Glucagon-Like Peptide-1 Receptor Agonists and Diabetic Cardiovascular Disease: Implications of the LEADER Study

Diabetes care has historically focused on glucose control, but at least since the publication of the UGDP in 1970 there has been interest in the effects of glucose-lowering treatments of diabetes on cardiovascular disease (CVD). More recently, since 2008, the FDA has required that new diabetes agents be evaluated for cardiovascular safety, as well as possible benefit. ¹ Although noninferiority to standard therapy in terms of cardiovascular risk has been acceptable to the FDA for new drugs, we are starting to see new hypoglycemic agents that actually appear to confer benefit on cardiovascular end points. To this end, the journal has introduced a perspectives/commentary section appraising these clinical trials for the busy clinician. ² We previously reported on the significant benefits of the SGLT-2 inhibitor, empagliflozin, compared with placebo on the primary end point: a composite of death from cardiovascular causes, nonfatal myocardial infarction (MI), and nonfatal stroke. Furthermore, there were significant reductions in hospitalizations for heart failure and total mortality. These findings may be unique to empagliflozin in the class of SGLT2 inhibitors. It is unlikely that the benefits could be ascribed to an amelioration of atherosclerosis and neurohormonal and hemodynamic mechanisms have been suggested.

In this perspective, we focus on the Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results (LEADER) study. ³

In this trial, 9340 diabetic patients were randomized to either liraglutide (1.8 mg or the maximum tolerated dose) or placebo. The major inclusion criteria were age of 50 years or greater and at least one cardiovascular coexisting condition (CHD, cerebrovascular disease, peripheral arterial disease, chronic kidney disease of stage 3 or greater, or chronic heart failure of New York Heart Association class II or III) or age of 60 years or older with at least one cardiovascular risk factor (microalbuminuria, hypertension, LVH, left ventricular systolic or diastolic dysfunction, or ankle/brachial index <0.9). Major exclusion criteria included, type 1 diabetes mellitus, use of glucagon-like peptide-1 (GLP-1)-receptor agonists, DPP4-inhibitors, pramlintide, rapid acting insulin, familial or personal history of medullary thyroid cancer or MEN type-2 or the occurrence of an acute coronary or cerebrovascular event in the 14 days before randomization. It is important to note that the majority of patients (81%) had established CVD and 24.7% had chronic kidney disease of stage 3 or greater, making this a very high-risk group for future cardiovascular events. Compliance was very good and 97% completed the study with the median time to exposure to drug or placebo of 3.5 years with a median follow-up of 3.8 years. At baseline, the mean duration of diabetes was 12.8 years and mean HbA1c was 8.7%. At 36 months, there was a significant 0.4% reduction in HbA1c, 2.3 kg reduction in weight, and 1.2 mmHg reduction in systolic BP. Interestingly, the data on the lipid profile were not commented on in this study. The primary composite outcome was death from cardiovascular causes, nonfatal MI, and nonfatal stroke.

There was a significant reduction in the composite primary outcome in the liraglutide group: hazard ratio, 0.87; confidence intervals (0.78–0.97), P < 0.001 for noninferiority and P = 0.01 for superiority. Also there was a significant 12% reduction in the expanded composite outcome comprising the primary end point plus coronary revascularizations and hospitalizations for angina and heart failure. There was a significant 22% reduction in cardiovascular death and 15% reduction in total mortality. The reductions in nonfatal MI and nonfatal stroke were not significant and there was no significant benefit on heart failure admissions. Subgroup analyses suggested a greater benefit for patients with glomerular filtration rate (GFR) <60 and patients with CVD. With respect to microvascular disease, although there was a significant reduction of 22% in the defined nephropathy outcomes, there was no benefit on retinopathy. There were no significant increases in pancreatitis or cancers in the liraglutide group with the caveat that there was a trend to increase in pancreatic carcinoma (P = 0.06). However, there was a significant increase in acute gall stone disease and other moderate gastrointestinal problems. Very interestingly, there was a significant decrease in severe hypoglycemic episodes in the liraglutide group.

So what might be the mechanisms for these cardiovascular benefits? We would like to suggest that the timeline and reduction in events are supportive of antiatherosclerotic mechanisms. ^4,5 The collective effects of reduction in HbA1c, body weight, and systolic blood pressure are important mediators of risk reduction. Also liraglutide has been shown to decrease inflammation, postprandial hyperlipidemia, and coagulation; all of these can have a beneficial effect on atherothrombosis. In addition, it can afford cardiac protection by improving glucose utilization and decreasing fatty acid utilization. Finally, the reduction in episodes of severe hypoglycemia in the liraglutide group could have also contributed, given that hypoglycemia induces a proinflammatory–procoagulant diathesis with the attendant deleterious hemodynamic and electrophysiological effects of a catecholamine surge. ⁶ Thus the totality of evidence on reduced cardiovascular mortality is extremely positive for the diabetic patient and the timeline is very suggestive of a decrease in atherosclerotic CVD. However, we must await future studies examining these mechanisms before any firm conclusion on the mechanism can be drawn.

An earlier study, examining another member of the GLP-1 agonist class, lixisenatide, under the auspices of the ELIXA (Evaluation of Lixisenatide in Acute Coronary Syndrome) study, demonstrated noninferiority. ⁷ Although noninferiority in terms of cardiac risk is acceptable in a drug whose primary goal is to improve glucose control, superiority is probably better, if it is true. How did two similar drugs, evaluating a similar issue, achieve such different outcomes?

Both studies were randomized controlled trials, comparing a GLP-1 agonist against placebo. In ELIXA, glycemic control was otherwise managed by individual investigators in accordance with local guidelines by adjustment of concomitant medications or the addition of new medications, with the exception of incretin-directed therapies. Patients were randomly assigned in a double-blind manner, to treatment or placebo. Glycemic control, as assessed by glycated hemoglobin, and prevalence of diabetes complications were similar in the treatment and placebo groups. A total of 6068 patients were randomized. The primary composite end point was cardiovascular death, myocardial infarction, stroke, or hospitalization for unstable angina. ⁶ In LEADER, randomization was stratified by calculated estimated GFR at screening, following which patients were randomly assigned to treatment or placebo. Glycemic control was maintained at target by addition of antihyperglycemic agents other than GLP-1 agents, DPP-4 inhibitors, or pramlintide. A total of 9340 patients were randomized. Similarly, the primary composite end point was cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke. Hospitalization for unstable angina was tabulated, but was not part of the composite end point. ³

However, in ELIXA, patients were selected if they had had a myocardial infarction or hospitalization for unstable angina within 180 days before study entry. ⁷ In LEADER, patients were selected if they were 50 years of age or more with at least one coexisting cardiovascular condition (coronary artery disease, cerebrovascular disease, chronic kidney disease state 3 or greater, or chronic congestive heart failure of New York Heart Association class II or III) or 60 years of age or more with at least one cardiovascular risk factor (microalbuminuria, proteinuria, hypertension and left ventricular hypertrophy, left ventricular dysfunction, or an ankle-brachial index <0.9); an acute coronary or cerebrovascular event within 14 days before screening were exclusionary criteria. ³ In simpler terms, in ELIXA, patients appeared to be at higher acute risk, whereas in LEADER, a broader population, but at less acute risk, was recruited.

So these major studies differed in the end point, selection, and randomization processes. The agents also differ. Lixisenatide has a shorter Tmax after administration than liraglutide (1–3.5 hr compared with 8–12), a larger volume of distribution (100 L compared with 20–25 L), a faster T-1/2 for elimination (3 compared with 13 hr), a faster clearance rate (35 L/hr compared with 0.9–1.4), and less protein binding (55% compared with >98%). ⁸ These various pharmacokinetics may predispose a longer active period for liraglutide. Meier has speculated that the short-acting GLP-1 agonists (e.g., lixisenatide) primarily impact postprandial glucose through inhibition of gastric emptying, whereas longer acting agents (e.g., liraglutide) may preferentially impact fasting glucose through their “insulinotropic” and “glucagonostatic” actions ⁹ ; how these glucose-controlling mechanisms impact cardiovascular function is unclear. There is no evidence that the longer exposure to liraglutide is associated with adverse events such as hypoglycemia or pancreatitis.

Previous comparisons of GLP-1 agonists showed no significant differences among available agents in terms of glycemic effect or potential cardiac outcomes; the lack of head-to-head comparisons made rational choice difficult, in what has been described as a “triumph of innovation” ¹⁰ in terms of new options for diabetes management. ¹¹ Similarly, Monami et al., in a meta-analysis of published data, opined that GLP-1 agonists were likely safe from the cardiovascular perspective and could well have benefit, at least in comparison with placebo. ¹² A subclass of GLP-1 agonists, those that could be given weekly, were also felt to be same; direct comparisons could not be made between agents. ¹³

DPP-4 inhibitors also represent a novel pharmacologic approach to the treatment of type 2 diabetes, and act along a pathway similar to the GLP-1 agonists by preventing the breakdown of endogenous GLP-1, stimulating insulin release, slowing gastric emptying, reducing glucagon secretion, etc. ¹⁴ The DPP-4 agents, unlike the GLP-1 agonists, have the advantage of being orally administered. Several recent DPP-4 agent trials have shown noninferiority in terms of cardiovascular safety when compared with placebo; again there are no head-to-head comparisons. ¹⁵ However, there is a concern for heart failure risk with saxagliptin.

So what do we have at the end of the day? At minimum, we have potent new hypoglycemic agents that are safe and generally well tolerated, effective at lowering blood sugar, and not dangerous in terms of cardiovascular safety. We have the possibility that, in some high-risk patients, cardiac safety can be enhanced compared with placebo. We do not know whether this apparent safety reflects intrinsic qualities of the drugs, such as pharmacokinetic performance or other features, or studies designed to optimize certain dramatic end points. We do not have the luxury of head-to-head comparisons, so we may never know how different drugs might perform in a different study design. Until some future paradise where we will have these data, clinicians will have to make their best decisions on behalf of their diabetic patients, decisions likely to be impacted by bureaucratic and financial pressures.

This study was supported, in part, by the Research Service, Department of Veterans Affairs. This material is the result of work that was supported, in part, by resources from the VA Northern California Health Care System, Martinez and Sacramento, California. The contents presented here do not represent the views of the Department of Veterans Affairs or the United States Government. The authors acknowledge the dedicated assistance of Ms. Alba Scott, MILS, Medical Librarian.