In choosing drugs to treat Type 2 diabetes, insulin, sulfonylureas and other secretagogues are not a prudent initial choice due to the increased risk of hypoglycemia associated with their use.1 Similarly, drugs that lead to weight gain (insulin, secretagogues, thiazolidinediones) should be avoided due to the potential of increasing insulin resistance.2
Since the vast majority of Type 2 diabetes patients are obese, there is a need to not only avoid weight gain but also to use therapies that, in addition to lowering glucose levels, promote weight loss, which will result in decreased intramuscular fat (lowering insulin resistance) and myocardial fat (improving cardiac function). In addition, decreasing beta cell fat will decelerate beta cell apoptosis (programmed cell death) and improve insulin secretion, and a reduction in hepatic fat will lower hepatic resistance to insulin’s action, which in turn will decrease hepatic glucose production.2,3 Therefore, drugs that have the potential to lower body weight, i.e., GLP-1 receptor agonists and sodium-glucose cotransporter-2 (SGLT2) inhibitors, should, along with metformin (mostly weight neutral), be the preferred initial therapies for Type 2 diabetes. The use of this triad (metformin, GLP1-receptor agonists and SGLT2 inhibitors) is consistent with the recommendations of the American Association of Clinical Endocrinologists’ (AACE) comprehensive diabetes-management algorithm.4
The liraglutide and SGLT2 inhibitor connection
To date, with or without metformin, prospective randomized studies of the majority of GLP1-receptor agonists (liraglutide, dulaglutide, lixisenatide or albiglutide) in combination with an SGLT2 inhibitor have not been reported. However, a 28-week study of the oral SGLT2 inhibitor dapagliflozin once daily combined with weekly exenatide showed an HbA1c reduction of 2 percent accompanied by a 3.4-kg weight loss, both of which were superior to monotherapy with either drug alone.5
The combination of liraglutide and an SGLT2 inhibitor has not been studied in a prospective, randomized trial. However, to illustrate his clinical experience that this is a powerful combination, David S. H. Bell, MB, FACE, FACP (the lead author of this paper), reported a case in which the combination of liraglutide and canagliflozin resulted not only in better glycemic control but also the discontinuation of insulin in addition to a weight loss of 12 kg (from 112 kg to 100 kg).6 The case report included a retrospective study of 15 subjects utilizing this combination, which saw HbA1c drop from 9.1 to 7 percent. Additionally, the mean body weight reduction from 113 kg to 108 kg is similar to that achieved in the prospective study of dapagliflozin and exenatide.5,6 Extending this observational study to three years resulted in a 1.5 percent decrease in HbA1c (from 8.1 to 6.6 percent) accompanied by an 8-kg weight loss (from 102 kg to 94 kg). A similar retrospective study showed that adding canagliflozin to the treatment regimen of 75 subjects with Type 2 diabetes who had been on a GLP-1 agonist for at least one year resulted in an additional 0.39 percent drop in HbA1c and an additional 4.6-kg weight loss.7
These striking outcomes of combining an SGLT2 inhibitor and a GLP-1 receptor agonist in both prospective and retrospective studies revealed a need to evaluate long-term outcomes of the combination. To that end, Edison Goncalves, MD, and Bell reviewed the records of 79 patients in a large practice who were being treated with the combination of an SGLT2 inhibitor and the GLP-1 receptor agonist liraglutide.8 Since liraglutide was approved three years before SGLT2 inhibitors became available, the authors were able to discriminate between sequential and simultaneous initiation of therapy.8 The review included 46 sequentially treated subjects (age 60.5 ± 7.1 years, 23 percent African American, 77 percent of European origin, with a duration of diabetes of 11 ± 6 years) and 33 simultaneously treated subjects (age 51 ± 10 years, 27 percent African American, 70 percent of European origin, with a duration of diabetes of 9.3 ± 6 years). In the sequential group, HbA1c dropped from 8.9 ± 1.3 percent to 7.6 ± 1.1 percent (p=0.001), which compared favorably to the simultaneous group, in which HbA1c dropped from 9.1 ± 1.4 percent to 7.1 ± 0.9 percent (p<0.001). With liraglutide alone, weight dropped from 111 ± 21 kg to 105 ± 20 kg (p<0.001). Adding an SGLT2 inhibitor resulted in further weight loss to 101 ± 20 kg (p=0.001). With simultaneous initiation, weight dropped from 116 ± 25 kg to 106 ± 22 kg (p<0.001). Of note, weight loss plateaued at six months in the sequential group and did not increase over the subsequent 141 ± 74 weeks before an SGLT2 inhibitor was added (Figure 1).8
In this long-term study, systolic blood pressure dropped significantly—by 6 mm Hg with liraglutide alone and by a further 7 mm Hg when an SGLT2 inhibitor was added. This compares favorably with the simultaneous initiation, in which systolic blood pressure dropped significantly by 13 mm Hg. There were no significant changes in diastolic blood pressure. Unlike the symbiotic effects of these drugs on blood pressure, there were opposing effects on the lipid profile—liraglutide significantly decreased total cholesterol, LDL cholesterol and non-HDL cholesterol. However, these positive decreases became nonsignificant when an SGLT2 inhibitor was added. The simultaneous group had a significant decline in triglycerides and non-HDL cholesterol, possibly due to improved glycemic control.8
With this combination, the decreases in the cardiac risk factors (body weight, systolic blood pressure, non-HDL cholesterol, triglycerides and HbA1c) accompanied by deceleration of the decline in renal function and the absence of severe hypoglycemia should result in decreased cardiac events. There is already evidence that this may be the case.
The Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results (LEADER) trial randomized 9,340 patients to either liraglutide or placebo. Over an average of 3.8 years, significantly fewer subjects in the liraglutide group had a cardiac event (nonfatal myocardial infarction, stroke or cardiac death) than in the placebo group (13 percent versus 14.9 percent; HR 0.87; CI 0.78 to 0.97; p<0.001 for inferiority; p=0.01 for superiority).9 In a similarly designed study, semaglutide—a long-acting GLP-1 receptor agonist—decreased cardiac events compared to placebo (6.6 percent versus 8.9 percent; HR 0.74; CI 0.58 to 0.95; p=0.01 for non-inferiority).10 However, the short-acting lixenatide was not shown to increase or decrease cardiac events.11
As far as evaluating SGLT2 receptor blockers, the Empagliflozin Cardiovascular Outcomes and Mortality in Type 2 Diabetes (EMPA-Reg) study, in which 7,020 patients were randomized to empagliflozin or placebo over an average of 3.8 years, those on empagliflozin had a lower rate of cardiovascular outcomes (myocardial infarction, stroke and cardiovascular death) (10.5 percent versus 12.1 percent; HR 0.86; CI 0.74-0.99; p=0.04 for superiority).12 In 2017, the results of the CANagliflozin CardioVascular Assessment Study (CANVAS) will be reported and, if positive, will suggest a cardioprotective effect of SGLT2 inhibitors. As far as retrospective studies are concerned, a recently published population-based retrospective cohort study of a large primary care British database looked at 22,124 Type 2 diabetes subjects (average HbA1c 7.7 percent, age 58.4 years, BMI 34.8), including 4,350 with previous cardiac events and 4,444 who had used dapagliflozin. Over an average of 33 months, those who had taken dapagliflozin had a significantly decreased risk of death from any cause (HR 0.50; CI 0.33-0.75; p=0.001). After adjusting for key variables in those without a previous event and who had used dapagliflozin, the reduction in risk was still significant (HR 0.44; CI 0.25 to 0.78; p=0.002).13
The mechanisms of cardioprotection probably differ between a SGLT2 receptor blocker and a GLP-1 receptor agonist. With SGLT2 inhibitors, cardioprotection likely results from improved myocardial function owing to decreased cardiac workload, possibly deriving from utilization of ketones as an energy substrate, increases in myocardial magnesium and calcium, or simply preload and afterload reduction.14 With GLP-1 receptor agonists, cardioprotection is more likely to be vascular and may well result from decreased coronary artery events, since GLP-1 receptors are expressed on both myocardiocytes and vascular endothelium. Receptor activation will likely result in increased arterial dilatation and decreased inflammation.15 Therefore, these proposed cardioprotective mechanisms of SGLT2 inhibitors and GLP-1 receptor agonists are likely to complement each other and follow a different time course.
The combination of an SGLT2 inhibitor and a GLP-1 receptor agonist, which has a complementary effect on glycemic control without causing hypoglycemia, increases weight loss, decreases cardiovascular risk factors and potentially cardiovascular events, should be ideal with or without metformin for the Type 2 diabetes patient. Why, in spite of a recommendation from a major professional association, is this “Cadillac combination” not being utilized?4
One explanation may be higher cost, resulting in third-party payers’ reluctance to approve and pay for the components of the combination. However, coverage cannot be expected in the absence of prospective, randomized studies. So why have these studies not been performed? Pharmaceutical companies are likely reluctant to study a combination of one of their drugs with a non-generic drug for which they do not have a patent. The one study that has been performed looked at the combination of once-weekly exenatide with once-daily dapagliflozin, the patents for both of which are held by the same company.6 Therefore, more prospective studies of the combination of SGLT2 inhibitors and GLP-1 receptor agonists are urgently needed.
Access additional resources and practical information to enhance the care and treatment of your diabetes patients.
About our experts:
David S. H. Bell, MB, FACE, FACP, Professor of Medicine (retired), University of Alabama and Southside Endocrinology, Birmingham, Alabama
Edison Goncalves, MD, FACE, FACP, Diabetes and Thyroid Associates, Birmingham, Alabama