Cardiovascular Outcomes of New Medications for Type 2 Diabetes

Abstract

Cardiovascular (CV) disease remains the leading cause of death in people with diabetes, highlighting the importance of using treatment options that do not increase CV risk or possibly decrease CV outcomes. Since 2008, the Food and Drug Administration has required demonstration of CV safety for all new medications developed for the glycemic management of diabetes. Seven trials have been published that have established CV safety for three DPP-4 inhibitors (alogliptin, saxagliptin, and sitagliptin), three GLP-1 receptor agonists (liraglutide, lixisenatide, and semaglutide), and one sodium–glucose cotransporter-2 inhibitor (empagliflozin). Three of those studies also established superiority with liraglutide, empagliflozin, and semaglutide at reducing the composite primary endpoint of major CV events (CV death, nonfatal myocardial infarction, and nonfatal stroke). In addition, one trial found an increase in heart failure hospitalizations with saxagliptin. The findings of these trials must be compared and contrasted cautiously given the differences in patient populations and trial designs, but together they provide important information that can be used to shape our treatment guideline recommendations and patient-specific treatment decisions.

Introduction

Approximately 415 million adults worldwide have diabetes, the vast majority of whom have type 2 diabetes (T2D).¹ In the United States, cardiovascular disease (CVD) remains the leading cause of death in people with T2D.² Despite rates of myocardial infarction (MI) decreasing significantly over the past two decades in people with diabetes, event rates remain significantly higher in people with diabetes than those without diabetes.³

Goals of treatment, therefore, include the reduction of long-term complications, including CVD. Several large clinical trials have evaluated the potential cardiovascular (CV) effects of intensive glycemic control, but none was initially able to demonstrate CV benefit with intensive glycemic control compared with standard glycemic control, and none was designed to evaluate specific diabetes medications.^4

–8 The UK Prospective Diabetes Study extension trial did show risk reductions for MI and death from any cause that emerged over time with intensive glycemic control after 10 years of follow-up.⁹ Conversely, the ACCORD trial found an increase in mortality with treatment aimed at achieving a goal of glycosylated hemoglobin (A1C) of less than 6% in high-risk patients.⁶ Thus, the impact of intensive glycemic control on CVD remains unclear.

In the past, concern was raised about the CV safety of diabetes medications. In 2007, a highly publicized meta-analysis found a 43% increased risk of MI (P = 0.03) and 64% increased risk of CV death (P = 0.06) with rosiglitazone, which subsequently led to the addition of a black box warning to the rosiglitazone prescribing information, development of a risk evaluation, and mitigation strategy (REMS), and prescribing restrictions.^10,11 It is worth noting that a subsequent prospective, randomized controlled trial (RECORD) found that rosiglitazone did not increase the risk of overall CV morbidity and mortality compared with standard antihyperglycemic therapy, and some rosiglitazone prescribing and dispensing restrictions were eventually removed.^12,13

In 2010, the Food and Drug Administration (FDA) issued a “Guidance for Industry” document requiring sponsors for all new diabetes medications to demonstrate CV safety.¹⁴ This guidance recommends evaluation of CV endpoints, including CV mortality, MI, and stroke, and states that trials can demonstrate safety by achieving an upper bound of the two-sided 95% confidence interval (CI) for the estimated risk ratio of less than 1.8 for premarketing trials and 1.3 for postmarketing trials.¹⁴

A handful of trials have evaluated the CV effects of specific diabetes medications in differing patient populations with varying results^15
–17; however, this review focuses specifically on CV outcome trials of new diabetes medications in T2D patients that were conducted based on the 2008 FDA requirement.

Clinical Evidence

To meet the FDA standard, the trials described in this article were all noninferiority trials designed to establish CV safety, not to demonstrate CV benefit. Throughout this review, the composite of CV death, nonfatal MI, or nonfatal stroke is referred to as major adverse cardiovascular events (MACEs), and the four-point MACE is MACE plus hospitalization for unstable angina. Published clinical trials are summarized in Table 1. Future or ongoing trials are summarized in Table 2.

Table 1.

Cardiovascular Outcome Trials of Type 2 Diabetes Medications

						Primary outcome results
Trial (drug)	Size	Median follow-up	Major CV inclusion criteria	Baseline characteristics	Primary outcome	Drug (%)	Placebo (%)	HR (95% CI)
DPP4 inhibitors
EXAMINE²⁰ (alogliptin)	5380	18 Months	Acute coronary event (acute MI or unstable angina requiring hospitalization) within previous 15–90 days	Average age, 65 years; duration of diabetes, 10.3 years; A1C 8.0% ± 1.1%	MACE	11.3	11.8	0.96 (≤1.16)
SAVOR-TIMI 53¹⁹ (saxagliptin)	16,492	2.1 Years	Established CVD or multiple risk factors for vascular disease	Average age, 61 years; duration of diabetes, 7 years; A1C 8.0% ± 1.4%	MACE	7.3	7.2	1.0 (0.89–1.12)
TECOS²¹ (sitagliptin)	14,671	3 Years	History of major coronary artery disease, ischemic stroke, or peripheral artery disease	Average age 65 years; duration of diabetes 11.6 years; A1C 7.2% ± 0.5%	Four-point MACE	11.4	11.6	0.98 (0.88–1.09)
SGLT-2 inhibitors
EMPA-REG OUTCOME²⁷ (empagliflozin)	7020	3.1 Years	Established CVD, BMI <45 kg/m², eGFR >30 mL/min/1.73 m²	Average age, 63 years; duration of diabetes varied (more than 80% of patients diagnosed at least 5 years prior, and more than 55% of patients diagnosed at least 10 years prior; A1C 8.1% ± 0.9%	MACE	10.5	12.1	0.86 (0.74–0.99)
GLP-1 receptor agonists
ELIXA²⁹ (lixisenatide)	6068	25 Months	Acute coronary event within previous 180 days	Average age, 60 years; duration of diabetes, 9 years; A1C 7.7% ± 1.3%	Four-point MACE	13.4	13.2	1.02 (0.89–1.17)
LEADER³⁰ (liraglutide)	9340	3.8 Years	Age ≥50 years with established CVD or chronic renal failure or age ≥60 years with at least one CV risk factor	Average age, 64 years; duration of diabetes, 12.8 years; A1C 8.7% ± 1.6%	MACE	13.0	14.9	0.87 (0.78–0.97)
SUSTAIN-6³¹ (semaglutide)	3297	2.1 Years	Age ≥50 years with established CVD, chronic HF, chronic kidney disease, or age ≥60 with at least one CV risk factor	Average age, 64 years; duration of diabetes, 13.9; A1C 8.7% ± 1.5%	MACE	6.6	8.9	0.74 (0.58–0.95)

A1C, hemoglobin A1C; CI, confidence interval; CV, cardiovascular; CVD, cardiovascular disease; four-point MACE, MACE + unstable angina hospitalizations; HF, heart failure; HR, hazard ratio; MACE, major adverse cardiovascular event (cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke); MI, myocardial infarction

Table 2.

Future or Ongoing Cardiovascular Outcome Trials of Type 2 Diabetes Medications

Trial (drug)	Comparator	Size	Major inclusion criteria	Primary outcome	Anticipated completion date	Clinicaltrials.gov identifier
DPP-4 inhibitors
CAROLINA (linagliptin)	Glimepiride	6115	Established CVD or diabetes mellitus end organ damage or ≥2 specified cardiac risk factors	Four-point MACE	March 2019	NCT01243424
CARMELINA (linagliptin)	Placebo	8300	High risk for CV events	Four-point MACE	January 2018	NCT01897532
SGLT-2 inhibitors
CANVAS (canagliflozin)	Placebo	4331	History of or high risk for CV events	MACE	February 2017	NCT01032629
DECLARE-TIMI 58 (dapagliflozin)	Placebo	17,276	High risk for CV events	MACE	April 2019	NCT01730534
VERTIS-CV (ertugliflozin)	Placebo	8000	Evidence or a history of atherosclerosis involving the cerebral, coronary, or peripheral vascular systems	MACE	October 2019	NCT01986881
GLP-1 receptor agonists
EXSCEL (exenatide once weekly)	Placebo	14,000	A1C 6.5%–10.0%	MACE	April 2018	NCT01144338
FREEDOM-CVO (exenatide ITCA 650)	Placebo	4000	History of coronary, cerebrovascular, or peripheral artery disease	MACE	Completed April 2016	NCT01455896
HARMONY Outcomes (albiglutide)	Placebo	9400	Age ≥40 years with established CVD	MACE	May 2019	NCT02465515
REWIND (dulaglutide)	Placebo	9622	Age ≥50 years with established clinical vascular disease, or age ≥55 years with subclinical vascular disease, or age ≥60 years with ≥2 CV risk factors	MACE	July 2018	NCT01394952
Insulin
DEVOTE (insulin degludec)	Insulin glargine	7637	Age ≥50 years with established CVD or renal disease, or age ≥60 years with CV risk factors	MACE	Completed September 2016	NCT01959529

DPP-4 inhibitors

There are currently four FDA-approved dipeptidyl-peptidase-4 (DPP-4) inhibitors in the United States. It has been demonstrated in preclinical studies that DPP-4 inhibition improves cardiac perfusion and minimizes postinfarction size, most likely as a result of increasing circulating glucagon-like peptide 1 (GLP-1) approximately threefold.¹⁸ DPP-4 inhibition also alters other cardioactive peptides that may contribute to these preclinical outcomes.¹⁸ Unlike GLP-1 receptor agonists (RAs), DPP-4 inhibition has little effect on weight, blood pressure, or lipoproteins.

Three of the DPP-4 inhibitors (alogliptin, saxagliptin, and sitagliptin) have had major CV outcome trials published in the past 3 years.^19
–21 In general, all three trials found that DPP-4 inhibitors do not increase or decrease composite CV outcomes in patients with T2D. It is important to note that the patient populations and trial designs were markedly different from one trial to another.

The SAVOR-TIMI 53 trial compared saxagliptin with placebo in patients with T2D who had a history of or were at risk for CV events.¹⁹ There was no significant difference in the primary MACE endpoint between the saxagliptin group and the placebo group (P < 0.001 for noninferiority; P = 0.99 for superiority). There were also no significant differences in individual endpoints. However, there was a marked difference between groups in hospitalization for heart failure (HF) despite similar rates of HF at baseline between groups (12.8%). More patients in the saxagliptin group were hospitalized during the trial period for HF than placebo (3.5% vs. 2.8%, respectively; HR 1.27; 95% CI 1.07–1.51; P = 0.007).¹⁹

Alogliptin was compared with placebo in patients with T2D in the EXAMINE trial.²⁰ There was no significant difference in the primary MACE outcome between the alogliptin group and the placebo group (P < 0.001 for noninferiority; P = 0.32 for superiority). Baseline HF was present in 27.8% and 28.0% of subjects randomized to placebo and alogliptin, respectively.²⁰

A subsequent trial evaluated risk of hospitalization for HF between the groups.²² The rate of hospitalizations for HF was 3.9% for alogliptin and 3.3% for placebo (HR 1.19; 95% CI 0.90–1.58; P = 0.220) during the 1.5-year trial period. There was no difference in the rate of hospital admissions for patients with existing HF at baseline (HR 1.00; 95% CI 0.71–1.42; P = 0.996), but in individuals with no history of HF at baseline, the difference between groups was more pronounced (alogliptin 2.2% vs. placebo 1.3%; HR 1.76; 95% CI 1.07–2.09; P = 0.026).

A recent publication further assessing CV mortality between alogliptin and placebo demonstrated CV death at a rate of 4.1% for alogliptin group and 4.9% for placebo group (HR 0.85; 95% CI 0.66–1.10), which were not significantly different.²³

The TECOS trial was published in 2015 and compared sitagliptin with placebo in patients with T2D.²¹ There was no significant difference in the primary four-point MACE endpoint between the sitagliptin group and the placebo group (P < 0.001 for noninferiority; HR 0.98; 95% CI 0.89–1.08; P = 0.65 for superiority). There was no difference in the rate of hospitalizations for HF between groups, as demonstrated by a total of 228 patients in the sitagliptin group (3.1%) and 229 patients in the placebo group (3.1%) who were hospitalized for HF (HR 1.00; 95% CI 0.83–1.20; P = 0.98).²¹

The CV outcomes associated with linagliptin are currently being assessed in the CAROLINA (https://clinicaltrials.gov/show/NCT01243424) and CARMELINA (https://clinicaltrials.gov/show/NCT01897532) trials (Table 2). The CAROLINA trial differs from other DPP-4 inhibitor trials in that it will evaluate patients receiving linagliptin compared with glimiperide. The CARMELINA trial will also evaluate the primary renal outcome of time to first occurrence of renal death, sustained end-stage renal disease, or sustained decrease of more than or equal to 50% in estimated glomerular filtration rate.

Sodium–glucose cotransporter-2 inhibitors

The sodium–glucose cotransporter-2 (SGLT-2) inhibitors represent the newest class of medications approved for use in patients with T2D in the United States. There have been numerous theorized CV benefits that agents within this class may produce independent of their glucose-lowering effects, including a reduction in arterial stiffness and subsequent blood pressure lowering, a reduction in body fat mass, and reductions in urinary albumin excretion.^24
–26 There are currently three FDA-approved agents within this class (canagliflozin, dapagliflozin, and emagliflozin), one of which recently published data were from a CV safety trial (empagliflozin).

The EMPA-REG OUTCOME trial was published in November 2015, and was the first diabetes medication CV safety study to demonstrate not only CV safety but also a reduction in MACE with the use of empagliflozin compared with placebo.²⁷ Major trial results are reported in Tables 1 and 3. The primary endpoint occurred in significantly fewer patients treated with empagliflozin compared with placebo (P < 0.001 for noninferiority; P = 0.04 for superiority).²⁷ Of note, this benefit was seen early, with the primary outcome curves diverging after ∼3 months of treatment.

Table 3.

Cardiovascular Results of EMPA-REG OUTCOME, LEADER, and SUSTAIN-6 Trials

	EMPA-REG OUTCOME ²⁷			LEADER ³⁰			SUSTAIN-6 ³¹
	Empagliflozin (%)	Placebo (%)	HR (95% CI)	Liraglutide (%)	Placebo (%)	HR (95% CI)	Semaglutide (%)	Placebo (%)	HR (95% CI)
MACE	10.5	12.1	0.86 (0.74–0.99)	13	14.9	0.87 (0.78–0.97)	6.6	8.9	0.74 (0.58–0.95)
CV death	3.7	5.9	0.62 (0.49–0.77)	4.7	6.0	0.78 (0.66–0.93)	2.7	2.8	0.98 (0.65–1.48)
Nonfatal MI	4.5	5.2	0.87 (0.70–1.09)	6.0	6.8	0.88 (0.75–1.03)	2.9	3.9	0.74 (0.51–1.08)
Nonfatal stroke	3.2	2.6	1.24 (0.92–1.67)	3.4	3.8	0.89 (0.72–1.11)	1.6	2.7	0.61 (0.38–0.99)
HF hospitalization	2.7	4.1	0.65 (0.50–0.85)	4.7	5.3	0.87 (0.73–1.05)	3.6	3.3	1.11 (0.77–1.61)
All-cause death	5.7	8.3	0.68 (0.57–0.82)	8.2	9.6	0.85 (0.74–0.97)	7.4	9.6	0.77 (0.61–0.97)

Examining individual components of the composite primary outcome, there were no significant differences between empagliflozin and placebo in the rates of fatal or nonfatal MI or fatal or nonfatal stroke, but empagliflozin demonstrated a lower rate of death because of CV causes than placebo, translating to a 38% relative risk reduction. The rate of hospitalization for HF was also lower in the empagliflozin group than in placebo, as was death from any cause. There were no significant differences between the 10 and 25 mg empagliflozin treatment groups.²⁷

The results from the EMPA-REG OUTCOME trial demonstrated that 63 patients would need to be treated for 3 years to prevent one MACE and 39 patients would need to be treated with empagliflozin for 3 years to prevent one death from any cause.

Additional CV safety trials for other SGLT-2 inhibitors include the CANVAS (canagliflozin; https://clinicaltrials.gov/show/NCT01032629), DECLARE-TIMI 58 (dapagliflozin; https://clinicaltrials.gov/show/NCT01730534) that was intentionally powered to demonstrate superiority, rather than noninferiority, and VERTIS CV (ertugliflozin; https://clinicaltrials.gov/show/NCT01986881) trials.

GLP-1 receptor agonists

Like the SGLT-2 inhibitors, there have been many theorized CV benefits of the GLP-1 RAs including reductions in glucose, body weight, and blood pressure, as well as anti-inflammatory, antithrombotic, and lipid lowering effects.²⁸ There are currently six GLP-1 RAs approved for use in the United States by the FDA (albiglutide, dulaglutide, exenatide twice daily, exenatide once weekly, liraglutide, and lixisenatide).

The first trial evaluating CV outcomes of a GLP-1 RA was the ELIXA trial, which evaluated the impact of lixisenatide on CV morbidity and mortality in patients with an acute coronary event in the previous 180 days.²⁹ There was no significant difference between lixisenatide and placebo for the primary four-point MACE endpoint (P < 0.001 for noninferiority; P = 0.81 for superiority). There were also no significant differences found between lixisenatide and placebo for the secondary endpoints of the composite of the primary endpoint or hospitalization for HF (15.0% vs. 15.5%; HR 0.97; 95% CI 0.85–1.10; P = 0.63), or the composite of the primary endpoint, hospitalization for HF, or coronary revascularization procedures (21.8% vs. 21.7%; HR 1.00; 95% CI 0.90–1.11; P = 0.96).²⁹

The recently published LEADER trial was the first trial to demonstrate CV benefit with a GLP-1 RA in patients with T2D (Tables 1 and 3).³⁰ The trial compared liraglutide with placebo in patients with T2D and high CV risk. Major trial results are reported in Table 3. The primary outcome occurred in significantly fewer patients taking liraglutide than those taking placebo (P < 0.001 for noninferiority; P = 0.01 for superiority), translating to a 13% relative risk reduction. The benefit seen in the primary outcome began after ∼12–18 months of treatment. Death from CV causes was significantly lower in the liraglutide group than in placebo as was death from any cause. Rates of nonfatal MI and nonfatal stroke were lower in the liraglutide group than in the placebo group, but these differences were not statistically significant.

Subgroup analyses for the primary outcome showed significant interactions with renal function and presence or absence of CVD at baseline. Benefit with liraglutide was shown in patients with an eGFR less than 60 mL/(min·1.73 m²) compared with those with higher eGFRs and patients with established CVD compared with patients with risk factors for CVD. The results of the LEADER trial demonstrate that 66 patients would need to be treated for 3 years to prevent one MACE and 98 patients would need to be treated for 3 years to prevent one death from any cause.³⁰

The recently published SUSTAIN-6 trial is the first preapproval study to demonstrate CV benefit (Tables 1 and 3).³¹ The trial compared semaglutide, a once-weekly GLP-1 RA not yet approved in the United States, with placebo in patients with T2D and high CV risk. The primary MACE outcome occurred in significantly fewer patients taking semaglutide than in those in placebo (P < 0.001 for noninferiority; 0.02 for superiority), translating to a 26% relative risk reduction. The lower risk of MACE was driven mainly by rates of nonfatal MI and nonfatal stroke, which were both lower in the semaglutide group than in the placebo group but only statistically significant for nonfatal stroke (P = 0.12 and 0.04, respectively). Death from CV causes was similar in both groups. The results of the SUSTAIN-6 trial demonstrate that 45 patients would need to be treated for 2 years to prevent one MACE.³¹

Additional CV safety trials for other GLP-1 RAs are anticipated, including the EXSCEL (exenatide once weekly), FREEDOM-CVO (ITCA-650; https://clinicaltrials.gov/show/NCT01455896), REWIND (dulaglutide; https://clinicaltrials.gov/show/NCT01394952), and HARMONY outcomes (albiglutide; https://clinicaltrials.gov/show/NCT02465515) trials. The FREEDOM-CVO trial evaluated ITCA 650, a continuous subcutaneous delivery of exenatide that is not yet approved by the FDA. The trial is completed and top-line results (announced through the manufacturer's press release) indicate that the trial achieved its primary endpoint of showing noninferiority of MACE with ITCA 650 compared with placebo.³²

Discussion

DPP-4 inhibitors

Data from the three CV outcome trials using DPP-4 inhibitors have similar outcomes despite these trials occurring in different patient populations. The TECOS trial (sitagliptin) included patients with established CVD, whereas the SAVOR-TIMI trial included patients with either established CVD or those with multiple risk factors for CVD. There was no statistical difference in MACE rates versus placebo comparators in either trial, although overall rates were higher in the TECOS trial. This is most likely a result of a slight difference in the composite endpoint and the patient populations all having established CVD.

The EXAMINE trial (alogliptin) only included patients who experienced an acute coronary syndrome (ACS) within the previous 15–90 days. There were no differences in the composite endpoints compared with placebo in this trial despite a higher risk population. The primary outcome in this trial occurred at a rate similar to that seen in the TECOS trial. Despite these agents being evaluated in slightly different patient populations, the findings of each support that DPP-4 inhibitors do not improve composite CV outcomes in the first 1.5–3 years after randomization. Long-term trials are needed to determine whether there are potential CV differences with this drug class.

It is possible that patients included in these DPP-4 inhibitor trials may have had more advanced disease that could not be reversed with DPP-4 inhibitors and, therefore, additional trials are needed in patients with shorter durations of diabetes who do not have established CVD to determine whether these agents can prevent the onset of CVD.

An interesting finding from these trials is the differences in rates of hospitalizations for HF between the different agents. In 2015, the FDA added a warning about HF risk to the labels for saxagliptin and alogliptin, but not for sitagliptin. Based on several recent meta-analyses, it appears that saxagliptin was the only DPP-4 inhibitor that was significantly associated with HF hospitalizations, although there does appear to be a signal associated with alogliptin as well.^33
–35 It is unclear from these studies whether the HF risk associated with these agents is drug specific or a class effect as the studies were not powered to detect such differences.

Pathophysiological effects of DPP-4 inhibitors do not appear to be associated with direct cardiotoxicity or result in fluid retention.¹⁹ There is a growing level of evidence associated with DPP-4 substrates having varying effects on peripheral tissue including the CV system, which could explain the potential for increased HF.³⁶ Differences in HF risk between various DPP-4 inhibitors are not conclusive, but one could speculate differences could exist as these agents differ significantly by chemical properties, metabolism, coupling mechanisms with DPP-4 and elimination.³⁷ Additional trials are needed in both patients with existing HF and those without to better understand this possible association and to determine whether this adverse event is drug or class related.

SGLT-2 inhibitors

The positive CV effects of empagliflozin demonstrated in the EMPA-REG OUTCOME trial have generated substantial attention. Treatment with empagliflozin resulted in a significant reduction in MACE compared with placebo with effects starting as early as 3 months after treatment initiation. Interestingly, more than 3 years of treatment did not demonstrate a clear effect on the incidence of either MI or stroke, suggesting that benefits seen with empagliflozin are independent of atherothrombotic events, and that the composite endpoint used may not be the most relevant indicator. The rapidity of the effects on both CV death and HF hospitalization suggests that these results are independent of the drug's effects on glucose lowering, modest weight loss, and blood pressure lowering, all of which would take longer to demonstrate a CV impact.

The diuretic properties that empagliflozin displays could partially explain the large, prompt effects on both CV death and HF; this is consistent with HF trials, in which pronounced mortality benefits occurred quickly.^38,39 Many of the patients in the EMPA-REG OUTCOME trial were already on diuretics for blood pressure control, which raises the questions of whether additional osmotic diuresis caused by the SGLT-2 inhibitor would generate such a pronounced effect so quickly. Possibly the SGLT-2 inhibitor provides more sustained diuresis than short-term or mild diuresis induced by diuretics despite a sustained effect on blood pressure. The overall influence of diuresis by empagliflozin on CV benefit remains unclear at this point.

There have been other proposed hypotheses to explain this pronounced, rapid benefit as well. One prominent theory is that the higher levels of ketone bodies produced with SGLT-2 therapy shift the fuel metabolism in the heart away from fat and glucose oxidation, known to be less efficient in T2D patients, to use ketone bodies as a more efficient fuel source, thus improving myocardial efficiency and function.^40,41 This mechanism would work synergistically with increased oxygen release seen with SGLT-2 inhibition to improve myocardial contractility and cardiac efficiency, with possible dramatic effects in patients at risk for HF.⁴¹ If true, this mechanism would open up new avenues for research and treatment of heart disease.

The completion of both the CANVAS and DECLARE-TIMI 58 trials over the next 3 years will be essential to gauge whether the CV benefits are consistent across the SGLT-2 class, or are specific to empagliflozin. It will also be important to evaluate these effects across other populations of diabetes patients, to gauge whether these results can be generalizable to any T2D patient. Nevertheless, the available data prompted the FDA Endocrine and Metabolic Advisory Panel in June of 2016 to approve an expanded indication for empagliflozin in a close vote of 11–10, voting to include language that empagliflozin may reduce the incidence of CV death in patients with T2D and established CVD.⁴²

GLP-1 receptor agonists

The ELIXA trial established the CV safety of lixisenatide, finding no significant difference in MACE in patients treated with lixisenatide versus placebo. This safety was consistent across all components of the primary outcome and in all subgroup analyses.

Contrastingly, the LEADER and SUSTAIN-6 trials not only showed CV safety but also showed significant CV benefit with liraglutide and semaglutide compared with placebo. The benefit with liraglutide was driven primarily by a reduction in CV death, but was consistent across all components of the primary composite outcome and other secondary outcomes, although not all reached statistical significance. It is important to note that subgroup analyses suggest that the benefit was primarily in patients with moderate to severe renal disease and those with established CVD, which warrants future investigation and consideration. The benefit with semaglutide was driven primarily by reductions in nonfatal MI and nonfatal stroke.

The differing results seen in the ELIXA, LEADER, and SUSTAIN-6 trials raise several questions regarding the CV effects of the GLP-1 RA class. At the forefront is the question of whether the CV benefit can be considered a class effect. There are important differences between the drugs within the GLP-1 RA class in terms of molecular structure, pharmacokinetics, and pharmacodynamics. Lixisenatide is structurally dissimilar to liraglutide and semaglutide and is shorter acting, thereby only intermittently increasing the GLP-1 RA levels. Short-acting GLP-1 RAs are associated with more significant delays in gastric emptying and more targeted reductions in postprandial glucose. Liraglutide is long acting, providing continuously elevated levels. Long-acting GLP-1 RAs have less effect on gastric emptying and rely more on the insulin secretion mechanisms, which lower both postprandial and fasting plasma glucose. Such differences have translated into clinical differences seen in glycemic efficacy, frequency of side effects, and dosing requirements, and could also possibly explain the differences seen in the CV outcome trials.

Differences in the trial designs, patient populations, and durations of the trials could also explain the different results. The ELIXA trial was conducted in a smaller population of post-ACS patients, over a short, 2-year time frame. The LEADER trial was conducted in a larger patient population with established CVD or CVD risk, over a longer 3.5- to 5-year duration. The SUSTAIN-6 trial was conducted in a smaller patient population with established CVD or CVD risk, over a short 2-year time frame. These substantial differences make it impossible to compare the three trials and no conclusion can be drawn regarding the potential CV impact of longer treatment with lixisenatide on patients with T2D and established CVD.

Empagliflozin versus liraglutide and semaglutide: Different paths to CV benefit

Empagliflozin, liraglutide, and semaglutide have all demonstrated CV benefit in patients with T2D and established CVD. It is difficult to compare studies performed in different patient populations with different methodology, but it is useful to evaluate their similarities and differences to incorporate their results into clinical practice decisions (Table 3).

Both the EMPA-REG OUTCOME and LEADER trials showed reductions in MACE, CV death, and all-cause death, but liraglutide's beneficial effects were seen more consistently across the individual components of the MACE endpoint with the hazard ratios (HRs) also favoring liraglutide for nonfatal MI and stroke (although not reaching statistical significance). Empagliflozin's beneficial effects came primarily from all-cause mortality and CV mortality and were less consistent across the MACE components with the HR for nonfatal stroke favoring placebo (although not reaching statistical significance). Semaglutide's beneficial effects came primarily from nonfatal MI and nonfatal stroke and were not consistent across all MACE components.

Both liraglutide and semaglutide also had no impact on HF hospitalizations, whereas empaglifozin resulted in lower rates of HF hospitalizations. Another important distinction is that the beneficial effects seen with empagliflozin in the EMPA-REG OUTCOME trial started very early after 3 months of treatment, whereas the beneficial effects of liraglutide in the LEADER trial were more gradual, starting after 12–18 months of treatment. Beneficial effects of semaglutide in the SUSTAIN-6 trial started to appear at 4 months.

These differences suggest different mechanisms beyond their shared effects on surrogate markers such as glucose, blood pressure, and weight. The mechanism of empagliflozin may be more related to hemodynamics or ketone body formation, whereas the mechanism of liraglutide may be more related to the basic biology of atherosclerosis.

The complexity of macrovascular complications

The effects of good glucose control on prevention of microvascular complications of diabetes have been well demonstrated and are unequivocal.^{4,5,7,9,43
–45} Since hyperglycemia is considered to be the main and direct cause of microvascular complications, it is not surprising that controlling hyperglycemia prevents these complications. Since these microvascular complications significantly impair quality of life and may also predispose to a shortened life span, prevention of microvascular complications alone is sufficient justification for good glucose control.

The effects of good glucose control on prevention of macrovascular complications, specifically CV outcomes, have been variable and for the most part have not supported the notion that CV events can be prevented by attention to glycemic control alone.^4

–8 The pathophysiology of macrovascular disease clearly extends beyond hyperglycemia alone. Multiple CV risk factors such as, but not limited to, genetics, hypertension, hyperlipidemia, poor diet, physical inactivity, and smoking all uniquely and in combination contribute to the development of macrovascular disease. This may explain why control of hyperglycemia has not been consistently effective in preventing CV events. It may be that the adverse influence of multiple CV risk factors in combination greatly outweighs the beneficial but smaller effects of glucose control.

Alternatively, the concomitant control of multiple other risk factors may be so beneficial that the incremental benefit from glycemic control is difficult to discern. In addition, the harmful effects of intermittent hypoglycemia, a frequent occurrence in many of the glucose control trials, likely counterbalance and offset the beneficial effects.⁶ Therefore, it is quite possible that lowering blood glucose profiles would have much more obvious beneficial CV outcome effects on the absence of intermittent hypoglycemia.

Although control of glucose alone, by any number of treatment strategies, has not been shown to provide a CV benefit, a different picture has emerged from the EMPA-REG OUTCOME, LEADER, and SUSTAIN-6 trials. It is particularly interesting that two of the medications with reported beneficial CV outcomes, a SGLT-2 inhibitor (empagliflozin) and a GLP-1 RA (liraglutide), only reduced blood glucose levels modestly and not completely into the target range in the safety trials. This suggests that at least part of their favorable CV benefit may derive from some of their nonglycemic actions.

Empagliflozin promotes weight loss, reduces blood pressure modestly without raising the heart rate, promotes renal sodium loss and consequent plasma volume reduction, raises the hematocrit through hemoconcentration, and raises HDL cholesterol (but also raises LDL). This medication also increases serum ketones, especially beta hydroxybutyrate, which may be a more efficient fuel for the myocardium.^40,41 Liraglutide also promotes weight loss, lowers blood pressure, and has modest beneficial effects on the serum lipid profile. Avoidance of hypoglycemia may also have been a major contributing factor.

Notably, both of these medications have significantly lower risks of hypoglycemia than insulin and medications with more strongly insulin-dependent mechanisms of action. Furthermore, subjects in the placebo arms of these studies often required escalating doses of these more hypoglycemia-prone medications compared with subjects in the active treatment arms.

Several unanswered, or incompletely answered, questions remain to be resolved through ongoing and future research: Are the demonstrated CV benefits reported with empagliflozin, liraglutide, and semaglutide specific to these agents or are they class effects that may later be shown to occur with other medications in the same class? Given their different mechanisms of action, might there be an additive CV benefit if these medication classes were used in combination? Do the mechanisms of CV benefit relate to direct effects on the myocardium, coronary vasculature, or peripheral vasculature? Given the short duration of these studies, is the CV benefit maintained over time? What is the relative importance of HF to overall CV outcomes and why have different diabetes medications resulted in differing results related to HF outcomes?

Clinical implications

Since CVD is the major cause of mortality in patients with T2D, the relative position of each medication in the clinical practice guideline hierarchy must be considered in the context of that medication's effects on CVD. Certainly it is important to first determine that a diabetes medication has no adverse CV consequences, such as an increase in MI, stroke, or HF. In the absence of adverse CV consequences, the beneficial effect of glucose lowering on microvascular complications is surely adequate justification for their use.

However, if diabetes medications, by any mechanism, also reduce adverse CV outcomes, it would be reasonable and perhaps wise to position these medications more prominently in clinical practice guidelines to be used preferentially over medications without demonstrated beneficial CV outcome data. The American Diabetes Association Standards of Medical Care currently acknowledge that empagliflozin is associated with a lower risk of CVD, but do not include any specific recommendations.⁴⁶ The Canadian Diabetes Association guidelines were updated in 2016 to recommend using an SGLT-2 inhibitor with demonstrated CV outcome benefit (i.e., empagliflozin) in people with CVD in whom glycemic targets are not met.⁴⁷

With these considerations in mind, both empagliflozin and liraglutide should be preferred agents for glucose control because of multiple clinical attributes. They effectively and durably lower glucose, they lower blood pressure and weight, have a low risk for causing hypoglycemia, and they target specific pathophysiological defects of T2D and have good safety profiles. They have also both demonstrated improvements in CV outcomes in patients with T2D and established CVD. Semaglutide may also be preferred after it gains FDA approval if other efficacy and safety features are comparable. Of course, the recognition of the need for an individualized, patient-centered approach is essential and must take specific patient characteristics, adverse effects, ease of use, and cost into consideration.

Conclusion

To date, three DPP-4 inhibitors, three GLP-1 RAs, and one SGLT-2 inhibitor have demonstrated CV safety in major CV outcome trials. Empagliflozin, liraglutide, and semaglutide have also demonstrated CV benefit. Although the mechanism of CV benefit remains unclear for these agents, the risk reduction of MACE cannot be ignored. For patients with T2D and established CVD, both empagliflozin and liraglutide should be considered as preferred second-line treatment options, as should semaglutide, once approved. An individualized, patient-specific treatment approach, with considerations for safety, cost, and treatment burden, remains crucial. Future trials are needed to bring clarity to questions related to mechanism of CV benefit, class effects, potential CV benefits of combination therapies, and the CV effects of T2D medications in lower CV risk patients.

Footnotes

Author Disclosure Statement

J.M.T. is an advisory board member and consultant for Sanofi. S.L.E., W.A.N., S.A.W., and M.T.M. have no competing financial interests.

References

International Diabetes Federation: IDF Diabetes, 7th ed. Brussels, Belgium: International Diabetes Federation, 2015. www.diabetesatlas.org (accessed June 20, 2016 ).

Centers for Disease Control and Prevention: National Diabetes Statistics Report: Estimates of Diabetes and Its Burden in the United States, 2014. Atlanta, GA: US Department of Health and Human Services, 2014.

Gregg

, Li

, Wang

, et al.: Changes in diabetes-related complications in the United States, 1990–2010. N Engl J Med, 2014; 370:1514–1523.

UK Prospective Diabetes Study (UKPDS) Group: Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet, 1998; 352:837–853.

UK Prospective Diabetes Study (UKPDS) Group: Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). Lancet, 1998; 352:854–865.

Gerstein

, Miller

, Byington

, et al.; Action to control cardiovascular risk in Diabetes Study Group: Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med, 2008; 358:2545–2559.

Patel

, MacMahon

, Chalmers

, et al.; ADVANCE Collaborative Group: Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med, 2008; 358:2560–2572.

Duckworth

, Abraira

, Moritz

, et al.; VADT Investigators: Glucose control and vascular complications in veterans with type 2 diabetes. N Engl J Med, 2009; 360:129–139.

Holman

, Paul

, Bethel

, et al.: 10-year follow-up of intensive glucose control in type 2 diabetes. N Engl J Med, 2008; 359:1577–1588.

10.

Nissen

, Wolski

: Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med, 2007; 356:2457–2471. [Erratum, N Engl J Med 2007;357:100.]

11.

Hiatt

, Kaul

, Smith

: The CV safety of diabetes drugs-insights from the rosiglitazone experience. N Engl J Med, 2013; 369:1285–1287.

12.

Home

, Pocock

, Beck-Nielsen

, et al.: Rosiglitazone evaluate for cardiovascular outcomes in oral agent combination therapy for type 2 diabetes (RECORD): a multicenter, randomized, open-label trial. Lancet, 2009; 373:2125–2135.

13.

U.S. Food and Drug Administration: Drug Safety Communication: FDA Requires Removal of Some Prescribing and Dispensing Restrictions for Rosiglitazone-Containing Medicines. www.fda.gov/Drugs/DrugSafety/ucm376389.htm (accessed October 1, 2016 ).

14.

U.S. Food and Drug Administration: Guidance for Industry: Diabetes Mellitus—Evaluating CV Risk in New Antidiabetic Therapies to Treat Type 2 Diabetes. Silver Spring, MD: Department of Health and Human Services, Center for Drug Evaluation and Research, 2008. www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm071627.pdf (accessed July 22, 2016 ).

15.

The ORIGIN Trial Investigators: Basal insulin and CV and other outcomes in dysglycemia. New Engl J Med, 2012; 367:319–328.

16.

Dormandy

, Charbonnel

, Eckland

DJA

, et al.: Secondary prevention of macrovascular events in patients with T2D in the PROactive trial (prospective pioglitazone clinical trial in macrovascular events): a randomized controlled trial. Lancet, 2005; 366:1279–1289.

17.

Kernan

, Viscoli

, Furie

, et al.: Pioglitazone after ischemic stroke or transient ischemic attack. N Engl J Med, 2016; 374:1321–1331.

18.

Ussher

, Drucker

: Cardiovascular actions of incretin-based therapies. Circ Res, 2014; 114:1788–1803.

19.

Scirica

, Bhatt

, Braunwald

, et al.: Saxagliptin and CV outcomes in patients with T2D mellitus. N Engl J Med, 2013; 369:1317–1326.

20.

White

, Cannon

, Heller

, et al.: Alogliptin after acute coronary syndrome in patients with T2D. N Engl J Med, 2013; 369:1327–1335.

21.

Green

, Bethel

, Armstrong

, et al.: Effect of sitagliptin on CV outcomes in T2D. N Engl J Med, 2015; 373:232–242.

22.

Zannad

, Cannon

, Cushman

, et al.: Heart failure and mortality outcomes in patients with T2D taking alogliptin versus placebo in EXAMINE: a multicentre, randomised, double-blind trial. Lancet, 2015; 385:2067–2076.

23.

White

, Kupfer

, Zannad

, et al.: Cardiovascular mortality in patients with type 2 diabetes and recent acute coronary syndromes from the EXAMINE trial. Diabetes Care, 2016:dc160303.

24.

Inzucchi

, Bergenstal

, Buse

, et al.: Management of hyperglycemia in type 2 diabetes, 2015: a patient centered approach: update to a position statement of the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care, 2015; 38:140–149.

25.

Cherney

, Perkins

, Soleymanlou

, et al.: The effects of empagliflozin on arterial stiffness and heart rate variability in subjects with uncomplicated type 1 diabetes mellitus. Cardiovasc Diabetol, 2014; 13:28.

26.

Yale

, Bakris

, Cariou

, et al.: Efficacy and safety of canagliflozin in subjects with type 2 diabetes and chronic kidney disease. Diab Obes Metab, 2013; 15:463–473.

27.

Zinman

, Wanner

, Lachin

, et al.: Empagliflozin, CV outcomes, and mortality in T2D. N Engl J Med, 2015; 373:2117–2128.

28.

Drucker

: The CV biology of glucagon-like peptide-1. Cell Metabolism, 2016. DOI:10.1016/j.cmet.2016.06.009

29.

Pfeffeer

, Claggett

, Diaz

, et al.: Lixisenatide in patients with T2D and acute coronary syndrome. N Engl J Med, 2015; 373:2247–2257.

30.

Marso

, Daniels

, Brown-Frandsen

, et al.: Liraglutide and cardiovascular outcomes in type 2 diabetes (LEADER). N Engl J Med, 2016. DOI:10.1056/NEJMoa1603827

31.

Marso

, Bain

, Consoli

, et al.: Semaglutide and cardiovascular outcomes in type 2 diabetes. N Engl J Med, 2016. DOI:10.1056/NEJMoa1607141

32.

Intarcia Therapeutics, Inc.: Press Release: Intarcia Announces Successful CV Safety Results in Phase 3 FREEDOM-CVO Trial for ITCA 650, an Investigational Therapy for T2D. www.intarcia.com/media/press-releases/2016-may-6-CV-safety.html (accessed June 24, 2016 ).

33.

Kundu

, Sardar

, Ghosh

, et al.: Risk of heart failure with dipeptidyl peptidase-4 inhibitors in patients with type 2 diabetes mellitus: a meta-analysis of randomized controlled trials. Int J Cardiol, 2016; 212:203–205.

34.

Kongwatcharapong

, Dilokthornsakul

, Nathisuwan

, et al.: Effect of dipeptidyl peptidase-4 inhibitors on heart failure: a meta-analysis of randomized clinical trials. Int J Cardiol, 2016; 211:88–95.

35.

, Li

, Deng

, et al.: Dipeptidyl peptidase-4 inhibitors and risk of heart failure in type 2 diabetes: systematic review and meta-analysis of randomised and observational studies. BMJ, 2016; 352:i610.

36.

, O'Leary

, Broxmeyer

: Implications of DPP4 modification of proteins that regulate stem/progenitor and more mature cell types. Blood, 2013; 122:161–169.

37.

Deacon

: Dipeptidyl peptidase-4 inhibitors in the treatment of type 2 diabetes: a comparative review. Diabetes Obes Metab, 2011; 13:7–18.

38.

Pitt

, White

, Nicolau

, et al.: Eplerenone reduces mortality 30 days after randomization following acute myocardial infarction in patients with left ventricular systolic dysfunction and heart failure. J Am Coll Cardiol, 2005; 46:425–431.

39.

Pitt

, Zannad

, Remme

, et al.: The effect of spironolactone on morbidity and mortality in patients with severe heart failure. N Engl J Med, 1999; 341:709–717.

40.

Mudaliar

, Alloju

, Henry

: Can a shift in fuel energetics explain the beneficial cardiorenal outcomes in the EMPA-REG OUTCOME trial? A unifying hypothesis. Diabetes Care, 2016; 39:1115–1122.

41.

Ferrannini

, Mark

, Mayoux

: CV protection in the EMPA-REG OUTCOME trial: a “thrifty substrate” hypothesis. Diabetes Care, 2016; 39:1108–1114.

42.

U.S.. Food and Drug Administration Briefing Document. Endocrine and Metabolic Drug Advisory Committee Meeting, June 28, 2016. www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/EndocrinologicandMetabolicDrugsAdvisoryCommittee/UCM508422.pdf (accessed August 3, 2016 ).

43.

Ismail-Beigi

, Craven

, Banerji

, et al.; for the ACCORD Trial Group: Effect of intensive treatment of hyperglycemia on microvascular outcomes in type 2 diabetes: an analysis of the ACCORD randomized trial. Lancet, 2010; 376; 419–430.

44.

Ohkubo

, Kishikawa

, Araki

, et al.: Intensive insulin therapy prevents the progression of diabetic microvascular complications in Japanese patients with non-insulin-dependent diabetes mellitus: a randomized prospective 6-year study. Diabetes Res Clin Pract, 1995; 28:103–117.

45.

Shichiri

, Kishikawa

, Ohkubo

, Wake

: Long-term results of the Kumamoto study on optimal diabetes control in type 2 diabetic patients. Diabetes Care, 2000; 23(Suppl 2): B21–B29.

46.

American Diabetes Association: Standards of medical care in diabetes 2016. Cardiovascular disease and risk management. Diabetes Care, 2016; 39(Suppl 1):S60–S71.

47.

Canadian Diabetes Association Clinical Practice Guidelines Expert Committee: Pharmacologic Management of Type 2 Diabetes—2016 Interim Update. http://guidelines.diabetes.ca/executivesummary/ch13_2016 (accessed October 2, 2016 ).