44], but it still needs further study to con rm.
Besides, our study showed that DEX could shorten the time of patients in ICU. However, the studies' heterogeneity is high, which may be related to different surgery types, different critical conditions, different ICU transport standards in the different centers, and whether to use fast-track anesthesia. No positive results were found in this study on the in uence of MV time and hospitalization time. However, Lin et al. [45] believe that DEX can lead to early extubation of patients after CABG. Compared with CABG patients, patients with valve disease, patients undergoing combined cardiac surgery, and patients undergoing large vessel surgery usually have higher postoperative complications and need more prolonged mechanical ventilation and hospitalization time. There are no restrictions on the types of adult cardiac surgery in our study, which may be one reason for not getting positive results.
There are some limitations in our study. First, it is not clear how long the antiarrhythmic effect of DEX can last, and most of the observation time included in the studies is no more than ve days after surgery. Second, some studies have potential bias risks, such as open-label design and pharmaceutical industry funds. Third, this study did not further analyze the effects of DEX dosage and timing on the results. Not all the included studies used load capacity.
The maintenance dose of DEX ranged from 0.04 to 1.5 ug/kg/h; the timing of intervention varies from study to study, including before, during, and even after surgery. The optimal time and dose of DEX intervention are unknown. Future large-sample clinical trials should focus on the above two points to maximize the antiarrhythmic effects of DEX.

Conclusions
Perioperative use of DEX can reduce the incidence of AF, SVT, VT and ICU stay. However, it has no effect on the incidence of VF, MI, and mortality, and has no effect on MV duration and hospital stay. Our research supports the application of DEX in the perioperative period of adult cardiac surgery, which provides a new way of thinking for the prevention and treatment of arrhythmia after adult cardiac surgery. WZW and JL conducted the primary data search, analyzed data and wrote the manuscript. HBY conducted the data analysis and helped in manuscript preparation. MSW solved the classi cation discrepancies in data collection designing the study and in the primary search and substantially contributed in manuscript preparation and data interpretation. TW designed the study, and drafted the manuscript. All the authors approved the nal version of the paper and agree to be accountable for all aspects of the work, thereby ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All authors read and approved the nal manuscript.
Funding: There is no funding source.
Availability of data and material: The datasets supporting the conclusions of this article are included within the article and its additional le.
Ethics approval and consent to participate: Not applicable.

Consent for publication: Not applicable
Competing interests: The authors of the manuscript claimed no con icts of interest.

Study
Number of patients      Forest plot of myocardial infarction Figure 10 Forest plot of mortality Figure 11 Forest plot of ICU stay Figure 12 Forest plot of mechanical ventilation time Figure 13 Forest plot of hospital stay Figure 14 Funnel plot of ICU stay Figure 15 Funnel plot of mechanical ventilation time