The Cardiovascular Effects of GLP-1 Receptor Agonists

Glucagon-like peptide-1 receptor (GLP-1R) agonists have been shown to regulate blood glucose concentrations by mechanisms including enhanced insulin synthesis/secretion, suppressed glucagon secretion, slowed gastric emptying, and enhanced satiety. GLP-1 receptors have also been identified in the heart, kidneys, and blood vessels, leading to the hypothesis that GLP-1R agonists may affect cardiovascular function or cardiovascular disease (CVD). The aim of this literature review was to assemble and assess preclinical and clinical data of potential medical importance regarding the cardiovascular effects of GLP-1R agonists. Preclinical studies with the GLP-1R agonists GLP-1, exenatide, or liraglutide provided evidence that GLP-1R stimulation favorably affects endothelial function, sodium excretion, recovery from ischemic injury, and myocardial function in animals. Similar observations have been made in exploratory studies on GLP-1 infusion in normal subjects and patients with type 2 diabetes. Post hoc analyses of phase III studies of patients with type 2 diabetes treated with exenatide(bid or qw) or liraglutide(qd) showed that these GLP-1R agonists reduced blood pressure, an effect largely independent of weight loss, and that liraglutide slightly increased heart rate. Preliminary data also indicated that GLP-1R agonists reduced markers of CVD risk such as C-reactive protein and plasminogen activator inhibitor-1. Ongoing studies are examining the effects of administering GLP-1R agonists to patients at risk of CVD, postangioplasty patients, post-CABG patients, and patients with heart failure. Additional studies should provide meaningful data to determine whether GLP-1R agonists provide unique treatment benefits to patients at risk for or with established CVD.


Introduction
Glucagon-like peptide-1 receptor (GLP-1R) agonists have been investigated for treating type 2 diabetes mellitus (T2DM) since the early 1990s because of their ability to enhance glucose-dependent insulin secretion. The cessation of GLP-1-stimulated insulin release when blood glucose concentrations are <55 mg/dL is probably responsible for the low incidence of severe hypoglycemia observed in phase III clinical trials of the GLP-1R agonists exenatide and liraglutide [1]. Uniquely, GLP-1R agonists exert coordinated effects on mechanisms of glucose release and nutrient uptake, suppress inappropriately elevated glucagon secretion, slow gastric emptying, and increase satiety with the net result of reduced body weight [2]. Weight loss and lower blood glucose concentrations may reduce patients' risk for cardiovascular disease (CVD) [3][4][5].
In addition, GLP-1R agonists are hypothesized to have pleiotropic effects on the cardiovascular system. This is a particularly important research area for patients with T2DM since these individuals are at increased risk for CVD and recover from cardiovascular events less well than patients without diabetes [6][7][8].
Type 2 diabetes is frequently associated with proatherogenic risk factors including hypertension, overweight or obesity, and dyslipidemia with elevated proinflammatory markers and procoagulant factors [9]. However, patients with T2DM also demonstrate endothelial dysfunction and impaired vasodilatation [9], microvascular disease (particularly in the myocardial microcirculation [10][11]), increased arterial stiffness [12], left ventricular hypertrophy [13], and cardiac fibrosis [14]. To improve the cardiovascular outcomes of patients with T2DM, comprehensive risk reduction was recommended by a joint committee of the American Heart Association and the American Diabetes Association, specifically to achieve an A1C <7% or as close to normal (<6.0%) as possible without significant hypoglycemia, weight loss to reduce multiple CVD risk factors, aggressive management of hypertension, treatment of dyslipidemia, and therapy to prevent platelet aggregation [15].
The purpose of this review is to summarize published preclinical and clinical studies on the cardiovascular effects of GLP-1R agonists and to assess the evidence as a whole. Search terms included "glucagon-like peptide," "GLP," "exenatide," and "liraglutide" and The Cardiovascular Effects of GLP-1 Receptor Agonists papers published prior to October 1, 2009 were selected based on their relevance to CVD, cardiovascular risk, or hypertension.

GLP-1 Receptor Agonists
Glucagon-like peptide-1 (GLP-1) is a peptide hormone primarily synthesized in the distal ileum and colon and released into the circulation in response to food intake. GLP-1(7-36)NH 2 is the primary active isoform, although it is degraded with a half-life of 2 min to GLP-1(9-36) by dipeptidyl peptidase-4 (DPP-4) [16]. In patients with T2DM, the postprandial GLP-1 secretory response to a meal is significantly diminished and it is hypothesized that this contributes to the pathogenesis of diabetes [17][18].
Discovery of the pleiotrophic effects of GLP-1 on blood glucose control led to the development of GLP-1R agonists with extended half-lives for the treatment of diabetes. The first GLP-1R agonist (incretin mimetic) approved for clinical use was exenatide (exendin-4), which is administered by subcutaneous injection. Discovered in 1990, this naturally occurring 39-amino-acid peptide was isolated from the saliva of the Gila monster (Heloderma suspectum). It shares 53% homology with mammalian GLP-1 and stimulates GLP-1R, yet is not degraded by DPP-4. Exenatide's circulating half-life is 60-90 min, which is sufficient for twice daily (bid) administration [2]. A once-weekly formulation (exenatide QW) has been submitted for regulatory review to the US FDA.
Liraglutide, a derivative of GLP-1 administered once daily (qd), has also been approved for clinical use in the United States and Europe. Liraglutide differs from native GLP-1 by the substitution of arginine 34 for lysine. However, the lysine at position 26 is acylated with a fatty acid [-CO-(CH 2 ) 14 CH 3 ] that facilitates binding to serum albumin, creating a liraglutide: albumin complex extending the half-life of liraglutide sufficiently for qd dosing. The addition of a fatty acid and formation of a complex with albumin are key molecular differences between liraglutide and GLP-1 [19][20]. The metabolites of liraglutide differ from those of exenatide due to liraglutide's susceptibility to cleavage by DPP-4.
The majority of studies on the cardiovascular effects of GLP-1R agonists have been completed with native GLP-1, exenatide, or liraglutide. Additional GLP-1R agonists in clinical trials include lixisenatide (AVE0100), taspoglutide, and albiglutide. Because GLP-1R agonists in development for clinical applications are resistant to cleavage by DPP-4, studies on the physiological effects of GLP-1 metabolites will not be addressed.

The GLP-1 Receptor
The human GLP-1R is a G-protein coupled receptor consisting of 463 amino acids including a large extracellular domain and seven transmembrane domains [21]. GLP-1R expression has been identified in pancreatic islet cells, lung, brain, stomach, kidney, liver, and heart in animal studies [21][22]. Recently, GLP-1R expression has been further localized in mice to cardiomyocytes, the endocardium, microvascular endothelium, and medial aortic smooth muscle cells [22] (Figure 1).
Multiple intracellular mediators for GLP-1R signaling have been identified in different tissues. It is generally agreed that the GLP-1R signals via the stimulatory G-protein Gs to increase cyclic AMP accumulation and activate protein kinase A. GLP-1R stimulation also increases intracellular calcium concentrations and activates calcium-dependent signaling cascades in pancreatic beta cells [21]. Activation of the phosphatidylinositol-3-kinase and mitogen-activated protein kinase pathways has been observed in insulin-secreting cell lines [23]. When GLP-1 was infused into isolated mouse hearts, cGMP production markedly increased compared with nontreated hearts, suggesting a role for this signaling pathway in GLP-1 activity [22]. If the signaling mechanisms observed in pancreatic cells are stimulated by GLP-1 in vascular and cardiac cells, vascular tone, muscle contractility, cell growth/differentiation, extracellular matrix remodeling, and apoptosis may be affected.
In humans, GLP-1R activity depends on GLP-1R agonist concentration [24][25]. This relationship may explain the observation that the DPP-4 inhibitor sitagliptin has lesser effects on postprandial glucose than exenatide, as sitagliptin yields a lower dose of GLP-1R agonist [26].

Preclinical Studies on the Cardiovascular Effects of GLP-1 Receptor Agonists
Numerous preclinical studies in rodents have investigated cardiovascular mechanisms of GLP-1R stimulation. Observed responses have included sympathetic nervous system activation, reduced triglyceride absorption, increased natriuresis, vasorelaxation, enhanced NO production, increased cardiac glucose uptake, and augmented cardiac contractility ( Figure 2; Table 1).
A recent study in mice has also suggested that exenatide reduces macrophage adhesion to the endothelium, the key initiating step in formation of atherosclerotic plaques [27]. The majority of these findings have been observed in >1 model system by multiple research groups, suggesting that the results of GLP-1R stimulation are reproducible. Observation of responses to GLP-1R agonists in isolated blood vessels and hearts suggests that GLP-1 agonists act directly on these target organs and not by modulating insulin release or systemic glucose concentrations, which are controlled in these experiments. The effects of GLP-1R agonists have also been studied in models of MI (ischemia with or without reperfusion) and heart failure (HF). While it is difficult to compare studies conducted with methodological differences, improvement in postischemia cardiac function or infarct size with GLP-1 or exenatide treatment was observed in seven of the nine studies identified (Table 2). Interestingly, improved cardiac function or reduced infarct size was observed with mild (e.g., low-flow or short-term) ischemia as well with permanent vessel occlusion and some effects were observed in isolated hearts as well as in intact animals ( Table 2). The most striking results were observed in in vivo studies with longer follow-up times, particularly the Timmers study in pigs (3 days) [40] and the Noyan-Ashraf study in mice (28 days) [41]. In the Timmers study, compared to saline treatment, exenatide significantly reduced infarct area (32.7 ± 6.4 vs. 53.6 ± 3.9%, P = 0.031, respectively) and myocardial stiffness (0.12 ± 0.06 mm Hg/mL vs. 0.22±0.07 mm Hg/mL, P = 0.004). In addition, systolic wall thickness (47.3 ± 6.3% vs. 8.1 ± 1.9%, P < 0.001) and dPdt m (1611 ± 91 mm Hg/s vs. 1345 ± 75 mm Hg/s, P < 0.05) increased with exenatide treatment [40]. In the Noyan-Ashraf study, treatment with liraglutide versus saline produced marked differences in outcome, respectively: cardiac rupture occurred in fewer mice (20% vs. 77%, P = 0.0001); infarct size was reduced (21 ± 2% vs. 29 ± 3%, P = 0.02); and cardiac output improved (12.4 ± 0.6 mL/min vs. 9.7 ± 0.6 mL/min, P = 0.02). In contrast, the negative studies examined infarct size in pigs after only 2 or 2.5 h of reperfusion, respectively [42][43].
GLP-1R stimulation has also been observed to improve cardiac function in animal models of HF. In a pacing-induced dilated cardiomyopathy canine model induced over 28 days and then treated for 2 days, an infusion of GLP-1 (7-36) increased myocardial glucose uptake versus saline (7.9 ± 0.5 μmol/min vs. 4.7 ± 0.3 μmol/min, P < 0.05) and improved LV ejection fraction (EF) over  Table 1 Physiological effects of GLP-1R agonists in animal studies
Although questions remain as to the role of the endothelium in GLP-1 stimulated vasodilatation and which of the preclinical models are most predictive of human response, these preclinical studies support a transition to human studies on the cardiovascular effects of GLP-1R agonists.

Clinical Data on the Cardiovascular Effects of GLP-1 Receptor Agonists
Current clinical data on the cardiovascular effects of GLP-1R stimulation in humans have been provided by small, exploratory studies on intravenous administration of GLP-1 or through post hoc analyses of data obtained from phase III clinical trials of subcutaneously administered GLP-1R agonists. Short-term studies of direct effects of GLP-1R agonists on the sympathetic nervous system, renal excretion of sodium, and vasodilatation have been published (Table 3). Less effect of GLP-1R agonists on cardiac sympathetic nerve stimulation seems to have been observed in humans than in animals, but this may differ between GLP-1R agonists and between species. In the LEAD-6 study, a larger increase in heart rate was observed with liraglutide qd than with exenatide bid (3 vs. 1 bpm, respectively, P = 0.0012) [49]. Furthermore, the renal and vascular effects of GLP-1 in humans are qualita-tively similar to those in animal studies. In contrast to the significant amount of preclinical data available, only one study has been published to date in patients with acute MI after successful reperfusion [57]. In this study, Dr. Nikolaidis and colleagues compared the in-hospital outcomes of a 72-h infusion of recombinant GLP-1 (1.5 pmol/kg/min) added to standard therapy with standard therapy alone in 21 high-risk patients with acute MI and LV dysfunction. In 10 patients, GLP-1 administration was initiated an average of 212 min after reperfusion and was associated with significantly improved LVEF (29-39% compared with no change in the control group, P < 0.01) and contractile function (−21% in regional wall motion score index vs. no change, P < 0.001) measured 6-12 h after infusion. In the small subset of patients (four per group) for whom 120-day data were available, the effect on LVEF persisted. Trends toward shorter hospital stays (6 vs. 10 days) and reduced in-hospital mortality (one death in the GLP-1 treated group vs. three in the control group) were also observed [57].
Two research groups have studied the effects of GLP-1 infusion on patients with or without diabetes undergoing coronary artery bypass graft (CABG) procedures [58][59]. In one study, GLP-1 (1.5 pmol/kg/min) was administered beginning 12 h before surgery and for 48 h afterward; in the second, GLP-1 (3.6 pmol/kg/min) was administered for 12 h after transfer to the intensive care unit. Neither study observed differences in postoperative hemodynamics, but both reported reduced use of inotropic and vasoactive infusions and improved glucose control postoperatively [58][59].
The effects of GLP-1R stimulation on HF have also been investigated. In a 48-h randomized, double-blinded crossover study of 20 nondiabetic patients with compensated HF, no differences in myocardial function or metabolism were observed with administration of GLP-1 (0.7 pmol/kg/min) [60]. The results of a 3-day IV infusion of GLP-1 for ∼1 h in seven healthy volunteers Increased muscle sympathetic nerve activity but no effect on HR, BP, or cardiac neural activity [51] Ph III study comparing exenatide bid and liraglutide qd in 464 patients over 26 weeks HR increased from baseline +0.69 bpm for exenatide versus +3.28 bpm for liraglutide (P = 0.0012) [49] Kidneys IV infusion of GLP-1 for 3 h in 15 normal and 16 obese subjects given salt bolus ↑Na + in both groups. ↓H + excretion and ↓glomerular hyperfiltration in obese subjects [52] IV infusion of GLP-1 for 3 h in 17 healthy salt-loaded men ↑Na + , ↓H + excretion [53] Blood vessels IV infusion of GLP-1 in 29 healthy subjects (2 h) Forearm blood flow increased, ↑response to ACh challenge [54] IV infusion of GLP-1 in 12 patients with T2DM and CAD (105 min) Flow-mediated vasodilation of brachial artery doubled [55] SC injection of exenatide prior to high-fat meal in 28 subjects with IGT or T2DM Improved endothelial function after high-fat meal [56] ACh, acetylcholine; DBP, diastolic blood pressure; CAD, coronary artery disease; HR, heart rate; IGT, impaired glucose tolerance; Na + , sodium ion; SBP, systolic blood pressure; T2DM, type 2 diabetes mellitus.
open-label study of GLP-1 infusion (4 pmol/kg/min) in six patients with HF and T2DM indicated that the therapy was well tolerated and there was a trend toward improved myocardial function [61]. Finally, a 5-week open-label trial of GLP-1 (2.5 pmol/kg/min infusion) in 12 patients with HF (with or without diabetes) compared with nine patients on standard therapy showed that GLP-1 therapy was associated with significant improvements in LVEF (from 21±3 to 27±3%, P < 0.01), maximal oxygen uptake (from 10.8 ± 0.9 mL O 2 /min/kg to 13.9 mL O 2 /min/kg, P < 0.001), and 6-min walking distance (from 232 ± 15 m to 286 ± 12 m, P < 0.001) [62]. In the latter trial, HF patients with or without diabetes appeared to benefit.
No phase III studies of exenatide or liraglutide have been conducted in patients with T2DM and CVD; however, blood pressure (BP) was recorded during all clinical studies. Reductions in BP were observed during studies of both exenatide and liraglutide [63][64]. Recently, 52-week follow-up data on administration of once-weekly exenatide demonstrated an systolic blood pressure (SBP) reduction of −6.2 mm Hg and a DBP reduction of −2.8 mmHg (n = 120; P < 0.05) [65]. A post hoc analysis of pooled data from six trials of exenatide bid in patients with T2DM compared with placebo and insulin (n = 2171) demonstrated a significant reduction in SBP relative to placebo and insulin comparators with a weak correlation to weight loss (r = 0.09, P = 0.002). No significant decrease in mean DBP was observed between treatment groups in this analysis. Patients with higher baseline SBP (≥150 mmHg) experienced the greatest blood pressure-lowering effects (Figure 3) [49]. A meta-analysis across six trials of liraglutide (n = 1363 for 1.8 mg, n = 896 for 1.2 mg) confirmed BP reductions of ∼2.5 mm Hg with treatment and demonstrated reductions within 2 weeks of therapy, before significant weight change occurred [64]. In addition, central pulse pressure and augmentation pressure have been shown to be reduced significantly (∼8 mm Hg) by exenatide compared with insulin glargine in patients with T2DM [66].
Reductions in risk factors or markers for CVD have also been observed following treatment with GLP-1R agonists [65,[67][68]. Decreases in triglycerides of 20% or more have been reported with exenatide qw or liraglutide qd treatment, but changes in other lipids have been less reproducible and may be related to weight loss [65,68]. A significant reduction in LDL-C and an increase in HDL-C (n = 151) was observed in a long-term followup study of exenatide bid in which patients lost −5.3 kg, but not in a 26-week study of exenatide bid versus liraglutide qd (n = ∼230 per group) in which patients lost ∼3 kg [49,67]. Other surrogate markers of CVD risk reported in preliminary studies to improve during therapy with a GLP-1R agonist include C-reactive protein, adiponectin, alanine transaminase, plasminogen activator inhibitor-1, and brain natriuretic peptide [65,[67][68].
In patients with T2DM, the effects of GLP-1R agonists on vasodilatation, BP, and triglycerides have been reproducible with different GLP-1R agonists studied by different researchers. As a whole, the results support additional investigation on the potential cardiovascular benefits of therapy with GLP-1R agonists. Qualitatively, the nature of the cardiovascular effects of the GLP-1R agonist class appear to differ from the cardiovascular effects of other glucoselowering therapies, including biguanides, sulfonylureas, thiazolinediones, and insulin.
Important questions for future research include the following: Can improvements in SBP be confirmed to occur largely independently of weight loss, and what are the effects on central aortic pressure and nocturnal dipping? Is the observed tachycardia related to vasodilatation or a direct effect? Does the change in sodium excretion reduce cardiac load? Do different GLP-1R agonists have similar effects on the cardiovascular system, or are their effects meaningfully different? Can GLP-1R agonists be designed that do not increase heart rate? How do the cardiovascular effects

The Safety and Tolerability of GLP-1R agonists
The cardiovascular safety characteristics of exenatide BID and liraglutide in clinical trial programs have been assessed, and no increase in the incidence of adverse cardiovascular events was apparent for either therapy [69] [70]. In the pooled clinical trial data for exenatide BID (N = 2279 subjects), the relative risk of at least one CV event was 0.69 (95%CI 0.46-1.04) for patients treated with exenatide BID compared with the pooled comparators. However, neither the size nor duration of the clinical trials was designed to detect differences in cardiovascular outcomes. Similarly, no increase in mortality was observed in a subset of patients in the ACCORD study treated with a GLP-1 receptor agonist [71]. The only GLP-1R mechanism of action identified to date that might have a negative effect on cardiovascular safety is the slight increase in heart rate. In epidemiological studies, 20 bpm increases are associated with higher mortality over 20 years [72]. General safety issues of interest for the GLP-1R agonist class include gastrointestinal issue such as nausea and vomiting, which are usually mild to moderate and decline after the first few months of therapy. Hypoglycemia requiring assistance is rare with GLP-1R agonists except when used in combination with a sulfonylurea [1]. Potential safety concerns include an increased risk of pancreatitis, although no causal relationship has been shown; patients with T2D appear to be at higher risk for this condition [73]. Data from rat and mouse studies demonstrated an increased risk of Ccell carcinoma for liraglutide, although this has not been proven in humans [70]. Liraglutide is contraindicated in patients with a personal or family history of medullary thyroid carcinoma or multiple endocrine neoplasia syndrome type 2, and is not recommended as first-line therapy for patients inadequately controlled on diet and exercise [74]. Hypersensitivity reactions have been documented in postmarketing reports for exenatide BID, and use of exenatide BID is contraindicated in patients with severe renal impairment or end-stage renal disease or in patients with severe gastrointestinal disease (e.g., gastroparesis) [75].

Ongoing Studies
The cardiovascular effects of GLP-1R agonists in patients with T2DM and or CVD are an active area of investigation (www.clinicaltrials.gov). As of October 2009, 11 studies were planned or recruiting patients, including studies on the clinical effects of exenatide or GLP-1 on outcomes of patients in the cardiac intensive care unit, post-PCI patients, post-CABG patients, and patients with HF (Table 4). These studies are essential in determining whether the unique observed effects of GLP-1R agonists on cardio-vascular function depend on the model systems studied, or are a relevant clinical observation.
The single most important issue to be addressed in this research field is the question of patients' cardiovascular outcomes on GLP-1R agonist therapy. To provide the needed evidence, large outcome trials are planned to investigate whether liraglutide and or exenatide QW reduce adverse cardiovascular outcomes in patients with T2DM. The Exenatide Study of Cardiovascular Event Lowering (EXSCEL) trial is recruiting patients for a >5-year study of the time to first confirmed cardiovascular event in 9500 patients treated with exenatide QW or placebo (NCT01144338).

Conclusion
Review of the existing literature supports the hypothesis that GLP-1R agonists exert physiological effects beyond glucose control that affect the cardiovascular system. While observed effects of GLP-1R agonists on triglycerides and markers of cardiovascular risk may be related to weight loss, exploratory data in both animals and humans are consistent with direct effects of GLP-1R agonists on endothelial function, excretion of sodium, improvement in SBP, limitation of ischemia/reperfusion injury, and with improved myocardial function in HF. Planned and ongoing clinical research studies will contribute significantly to knowledge of the cardiovascular effects of GLP-1R agonists. These studies may affect future patient care, as no current therapy for CVD provides the unique combination of potential actions proposed for GLP-1R agonists: improved glucose control with minimal hypoglycemia; improved endothelial function; reduced body weight, BP and serum triglycerides; improved recovery from ischemia; and improved hemodynamics in patients with reduced contractile function.