Efficacy of N-acetylcysteine in Preventing Acute Kidney Injury and Major Adverse Cardiac Events After Cardiac Surgery: A Meta-Analysis and Trial Sequential Analysis

Background The effect of N-acetylcysteine (NAC), an antioxidant, on preventing acute kidney injury (AKI) and major adverse cardiac events (MACE) remains controversial. Therefore, we conducted this meta-analysis and trial sequential analysis to evaluate its efficacy on cardiac surgery-related adverse events. Methods PubMed, Embase, and Cochrane Library were searched for relevant studies from inception to June 2021. We selected randomized controlled trials comparing NAC with controls in patients undergoing cardiac surgery. Results Twenty-five studies including 2,444 patients met the inclusion criteria. The pooled results showed that there was no significant difference in the incidence of AKI between the NAC and control groups [relative risk (RR) = 0.91, 95% confidence interval (CI) = 0.77, 1.08, P = 0.28], but the trial sequential analysis (TSA) could not confirm this result. No difference was observed in the need for renal replacement therapy (RRT), all-cause mortality, MACE, length of stay in the intensive care unit (ICU), and length of stay in the hospital. Results of subgroup analysis results showed that intravenous infusion instead of oral NAC could significantly reduce the incidence of AKI and arrhythmia (RR = 0.84, 95% CI = 0.71, 0.99, P = 0.03, I2 = 3% and RR = 0.74, 95% CI = 0.61, 0.91, P = 0.004, I2 = 48%, respectively). Conclusion Intravenous administration of NAC can reduce the incidence of AKI and arrhythmia in patients after cardiac surgery, but cannot reduce all-cause mortality, AMI, cardiac insufficiency, and the number of patients using RRT. Oral NAC has no significant effect on the outcomes of patients after cardiac surgery.

Results: Twenty-five studies including 2,444 patients met the inclusion criteria. The pooled results showed that there was no significant difference in the incidence of AKI between the NAC and control groups [relative risk (RR) = 0.91, 95% confidence interval (CI) = 0.77, 1.08, P = 0.28], but the trial sequential analysis (TSA) could not confirm this result. No difference was observed in the need for renal replacement therapy (RRT), all-cause mortality, MACE, length of stay in the intensive care unit (ICU), and length of stay in the hospital. Results of subgroup analysis results showed that intravenous infusion instead of oral NAC could significantly reduce the incidence of AKI and arrhythmia (RR = 0.84, 95% CI = 0.71, 0.99, P = 0.03, I 2 = 3% and RR = 0.74, 95% CI = 0.61, 0.91, P = 0.004, I 2 = 48%, respectively).
Conclusion: Intravenous administration of NAC can reduce the incidence of AKI and arrhythmia in patients after cardiac surgery, but cannot reduce all-cause mortality, AMI, cardiac insufficiency, and the number of patients using RRT. Oral NAC has no significant effect on the outcomes of patients after cardiac surgery.
Keywords: N-acetylcysteine, acute kidney injury, major adverse cardiac events (MACEs), cardiac surgery, trial sequential analysis BACKGROUND Coronary artery disease (CAD) is one of the major cardiovascular diseases worldwide and many CAD patients will be submitted for cardiac surgeries. The most common cardiac surgeries are coronary artery bypass graft (CABG) surgery and aortic or mitral valve repair or replacement (1,2). They are commonly performed on-pump, which indicates that cardiopulmonary bypass (CPB) is used in the surgery. However, there is also a considerable number of cardiac surgeries that are performed off-pump without CPB assistance (3). Although the overall prognosis after cardiac surgery has been improved over the past decades, the occurrence of acute kidney injury (AKI) and major adverse cardiac events (MACEs) remains unsatisfactorily high (4,5). AKI is the most common important complication in adult patients undergoing cardiac surgery and is associated with a prolonged hospital stay, use of dialysis, subsequent chronic kidney disease (CKD), and increased mortality (4). The incidence of AKI occurs in approximately 18% of patients undergoing cardiac surgery and approximately 2%-6% of them require renal replacement therapy (RRT). It is more likely for these kinds of patients to progress to CKD in the ensuing months and years than those who do not develop AKI and do not require RRT. A variety of risk factors, either renal or extrarenal, contribute to the development and progression of AKI after heart surgery, including renal ischemia, reperfusion, mechanical trauma, inflammation, hemolysis, oxidative stress, cholesterol emboli, and nephrotoxins (6). MACE is also an important composite primary endpoint assessed in most cardiovascular trials. Despite these unsatisfactory adverse events, cardiac surgery remains popular in CAD patients for its irreplaceable therapeutic effect. Thus, we must develop a more effective method to reduce the risk of postoperative complications of cardiac surgery.
N-acetylcysteine (NAC) is a cysteine prodrug and glutathione (GSH) precursor which has been used in clinical therapeutic practice as a mucolytic agent and for the treatment of numerous disorders including paracetamol intoxication, doxorubicin cardiotoxicity, ischemia-reperfusion cardiac injury and chemotherapy-induced toxicity associated with GSH deficiency for several decades (7). Also, NAC is a kind of free radical scavenger antioxidant agent and it is now well-known that it can reduce pro-inflammatory cytokines, oxygen free-radical production, and ameliorates ischemia-reperfusion injury which may consequently reduce postoperative complications in cardiac surgery (8). Researchers have investigated the efficacy and safety of NAC in several clinical trials in recent years. However, its effectiveness remains controversial. Although previous meta-analyses have investigated the role of NAC in preventing post-cardiac surgery complications, the results of them are conflicting (9)(10)(11)(12)(13)(14)(15). Therefore, we conducted this meta-analysis and trial sequential analysis (TSA) to further evaluate the efficacy of NAC in preventing AKI and MACEs after cardiac surgery.

METHODS
This meta-analysis adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA statement) guidelines (16).

Data Sources and Search Strategy
Electronic databases including PubMed, Embase, and the Cochrane Library were systematically searched from inception to June 2021, using items related to "n-acetylcysteine, " and "cardiac surgery." The search was limited to studies involving human subjects. No language restrictions or publication status were applied. The citations of included references were searched individually to identify potential additional relevant studies.

Eligibility Criteria
The inclusion criteria were as follows: (1) study design: randomized controlled trials (RCTs); (2) population: patients undergoing cardiac surgery (>18 years old); (3) intervention: n-acetylcysteine (NAC) compared with placebo or standard of care; and (4) outcome: assessed at least one of the following outcomes: the incidence of acute kidney injury (AKI), the need for renal replacement therapy (RRT), all-cause mortality, major adverse cardiac events (MACEs) including arrhythmia, cardiac insufficiency and acute myocardial infarction (AMI), length of stay in an intensive care unit (ICU) and hospital. The exclusion criteria were as follows: (1) studies that involved participants who are < 18 years old; (2) studies that evaluated different interventions or did not include a reference group; (3) studies that did not report predefined outcomes or the data could not be extracted.

Data Extraction and Risk of Bias Assessment
Two reviewers independently extracted the data using a predefined standardized form. The extracted data included first author, year of publication, sample size, patient characteristics, interventions, all clinical outcomes (the incidence of AKI, the need for RRT, all-cause mortality, MACEs including arrhythmia, cardiac insufficiency, and AMI, length of stay in ICU and length of stay in hospital). The risk of bias assessment was performed by two independent reviewers using the Cochrane risk of bias approach and a third reviewer was consulted if no consensus could be reached. The standard criteria included the following domains: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other biases.

Statistical Analysis
Outcomes were treated as dichotomous or continuous variables. We calculated risk ratios (RRs) with 95% confidence intervals (CIs) for dichotomous variables and mean differences (MDs) with 95% CIs for continuous variables. Heterogeneity among the studies was assessed by the I 2 statistic, and when I 2 was more than 50%, significant statistical heterogeneity was considered to be present (17). We used the fixed-effect model when the I 2 values were < 25%. Otherwise, we used the random-effects model. Sensitivity analyses were conducted to test the robustness of the overall pooled effect. The presence of publication bias was evaluated by using a funnel plot. All comparisons were two-sided, and a P < 0.05 was considered statistically significant. If the mean or standard deviation of the outcomes could not be directly extracted from the studies, we estimated them from the sample size, median, range, and/or interquartile range (18,19).
Review Manager (version 5.3, The Cochrane Collaboration, Oxford, United Kingdom) was used in all analyses.

Trial Sequential Analysis
We conducted trial sequential analysis (TSA) in this metaanalysis to control the risk of random errors and assess whether the results were conclusive (20). Firm evidence for accepting or rejecting the anticipated intervention effect is considered clear and no further studies are needed if the cumulative Z-curve crossed the trial sequential monitoring boundary or entered the futility area. No conclusion is made and more studies are required to confirm the results if the Z-curve did not cross any of the boundaries or the required information size (RIS) has not been reached (21). In our meta-analysis, we performed the TSA with an overall risk of 5% of the type I error and estimated the RIS based on a RR reduction of 20% with a power of 80%. The control event rate was calculated according to the comparator group (22).

Search Results and Study Characteristics
The flow chart describing the selection of the trials for this meta-analysis is presented in Figure 1. According to our search strategy, 531 potential studies were identified. After removing the duplicates and the studies that failed to meet the inclusion criteria, 33 studies were eligible for full-text reviews. Finally, only twenty-five (23-47) studies involving 2,444 patients were included in this meta-analysis. The characteristics of the included individual studies were summarized in Table 1. All included trials were reported between 2003 and 2018. The sample sizes of the included trials ranged from 20 to 295.

Risk of Bias Assessment and Publication Bias of the Included Studies
Review Manager 5.3 was used to assess the study quality in this study. A summary of the risk of bias in the included studies is presented in Figure 2. Random sequence generation was judged to be at a low risk of bias in all included studies. There is no significant publication bias of mortality (P = 0.652 for the Begg's test, P = 0.475 for the Egger's test), length of stay in ICU (P = 0.086 for the Begg's test, P = 0.163 for the Egger's test) and length of stay in hospital (P = 0.475 for the Begg's test, P = 0.181 for the Egger's test).

The Incidence of AKI
Fifteen studies (23-25, 28-31, 36, 38, 39, 42, 43, 45-47) reported the incidence rate of AKI. The data from the trials showed that there was no significant difference in the incidence of AKI between the NAC groups and the controlled groups (RR = 0.91, 95%CI = 0.77, 1.08, P = 0.28, I 2 = 15%) as shown in Figure 3A. However, the TSA could not confirm this result because the  cumulative Z-curve did not cross the conventional boundary or the trial sequential monitoring boundary and did not cross the futility boundary ( Figure 3B).

All-Cause Mortality
Eighteen (23-25, 28, 29, 31-33, 35, 36, 38-42, 44, 46, 47) of the included studies reported all-cause mortality. The mortality in the NAC group and the controlled group was 2.2 and 3.6%, respectively. The pooled results showed that the use of NAC could not reduce the risk of all-cause mortality compared with the use of placebo or just standard of care (RR = 0.72, 95% CI = 0.41, 1.25, P = 0.24, I 2 = 0%) as presented in Figure 5.

Effect of NAC Administration Methods on Outcomes
All the studies were divided according to NAC administration methods, and the effects on the outcome were analyzed respectively. The results showed that intravenous infusion instead of oral NAC could significantly reduce the incidence of AKI and arrhythmia (RR = 0.84, 95% CI = 0.71, 0.99, P = 0.03, I 2 = 3% and RR = 0.74, 95% CI = 0.61, 0.91, P = 0.004, I 2 = 48% respectively) as shown in Figures 8A,B. The addition of NAC to cardioplegia may reduce ICU and hospital stay, but significant statistical heterogeneity was observed. NAC has no significant effect on RRT, all-cause mortality, AMI, and cardiac insufficiency. Results of subgroup analysis showed administration methods of NAC had no significant effect on the need for RRT, all-cause mortality, AMI, and cardiac insufficiency among patients ( Table 2).

DISCUSSION
Our meta-analysis showed some novel findings. The incidence of AKI in postoperative cardiac patients who received intravenous rather than oral NAC treatment was significantly lower than that in the control group. Intravenous injection of NAC can reduce the incidence of arrhythmia in patients after cardiac surgery while no difference was found in the incidence between intravenous injection and oral subgroups. Conversely, NAC has no significant effect on RRT, all-cause mortality, AMI, and cardiac insufficiency.
The intravenous NAC to patients with cardiac surgery is associated with a lower incidence of AKI, and the efficacy of oral NAC in preventing CI-AKI is inconclusive. Cardiac surgeryrelated acute renal injury (CSA-AKI) is the most common major complication of cardiac surgery (48). Although there is no consensus on the definition of post-cardiac surgery ARI, most of the fifteen studies stated that an increase in serum creatinine concentration > 25% from baseline is one of the conditions for the diagnosis of AKI. NAC has been shown to reduce the level of oxidative stress and reduce acute renal failure induced by ischemia-reperfusion. It can also improve renal aging and renal interstitial fibrosis through Sirtuin 1 activation and p53 deacetylation (49). There is no conclusive evidence supporting the benefit of the administration of NAC to prevent CSA-AKI in previous meta-analyses but our subgroup analysis reveals different results (12,13). Of all the 15 studies that reported the incidence of AKI, 9 were administered intravenously with NAC, 6 were done orally. Results of a subgroup analysis according to the methods of administration show that the effect of NAC is related to the administration methods. Several possible mechanisms caused these differences such as dosage and duration. We listed the treatment protocol of NAC in each study in Table 1. The dosage of NAC by intravenous injection is often much larger than that by oral administration. For example, Aldemir gave patients 150 mg.kg −1 I.V. in 15 min after anesthesia induction, followed by 50 mg.kg −1 .4h −1 and 100 mg.kg −1 .16h −1 . While the drug was administered at 600 mg PO twice daily for a total of 14 doses in Adabag's study. Furthermore, oral NAC was first absorbed through the digestive system, which may have a first-pass elimination effect, resulting in a decrease in blood concentration. It is worth mentioning that high-dose NAC is effective, but it has been also demonstrated to frequently cause adverse effects (50,51). Before a large-scale application of NAC after cardiac surgery is recommended, further research is needed to explore the optimal dosage and method.
This meta-analysis shows that intravenous administration of NAC after cardiac surgery may reduce the incidence of  arrhythmias. Baker and other studies found that the incidence of postoperative atrial fibrillation in the NAC group was 36% lower than that in the control group. But they did not explore the effect of drug use on the outcome (52). On the one hand, this confirms that NAC can significantly reduce the incidence of arrhythmias after cardiac surgery. On the other hand, there is a significant correlation between the occurrence of postoperative atrial fibrillation and the level of postoperative inflammation. NAC may reduce the occurrence of arrhythmias by inhibiting inflammatory storms. In addition, we found that there was no significant difference in all-cause mortality, risk of AMI and cardiac insufficiency, rate of renal replacement therapy, and length of stay between the NAC group and the control group. This is consistent with previous studies (53). Given that we have included more RCT trials, the results may apply to a larger population.
Previous meta-analyses evaluating this topic have been published. However, there are several differences between the present study and the previous works. The present analysis includes the most RCT experiments representing the latest and most comprehensive study and we conducted a subgroup analysis of NAC for the first time based on the administration methods of NAC. Our meta-analysis also has some limitations. First, we only assessed the impact of NAC on the incidence of AKI but failed to assess the impact of the use of NAC on AKI to varying degrees because the vast majority of studies did not classify AKI or used different grading criteria. Second, a limitation is imposed on this meta-analysis because the relevant literatures included different degrees of differences such as the inclusion criteria of the study population and the dose and time of the use of NAC. These shortcomings may affect the results. Third, there is a lack of long-term follow-up results for patients, and a shortage of further subdivided MACEs, such as subgroup analysis of different types of arrhythmias.

CONCLUSION
To sum up, this meta-analysis suggests that intravenous administration of NAC can reduce the incidence of AKI and arrhythmia in patients after cardiac surgery, but cannot reduce all-cause mortality, AMI, cardiac insufficiency, and the number of patients using RRT. Oral NAC has no significant effect on the outcomes of patients after cardiac surgery.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.