Role of glucagon-like peptide-1 agonist in patients undergoing percutaneous coronary intervention or coronary artery bypass grafting: A meta-analysis

Abstract

Importance

Glucagon-like peptide-1 (GLP-1) protects against ischemia-reperfusion injury in patients with acute myocardial infarction (AMI). Controversy exists on the effects of GLP-1 on AMI patients undergoing percutaneous coronary intervention (PCI) and coronary artery bypass graft (CABG) surgery.

Study objective

We aimed to investigate the cardioprotective effects of GLP-1 in AMI patients after PCI and CABG.

Design

We searched PubMed, Web of Science, EBSCO, Scopus, and the Cochrane Library for relevant randomized controlled trials (RCTs) up to June 2021, with no restriction on publication date. The following search terms are used: “percutaneous coronary intervention” or “coronary artery bypass grafting” or “myocardial infarction” and “glucagon-like peptide 1” or “exenatide” or “liraglutide”.

Study selection

Articles were independently assessed by 2 reviewers. We included RCTs only that compared GLP-1 with control in AMI patients.

Data extraction and synthesis

Continuous data were pooled as mean differences (MDs), while dichotomous variables were pooled as odds ratios (ORs), with 95% confidence interval (CI), using R software (meta package) for windows. Subgroup analysis according to the intervention type and GLP-agents were conducted. We assessed the heterogeneity among RCTs using the Q statistic and I² statistic. We also tested publication bias by funnel plot–based methods. The quality of each study was assessed with the Cochrane risk of bias tool.

Main outcomes and measures

Primary outcomes were changes of left ventricular ejection fraction (LVEF), myocardial infarct characteristics, salvage index. Secondary outcomes included major adverse cardiac events (MACE), gastrointestinal events, and hypoglycemia.

Results

Nine RCTs (14 reports) including 1216 patients were included in this meta-analysis. At 3 months follow up, GLP-1 was associated with improved LVEF (MD = 2.81, 95% CI [0.69, 4.94]), infarct size in grams (MD = −5.71, 95% CI [−10.24, −1.18]), and salvage index (MD = 0.09, 95% CI [0.05, 0.14]). While, GLP-1 had less MACE rate than control (RR = 0.64, 95% CI [0.41, 0.99]), and higher gastrointestinal side effects (RR = 4.21, 95% CI [2.39, 7.41]).

Conclusions and relevance

This meta-analysis illustrated that GLP-1 was associated with better LVEF and reduced infarct size in patients with AMI undergoing PCI and CABG surgery, although the mechanism on how this agent provide this benefit is not clear.

Key points

Question: What is the effectiveness of Glucagon-like peptide-1 (GLP-1) agonist in patients with acute myocardial infarction (AMI) undergoing percutaneous coronary intervention (PCI) and coronary artery bypass graft (CABG) surgery.

Findings: This systematic review and meta-analysis illustrated that GLP-1 was associated with better left ventricular ejection fraction and reduced infarct size in patients with AMI undergoing PCI and CABG surgery, probably by reducing reperfusion injury.

Meaning: GLP-1 could improve systolic and diastolic function, lowering the cardiovascular risk of morbidity and mortality in AMI patients.

1. Introduction

Acute myocardial infarction (AMI) is a major cause of mortality and morbidity. Effective reperfusion therapies with percutaneous coronary intervention (PCI) and coronary artery bypass graft (CABG) surgery have decreased long-term cardiac mortality [1], [2]. Nevertheless, rapid blood flow restoration may provoke the death of cardiomyocytes. This phenomenon, known as reperfusion injury, is a contributing factor to adverse prognosis [3]. In the last years, several cardiovascular protective treatments were believed to prevent or reduce ischemia-reperfusion injury [4].

Glucagon-like peptide-1 (GLP-1), a natural antidiabetic hormone secreted in the distal intestine, plays an important role in the regulation of blood glucose by increasing insulin secretion, prolonging gastric emptying, and increasing satiety sensation. In addition to improved blood glucose profiles and body weight, animal studies revealed that GLP-1 had anti-apoptotic properties and corroborated to be cardioprotective [5], [6]. Previous clinical trials showed evidence of cardioprotective effects following GLP-1 administration in patients with type 2 diabetes mellitus [7]. However, controversy exists on the cardiovascular protective effects of GLP-1 agents in patients with AMI against both ischemia and reperfusion injuries.

We conducted the current systematic review and meta-analysis of randomized controlled trials (RCTs) to test the effectiveness of GLP-1 infusion on left ventricular function, myocardial infarct characteristics, and major adverse cardiovascular events in patients with AMI undergoing PCI or CABG surgery.

2. Materials and methods

All steps of the current study were carried out according to the Cochrane handbook of systematic reviews of interventions in addition to PRISMA statement guidelines. The protocol of this meta-analysis was published online at the PROSPERO International Prospective Register of Systematic Reviews under registration number (CRDXX). No institutional review board approval was required for this systematic review.

2.1. Databases and search strategy

Relevant studies were identified through searching PubMed, Web of Science, EBSCO, Scopus, and the Cochrane Library up to June 2021. The search terms were: (“percutaneous coronary intervention” OR “PCI” OR “coronary artery bypass grafting” OR “CABG” OR “myocardial ischemia” OR “coronary disease” OR “myocardial infarction”) AND (“glucagon-like peptide 1” OR “GLP-1” OR “exenatide” OR “liraglutide” OR “albiglutide” OR “taspoglutide” OR “dulaglutide” OR “lixisenatide” OR “semaglutide”). In addition, the reference lists of included studies were screened manually to identify additional potentially relevant articles.

2.2. Eligibility criteria

We assessed the included RCTs independently, and studies were considered eligible according to the PICO approach: P, adults patients diagnosed with coronary artery disease undergoing PCI or CABG; I, GLP-1 infusion or its receptor agonists at the time of PCI or CABG surgery and followed up for 1 month at least; C, comparison with placebo or insulin; and O, primary outcomes including changes of left ventricular ejection fraction (LVEF) between baseline and 3 months follow up, thrombolysis in myocardial infarction (TIMI) flow grade 3 (complete perfusion), troponin levels, myocardial infarct characteristics, salvage index, secondary outcomes included major adverse cardiac events (MACE) and other side effects such as gastrointestinal side effects and hypoglycemia.

Exclusion criteria included review articles, observational studies, case reports, comments or guidelines, animal studies, non-English articles, and insufficient data to calculate. Disagreements were solved by consensus.

2.3. Data extraction

For each paper, we extracted the following information: first author's name, publication year, research period, country, study design, number of patients, mean age, male percentage, type of GLP-1, dose, the plasma concentration of GLP-1, control, imaging methods, body mass index (BMI) (kg/m²), percentage of diabetes and hypertension, and related outcomes.

2.4. Assessment of bias risk

We used Cochrane collaboration's tool for assessing the risk of bias of included RCTs [8]. Risk of bias assessment included the following domains: 1) sequence generation, 2) allocation sequence concealment, 3) blinding of participants and personnel, 4) blinding of outcome assessment, 5) incomplete outcome data, 6) selective outcome reporting and 7) other potential sources of bias; the authors' judgment is categorized as ‘Low risk’, ‘High risk’ or ‘Unclear risk’ of bias.

2.5. Statistical analysis

Continuous data, as LVEF, were pooled as mean difference (MD) with 95% confidence interval (CI). Dichotomous data, as MACE outcome, were pooled as risk ratio (RR), with 95% CI. Heterogeneity was assessed by visual inspection of the forest plots and measured by Q statistic and I² statistic. Significant statistical heterogeneity was indicated by Q statistic P-value less than 0.1 or by I² more than 50%. In case of significant heterogeneity, a random effect model was employed. Otherwise, the fixed effect model was used. P < 0.05 was considered statistically significant, and statistical tests were two-sided. Further, we conducted subgroup analysis according to the intervention type (PCI or CABG) and GLP-1 agent (exenatide or liraglutide). The funnel plot method was used to assess publication bias. All meta-analyses of the enrolled studies were done in R version 4.0.3 (meta package) for Windows.

3. Results

3.1. Study screening

A total of 1825 relevant papers were initially identified using the search strategy. Some 375 duplicated papers were excluded by EndNote software, along with another 1421 papers that did not fit the criteria for analysis after the title and abstract screening. Based on full-text screening, 14 papers (nine articles) were finally included. The PRISMA flow process of study selection is shown in Fig. 1. A list of excluded studies after full-text screening is reported in Table 1 in the Supplement.

Fig. 1.

The PRISMA flow process of study selection.

3.2. Study characteristics and risk of bias of the included studies

Ultimately, nine RCTs (14 reports) including 1216 patients were included in this meta-analysis. Seven RCTs (11 reports) [9], [10], [19], [11], [12], [13], [14], [15], [16], [17], [18] included 593 patients undergoing PCI while two RCTs (three reports) [20], [21], [22] included 142 patients undergoing CABG surgery. Six RCTs (9 reports) [14], [15], [16], [17], [18], [19], [20], [21], [22] used exenatide (n = 722 patients) and two RCTs (four reports) [9], [10], [11], [12] used liraglutide (n = 302 patients), while one study [13] used native GLP-1 (7-36) amide rather than a pharmaceutical GLP-1 receptor agonist (produced by Bachem). The reason for using this rather than a GLP-1 RA such as exenatide was that breakdown products of human GLP-1 (7-36) such as GLP-1 (9-36) have been shown to exert cardiovascular effects in animal studies. All studies were published between 2012 and 2019; the number of patients per study ranged from 38 to 387 patients. Seven studies included patients with ST-elevation MI [9], [11], [12], [14], [15], [16], [17], [19] while one study included patients with non-ST elevation MI [10]. Follow-up duration ranged from 3 to 65 months. The percentage of patients with diabetes mellitus was 15%, while patients with hypertension represented 45% of the total patients included in this meta-analysis. The major characteristics of enrolled patients in each RCT are detailed in Table 1. The summary of risk of bias assessment is shown in Fig. 2.

Table 1.

Summary of included studies.

Author	Country	Study design	Study duration (months)	Type of intervention and Patients no.	Age (years, Mean ± SD)	Male, n (%)	Type of GLP-1, dose	Plasma concentration of GLP-1	Control	Imaging methods	BMI (kg/m2)	Diabetes, n (%)	Hypertension, n (%)
Giblett et al., 2019	UK	RCT	13	PCI, 193 (96/96)	T: 69.0 ± 9.5; C: 67.1 ± 8.2	T: 76 (79.2%); C: 76 (79.2%)	GLP-1, 1.2 pmol/kg/min for 30 min before reperfusion, up to 100 μg	NR	Placebo	NR	T: 28.4 [25.5–31.1]; C: 28.0 [25.5–30.1]	T: 9 (9.4%); C: 10 (10.4%)	T: 42 (43.8%); C: 40 (41.7%)
Besch et al., 2018	France	RCT	6	CABG surgery, 92 (49/43)	T: 71 [63–75]; C: 69 [61–76]	T: 46 (94%); C: 36 (84%);	Exenatide; 0.05 μg/min for 1 h, then constant infusion 0.025 μg/min	NR	Insulin	Echo	NR	T: 10 (20%); C: 9 (21%)	T: 29 (59%); C: 30 (70%)
Besch et al., 2017	France	RCT	6	CABG surgery, 104 (53/51)	T: 70 ± 9; C: 68 ± 11	T: 49 (92%); C: 41 (80%)	Exenatide; 0.05 μg/min for 1 h, then constant infusion 0.025 μg/min	NR	Insulin	Echo	NR	T: 12 (23%); C: 10 (20%)	T: 33 (62%); C: 36 (71)
Lipš et al., 2017	Czech	RCT	3	CABG surgery, 38 (19/19)	T: 66.6 ± 9.4; C: 65.9 ± 9.1	T: 14 (73.7%); C: 17 (89.5%)/	Exenatide; 10 μg 12 h before surgery and continuing for 72 h	NR	Placebo	Echo	T: 28.9 ± 6.5; C: 29.3 ± 5.8	T: 13 (68.4%); C: 13 (68.4%)	NR
Chen et al., 2016	China	RCT	3	PCI, 77 (39/38)	T: 57.1 ± 13.0; C: 58.7 ± 12.7	T: 27 (69%); C: 26 (68%)	Liraglutide; 0.6 mg for 2 days, 1.2 mg for 2 days and 1.8 mg for 3 days	0.03–0.3 nmol/L	Placebo	MRI	T: 25.2 ± 3.4; C: 25.4 ± 3.2	T: 5 (13%); C: 7 (18%)	T: 4 (10%); C: 5 (13%)
Chen et al., 2016	China	RCT	3	PCI, 90 (45/45)	T: 58.0 ± 11.7; C: 59.0 ± 12.1	T: 34 (76%); C: 32 (71%)	Liraglutide; 0.6 mg for 2 days, 1.2 mg for 2 days and 1.8 mg for 3 days	0.03–0.3 nmol/L	Placebo	Echo	NR	T: 9 (20%); C: 13 (28%)	T: 27 (60%); C: 29 (64%)
Kyhl et al., 2016 and Lonborg et al., 2012	Denmark	RCT	65	PCI, 334 (175/159)	T: 62 ± 11; C: 63 ± 12	T: 142 (81%); C: 122 (77%)	Exenatide; 0.12 μg/min for the first 15 min and 0.043 μg/min for 6 h	0.03–0.30 nmol/L	Placebo	MRI	T: 27 ± 4; C: 27 ± 4	T: 12 (7%); C: 17 (11%)	T: 62 (35%); C: 62 (39%)
Roos et al., 2016	Netherlands	RCT	4	PCI, 91 (42/49)	T: 57.2 ± 10.2; C: 57.5 ± 10.1	T: 33 (79%); C: 36 (73%)	Exenatide; 10 μg/h for 30 min before reperfusion, and 0.84 μg/h for 72 h	0.14 nmol/L	Placebo	MRI	T: 27.5 ± 4.1; C:26.4 ± 3.2	T: 0 (0%); C: 0 (0%)	T: 9 (23%); C: 6 (13%)
Chen et al., 2015	China	RCT	3	PCI, 210 (105/105)	T: 58.2 ± 11.5; C: 57.4 ± 11.3	T: 71 (68%); C: 67 (64%)	Liraglutide; 1.8 mg for 30 min before reperfusion	0.03–0.3 nmol/L	Placebo	Echo	T: 25.6 ± 3.3; C: 25.3 ± 3.2	T: 17 (16%); C: 21 (20%)	T: 59 (56%); C: 55 (52%)
Chen et al., 2015	China	RCT	3	PCI, 92 (45/47)	T: 57.7 ± 11.3; C: 59.2 ± 14.4	T: 33 (73%); C: 30 (64%)	Liraglutide; 0.6 mg for 2 days, 1.2 mg for 2 days and 1.8 mg for 3 days	0.03–0.3 nmol/L	Placebo	Echo	T: 25.5 ± 3.5; C:25.6 ± 3.9	T: 9 (20%); C: 7 (16%)	T: 23 (51%); C: 23 (51%)
Woo et al., 2013	Korea	RCT	6	PCI, 116 (39/77)	T:61.2 ± 10.8; C: 59.1 ± 11.6	T: 30 (77%); C: 63 (82%)	Exenatide; 10 μg before reperfusion, and 10 μg BID on 2 days	NR	Placebo	Echo + MRI	T:24.9 ± 3.0; C:25.3 ± 3.1	T: 10 (26%); C: 22 (28%)	T: 22 (56%); C: 43 (56%)
Lonborg et al., 2012	Denmark	RCT	65	PCI, 387 (196/191)	T: 63 ± 10.5; C: 61.5 ± 8.5	T: 127 (65%); C: 159 (83%)	Exenatide; 10 μg for 15 min before reperfusion, and 0.043 μg/min for 6 h	0.03 and 0.3 nmol/L	Placebo	MRI	NR	T: 1 (1%); C: 9 (4%)	T: 23 (12%); C: 9 (13%)
Bernink et al., 2012	Netherlands	RCT	4	PCI, 39 (19/20)	T: 60 ± 10; C: 58 ± 8	T: 16 (84%); C: 15 (75%)	Exenatide; 2 μg for 30 min, and 20 μg/24 h for 72 h	NR	Placebo	MRI	T: 27 ± 4; C: 27 ± 4	NR	T: 8 (42%); 6 (30%)

RCT: randomized controlled trial; GLP-1: Glucagon-Like Peptide 1; PCI: percutaneous coronary intervention; CABG: coronary artery bypass grafting; BMI: body mass index; Echo: echocardiography; T: treatment group; C: control group; NR: not reported; MRI: magnetic resonance imaging; BMI: body mass index.

Fig. 2.

The risk of bias summary of included RCTs.

3.3. Outcomes

3.3.1. LVEF

Changes in LVEF between baseline and 3 months follow-up were reported in five trials (n = 351 patients). The effect size showed that GLP-1 was associated with a higher LVEF compared to control (MD = 2.81, 95% CI [0.69, 4.94]). Pooled studies were homogenous (I² = 0%, p = 0.5), Fig. 3.

Fig. 3.

Forest plot comparing GLP-1 vs control in terms of changes of LVEF at 3 months.

3.3.2. Final TIMI 3

Final TIMI 3 flow rates were reported in four trials (n = 567 patients). The effect size showed no significant difference regarding final TIMI 3 between GLP-1 and control (RR = 0.99, 95% CI [0.89, 1.10]) and pooled studies were heterogenous (I² = 71%, p = 0.02), eFig. 1 in the Supplement.

3.3.3. Troponin levels

Troponin levels were reported by five trials (n = 637 patients). No significant differences were found between GLP-1 and control regarding six-hour cardiac troponin I (MD = 0.15, 95% CI [−0.82, 1.13]. While GLP-1 was associated with a lower 24-h cardiac troponin I (MD = −0.82, 95% CI [−1.28, −0.36]). Pooled studies were homogeneous at six- and 24-h cardiac troponin (I² = 22%, p = 0.28 and I² = 0%, p = 0.53, respectively), eFig. 2 in the Supplement.

3.3.4. Myocardial infarct characteristics

Five studies (n = 326 patients) reported data on myocardial infarct characteristics in patients undergoing PCI. The pooled estimate showed that GLP-1 was associated with reductions in terms of final infarct size in grams (MD = −5.71, 95% CI [−10.24, −1.18]), final infarct size in percentage of the left ventricle (MD = −4.61, 95% CI [−8.17, −1.06]), and infarct size (g)/are at risk (g) (MD = 0.10, 95% CI [−0.14, −0.06]), Fig. 4.

Fig. 4.

Forest plot comparing GLP-1 vs control in terms of myocardial infarct characteristics.

3.3.5. Salvage index

Two studies (n = 182 patients) reported on salvage index. The effect size showed that GLP-1 was associated with larger myocardial salvage index compared to control (MD = 0.09, 95% CI [0.05, 0.14]). Pooled studies were homogenous (I² = 0%, p = 0.49), eFig. 3 in the Supplement.

3.3.6. MACE rate

Six studies (n = 1015 patients) reported on MACE rate. The pooled estimate showed that GLP-1 was associated with less MACE risk than control (RR = 0.64, 95% CI [0.41, 0.99]). Pooled studies were homogenous (I² = 0%, p = 0.93), eFig. 4 in the Supplement.

3.3.7. Side effects

Six studies (n = 426 patients) reported on gastrointestinal side effects (nausea, vomiting, and diarrhea) and the pooled estimate showed that GLP-1 was associated with a higher risk of gastrointestinal side effects than control (RR = 4.21, 95% CI [2.39, 7.41]), eFig. 5 in the Supplement. Five studies (n = 386 patients) reported on hypoglycemia and the effect estimate showed no significant difference between GLP-1 and control (RR = 1.57, 95% CI [0.86, 2.87]). Pooled studies were homogenous regarding gastrointestinal side effects (I² = 0%, p = 0.65) and hypoglycemia (I² = 0%, p = 0.98), eFig. 6 in the Supplement.

3.4. Subgroup analysis according to GLP-1 agent; exenatide or liraglutide

We stratified the included studies according to the GLP-1 agent; exenatide or liraglutide. The results of subgroup analysis are consistent with overall results except for LVEF, MACE, TIMI flow, and gastrointestinal side effects. No significant difference was found between exenatide and control regarding LVEF change at 3 months (MD = 1.16, 95% CI [−1.74, 4.06]) and MACE risk (RR = 0.90, 95% CI [0.36, 2.29]). Liraglutide was associated with higher TIMI flow 3 than control (RR = 1.13, 95% CI [1.01, 1.26]). Further, there was no significant difference between liraglutide and control regarding gastrointestinal side effects (RR = 2.61, 95% CI [0.53, 12.78]). The results of subgroup analysis according to GLP-1 agent (exenatide or liraglutide) are shown in eFigs. 7 through 12 in the Supplement.

3.5. Subgroup analysis according to intervention type; PCI or CABG

We stratified the included studies according to the type of intervention PCI or CABG. The results of subgroup analysis are consistent with overall results except for LVEF and gastrointestinal side effects which there was no significant difference between GLP-1 and control in patients undergoing CABG (for LVEF: MD = −0.92, 95% CI [−22.82, 20.98] and for gastrointestinal side effects: RR = 0.98, 95% CI [0.00, 476.89]). The results of subgroup analysis according to the type of intervention PCI or CABG are shown in eFigs. 13 through 15 in the Supplement.

3.6. Publication bias

The funnel plot method was used to assess any evidence of publication bias. The results showed no funnel plot asymmetry in the included outcomes; hence, no publication bias was detected. The funnel plots of include outcomes are shown in eFigs. 16 through 20 in the Supplement.

4. Discussion

This meta-analysis of RCTs provides updated evidence on the cardiac outcomes of GLP-1 in patients with AMI undergoing PCI or CABG. It included 14 reports and 1216 patients. It confirms the cardioprotective effects of GLP-1, including the improvement of LVEF, cardiac troponin levels, final infarct size, and salvage index. Moreover, the results showed that GLP-1 infusion was associated with less MACE rate compared to control.

The exact mechanism on how GLP-1 provide this benefit is not fully understood, however, previous animal studies suggest that their cardioprotective effect could be related to their role in reduction of apoptotic processes, metabolism, and anti-inflammatory effects [6], [23], [24]. During ischemia and reperfusion, GLP-1 causes downstream phosphorylation and inhibition of the proapoptotic protein Bcl-xL/Bcl-2 in the myocardial cells [25]. These effects are modulated via the G-protein-coupled receptor that activates intracellular signaling pathways, particularly the Reperfusion Injury Salvage Kinase (RISK) pathway [25], [26], and stimulates mitochondrial K + -ATP channels [27]. Thus, GLP-1 offers a vital role in protecting mitochondrial function after ischemia and reperfusion injuries. Moreover, GLP-1 activates phospholipase C, inositol triphosphate, and diacylglycerol; thus, the production of protein kinase C, which is an important signaling pathway in the cardioprotective effects [28]. With diminishing myocardial apoptosis and oxidative stress, GLP-1 receptors might provide an anti-inflammatory effect at the time of ischemia and reperfusion injuries [26].

Infarct size has been considered a key factor of cardiovascular outcomes, such as cardiac mortality, reinfarction, and congestive heart failure. Our findings showed that GLP-1 caused significant reductions in the final infarct size in patients undergoing PCI or CABG. By infarct size reduction, GLP-1 could improve systolic and diastolic function, lowering the cardiovascular burden of risk [15]. The results of the current meta-analysis are consistent with previous reports [29], [30], [31]. Administration of GLP-1 receptor agonist was associated with a reduction in infarct size of approximately 40%, as postconditioning by PCI [32]. Further, GLP-1 treatment reduced the final infarct size by 30% in patients with short system delay (less than 132 min), and not in patients with a system delay of more than 132 min [17]. In patients with myocardial infarction with single blood vessel disease, Lønborg et al. reported that GLP-1 infusion (25 μg) for 6 h before and after PCI improved myocardial salvage index [18]. An improved salvage index is associated with an improved clinical outcome in patients with AMI [33]. The present meta-analysis included a relatively low percentage of patients with diabetes mellitus (15% of included sample size). Previous research showed that GLP-1 might be more useful in patients with diabetes mellitus, as glucose control might improve clinical outcomes [34]. Because of the preclinical evidence, it is unlikely that GLP-1 mediated cardioprotective effects are exclusively present in patients with diabetes mellitus.

On the other hand, Besch et al. failed to find cardioprotective effects of GLP-1 administration after CABG surgery [21]. Bernink et al. showed a trend toward a smaller infarction size following GLP-1 infusion; however, not significant difference [16]. Another trial by Roos et al. reported no benefit of GLP-1 on top of PCI in patients with ST myocardial infarction [14]. The lower GLP-1 dose prescribed in the previous studies could be insufficient to generate a significant cardioprotective effect; thus, the GLP-1 dose seems to have an essential role in mediating the protective effects [15].

Our results revealed that GLP-1 was associated with a less rate of MACE (cardiovascular death, nonfatal myocardial infarction, and nonfatal stroke) than control. These results are consistent with those of a meta-analysis of 37 RCTs that reported a significant reduction in the MACE rate in the GLP-1 group compared with placebo [35]. Contrary, Hirshberg et al. suggested that GLP-1 had no significant effect on MACE risk0 compared with placebo [36]. Similar results were reported in a meta-analysis by Huang et al. [37]

Subgroup analysis showed that liraglutide was associated with better LVEF, while exenatide had no significant effect on LFEF. This might be due to the variations of treatment duration between exenatide and liraglutide RCTs. In the exenatide trials, treatment duration ranged from 6 to 72 h, while it was 7 days in liraglutide trials. The shorter treatment duration in exenatide RCTs might contribute to the lower cardioprotective effect [37]. The imaging methods for evaluation of the cardiac function differed in the different RCTs. Two liraglutide trials were conducted with echocardiography, whereas exenatide studies used magnetic resonance imaging (MRI). Although previous research demonstrated that the functional parameters and mass interrelated closely between the echocardiography and MRI [38], [39], the variation in imaging methods may represent a potential source of heterogeneity.

4.1. Strengths and limitations

To the best of our knowledge, this study was the first to assess the cardioprotective events of GLP-1 agents in patients undergoing both PCI and CABG. We established a comprehensive search using many electronic databases and adhered to the PRISMA checklist during reporting this manuscript. Even though we acknowledge the existence of some limitations. The small number of included trials (nine articles) with small sample size (1216 patients) might constrain the applicability and generalizability of the results of this meta-analysis. The included RCTs had short treatment lengths; thus, we could not completely assess the long-term cardioprotective outcomes and major cardiac adverse events. This study involved RCTs only, to provide a high level of evidence and, thus, a high-grade recommendation.

5. Conclusion

This meta-analysis showed that GLP-1 achieved better LVEF, improved salvage index, reduced infarct size, and less MACE risk. Larger studies on sufficiently powered coronary artery disease populations are needed to assess the potential beneficial effects of GLP-1 on morbidity and mortality.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary material