Lixisenatide and Cardiovascular Outcomes: The Burden of Proof

After cardiovascular (CV) safety concerns emerged with rosiglitazone use, the Food and Drug Administration (FDA) now requires outcome studies to be performed for all new diabetes medications.^1-3  Naturally, we’d prefer to use medications to treat diabetes that actually reduce CV risk – but, at a minimum, they shouldn’t be harmful.  The first CV safety study (ELIXA) for a drug in the glucagon-like peptide 1-receptor agonist (GLP-1 RA) class – lixisenatide – was recently published.⁴  Lixisenatide is a once-daily injectable GLP-1 RA approved for use in Europe and awaits FDA approval in the United States.

The ELIXA study was a multicenter, randomized, double-blind, placebo-controlled trial that evaluated the impact of lixisenatide on CV morbidity and mortality. Patients included in the trial had type 2 diabetes (T2DM), an acute coronary event within 180 days before screening, an estimated glomerular filtration rate greater than 30 mL/min/1.73 m², and a glycated hemoglobin (A1C) level between 5 – 11%. Patients excluded were under 30 years of age, had a percutaneous coronary intervention (PCI) within the previous 15 days, a coronary-artery bypass graft surgery for the qualifying event, or a planned coronary revascularization procedure within 90 days after screening.  Using a time-to-event analysis, the primary endpoint was the 4-point major adverse cardiovascular events (MACE) score, which consists of the first occurrence of either death from CV causes, nonfatal myocardial infarction, nonfatal stroke, or hospitalization for unstable angina. Secondary endpoints included a composite of the primary endpoints or hospitalization for heart failure as well as a composite of the primary endpoints, hospitalization for heart failure, or coronary revascularization procedures.  Change in A1C, urinary albumin-to-creatinine ratio, body weight, heart rate, and blood pressure (BP) were also evaluated, as well as incidence of hypoglycemia, pancreatitis, cancer, and systemic allergic reactions.

The study enrolled slightly more than 6000 patients who were randomly assigned to lixisenatide 10 mcg daily (n=3,034) or a placebo injection (n=3,034). After 2 weeks, the dose of lixisenatide (or placebo) could be increased to 20 mcg daily at the investigator’s discretion. Baseline characteristics were similar between groups, including prescribed CV medications (ACEI/ARB=85%, statin=93%, antiplatelet therapy= 97%, beta-blocker=84%) and T2DM medications (metformin=66%, insulin=39%, sulfonylureas=33%, TZD=1.5%, other=5%). The majority of patients in the study were middle aged (mean = 60 years), male (70%), and white (75%); with a history of T2DM of approximately 9 years and an A1C of 7.7%.  Most patients (76%) had hypertension with an average systolic BP of 130 mmHg at baseline.  PCI was anticipated a prior in a significant portion (67%) of patients at the time of study enrollment.

After a mean follow-up of 25 months, there was no significant difference between lixisenatide (n=406, 13.4%) and placebo (n=399,13.2%) in the primary endpoint. Based on the hazard ratio for the primary endpoint of 1.02 (95% CI, 0.89-1.17), the upper boundary of the confidence interval met the non-inferiority marker of less than 1.3 (p<0.001), but was not superior (p=0.81). There were no statistically significant differences found between lixisenatide and placebo for the composite of the primary endpoints or hospitalization for heart failure (15.0% vs. 15.5% respectively; HR 0.97 [0.85-1.10]; p=0.63), or the composite of the primary endpoints, hospitalization for heart failure, or coronary revascularization procedures (21.8% vs. 21.7% respectively; HR 1.00 [0.90-1.11]; p=0.96). The results were consistent across multiple pre-specified subgroups.

Not surprisingly, lixisenatide was more effective than placebo in reducing A1C at 12 weeks (-0.6% vs. -0.2% respectively; p<0.001) and rates of hypoglycemia were similar (16.6% vs. 15.2% respectively; p=0.14). Modest weight loss with lixisenatide was seen at 12 weeks (-0.6 kg vs. -0.0 kg; p<0.001), which was sustained throughout the study period. Adverse effects (AEs) leading to therapy discontinuation were more frequent with lixisenatide (11.4% vs. 7.2%; p<0.001), which were primarily related to nausea and vomiting. The study found that rates of serious AEs were similar, but slightly higher for placebo (22.1%) than lixisenatide (20.6%). There were also a more cases of pancreatitis reported in the placebo group (n=8) than with lixisenatide (n=5).

Based on the results of ELIXA, lixisenatide is safe.  It neither helped nor harmed patients.  It had no impact on the CV morbidity and mortality of patients with T2DM and a recent acute coronary event.  Moreover, there were numerically higher rates of serious AEs as well as pancreatitis in the placebo group when compared to lixisentatide.  Patients assigned to lixisentide were more likely to stop therapy due to nausea and vomiting, well known AEs for the GLP-1 RAs. Weight loss, an attractive feature of the GLP-1 RA class, was modest. The internal validity of the study is strong based on the study design. The study enrolled patients from 49 different countries but the external validity may be limited as the majority of the patients were middle aged white males.

Five CV outcome studies have been published to date (Table 1). The ELIXA findings are consistent with three published studies which all found no significant differences in MACE in patients treated with a DPP-4 inhibitor versus placebo.^5-7  The patient populations, while not identical, were all patients at very high risk for CV events or with established CV disease. It is worth highlighting that there was no increased risk of heart failure hospitalizations with lixisenatide in ELIXA, which was an unexpected finding in SAVOR-TIMI 53 with saxagliptin.⁵

Table 1: Overview of Published Cardiovascular Safety Studies of New Diabetes Medications

^{CV= cardiovascular; Major adverse cardiovascular event (MACE): Cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke; 4-point MACE: MACE + unstable angina hospitalizations}

Demonstrating CV safety may no longer be good enough given the recent positive outcomes seen in EMPA-REG OUTCOME.⁸  This study not only demonstrated CV safety with a SGLT-2 inhibitor, but empagliflozin significantly reduced MACE, CV mortality, all-cause mortality, and heart failure hospitalization when compared to placebo (see The Heart of the Matter – Is EMPA-REG a Game Changer for Diabetes Management?). The different results seen in ELIXA and EMPA-REG OUTCOME raise several questions.

Were the findings from ELIXA ones that were anticipated?   The FDA required CV outcome studies have been powered to establish safety, not efficacy.  So the CV safety results of ELIXA were anticipated and the fact that ELIXA did not demonstrate CV benefit is (perhaps) not surprising.  Also, ELIXA was conducted over a short, two-year time frame. Based on prior evidence, it is unlikely to see beneficial effects of glycemic lowering in less than 4 years.  However, the results of ELIXA are disappointing.  Previous studies suggested that GLP-1 RAs might have a cardioprotective effect.  There was hope that the class would improve CV outcomes.^9,10

Why did empagliflozin lower CV outcomes, while lixisenatide did not?   GLP-1 RAs have similar effects on surrogate markers such as glucose, weight, and BP.  It’s possible that the CV benefit from empagliflozin is a result of a unique mechanism such as osmotic diuresis. The primary benefit of EMPA-REG OUTCOME was a reduction in CV death; not a reduction in nonfatal MI or stroke.  Moreover, a reduction in CV death and heart failure hospitalizations were seen as early as three months, which also supports the idea that the benefits are unrelated to glucose, weight, or BP.

Will these results be replicated in ongoing and future CV outcome studies?   Results of ongoing trials will certainly shed more light on the CV effects of GLP-1 RAs and SGLT-2 inhibitors.  The results of future studies including LEADER (liraglutide; completed in December 2015), EXSCEL (exenatide; April 2018), CANVAS (canagliflozin; June 2017) and DECLARE-TIMI 58 (dapagliflozin; April 2019) will be highly anticated!

What are the take home points from ELIXA?  ELIXA showed that lixisenatide neither increased nor decreased the rates of CV events in patients with T2DM and ACS.  ELIXA also showed no increased risk of heart failure or pancreatitis with lixisenatide, which should subdue earlier concerns regarding these potential risks. Clinicians should feel reassured that the GLP-1 RAs are reasonable options for the treatment of T2DM.   They are safe.  But in light of the results of EMPA-REG OUTCOME is being safe good enough?