GLP-1 Receptor Agonists and SGLT-2 Inhibitors as Adjuncts to Insulin in Type 1 Diabetes: Benefits and Concerns

The metabolic control of type 1 diabetes (T1D) is more difficult and complicated than type 2 diabetes (T2D). Although novel basal and prandial insulin analogues and systems for continuous glucose monitoring connected to smart insulin pumps/pens are now available, more than 70% of individuals with T1D do not meet the recommended glucose targets. Intensification of treatment with insulin is associated with an increased risk of hypoglycemia and weight gain. Furthermore, in T1D, there is an increasing prevalence of obesity, dyslipidemia, depression, and other risk factors for cardiovascular disease (CVD). The paracrine regulation of glucagon secretion by insulin in the islet environment is also lost in T1D leading to hyperglucagonemia. Together with these drawbacks, insulin resistance (IR) is present from the early stages of the disease contributing substantially to metabolic dysregulation and β-cell loss (1). The treatment of T1D is therefore staggering, and the way to tackle this is by assigning therapeutic strategies that include targets not only within but also beyond glycemic control. Subcutaneous insulin reaches peripheral tissues before it reaches the liver (2). This results in sustained peripheral hyperinsulinemia, induces, or aggravates IR and obesity, and predisposes the patients to frequent episodes of hyperglycemia and hypoglycemia thereby increasing glucose fluctuations and oxidative stress, major pathogenetic mechanisms of CVD (1, 2). Therefore, it is important for T1D to keep insulin doses per day as low as possible; this can be achieved by increasing insulin sensitivity (IS), suppressing hyperglucagonemia, and eliminating weight gain. The regulation of glucose metabolism is complex and involves insulin-dependent and insulin-independent mechanisms (3). During fasting, the insulin-to-glucagon ratio decreases to allow the increase of endogenous glucose production (EGP). In the postprandial state, the insulin-to-glucagon ratio increases to allow suppression of EGP and increase glucose uptake (GU) in muscle. During meals, glucagon-like peptide-1 (GLP-1) is an essential component of the mechanisms regulating glucose metabolism helping insulin to adjust carbohydrate ingestion to metabolic needs (3). GLP-1 retards gastric emptying and intestinal glucose absorption; potentiates insulin secretion and inhibits glucagon release by glucose-dependent mechanisms; helps insulin to suppress EGP and increase GU by direct effects on the liver and muscle; promotes satiety by effects on the central nervous system, thereby decreasing energy intake and body weight; and increases tissue blood flow by effects on the endothelium. Furthermore, GLP-1 stimulates β-cell proliferation/neogenesis and inhibits β-cell apoptosis, thereby helping to maintain β-cell mass, activates cardioprotective/neuroprotective mechanisms, increases natriuresis contributing to the regulation of blood pressure, and alleviates glucotoxicity/lipotoxicity and oxidative stress (4). These pleiotropic mechanisms make GLP-1 receptor agonists (GLP-1RA) appropriate candidates for adjunctive treatment in T1D (5). Sodium glucose cotransporter-2 inhibitors (SGLT-2i) block the reabsorption of glucose and sodium in the kidneys, inducing glucosuria and natriuresis. They decrease hyperglycemia and glucotoxicity without the intervention of insulin and decrease weight and blood pressure. Administration of SGLT-2i improves IS, β-cell function, and fasting/postprandial hyperglycemia, although plasma glucagon levels and EGP are increased, muscle GU is decreased and there is a shift in muscle substrate utilization from glucose to lipids (6). The differences in the efficacy and safety between GLP-1RA and SGLT-2i in T1D have never been so clearly demonstrated as in the study by Edwards et al (7). These authors performed a retrospective investigation of long-acting GLP-1RA and SGLT-2i as adjuncts to insulin in adults with T1D on multiple insulin injections or insulin pumps, in a real-world clinical setting. Although off-label, these agents are frequently prescribed in clinical practice. The efficacy and safety assessments for both treatments are the longest so far, and the fact that this study was performed under free-living conditions adds value to the results. After 1 year of therapy, GLP-1RA reduced weight, glycated hemoglobin A1c (HbA1c), and total (basal plus prandial) dose of insulin per day (32%). SGLT-2i showed reductions only in basal insulin (18%) and HbA1c, while prandial insulin and weight were not decreased. HbA1c reductions were similar between treatment groups. GLP-1RA but not SGLT-2i reduced total cholesterol and low-density lipoprotein by 14% and 26%, respectively. There was a 4-fold increase in diabetic ketoacidosis (DK) in SGLT-2i vs GLP-1RA users, while the incidence of severe hypoglycemia was small and identical in both groups, as expected (1.3% vs 2.6% for GLP-1RA and SGLT-2i, respectively). The conclusions from this study are clinically important. First, under free-living conditions, glycemia improved similarly under the 2 treatments. However, GLP-1RA were more effective than SGLT-2i in inducing reductions in weight and basal plus prandial insulin requirements, suggesting an increase of IS both in fasting and postprandial conditions. Second, GLP-1RA but not SGLT-2i decreased atherogenic lipids, supporting previous findings. Third, discontinuation of treatment because of any adverse events (AEs) was similar after GLP-1RA (26.9%, due mostly to gastrointestinal symptoms) or SGLT-2i (27.7%, due mostly to DK and genitourinary infections). As the authors point out, DK in the real world is more frequent than in clinical trials and should be closely monitored. Dapagliflozin was authorized as an adjunctive treatment in individuals with T1D and obesity; however, subsequently, this approval was withdrawn because of a 13-fold increase in the frequency of DK in this subpopulation (8). SGLT-2i-induced ketogenesis is due to hyperglucagonemia, insulin deficiency, and dehydration (8). SGLT-2i users should measure ketone bodies frequently in blood since these medications may diminish their excretion in the urine, making these measurements less reliable. Several issues remain to be determined. In the study by Edwards et al (7) and all previous reports on T1D, liraglutide was the most frequently used GLP-1RA. It is important to investigate whether there are differences in compliance, effectiveness, and AEs with the recently released once-weekly GLP-1RA. These novel agents are more effective and induce fewer gastrointestinal symptoms following dose titration. The new dual GLP-1/GIP agonist may also be worth investigating. Measuring C-peptide before initiating treatment with GLP-1RA or SGLT-2i may be useful. Residual insulin secretion facilitates glycemic control with fewer insulin units per day and reduces the risk for hypoglycemia and DK; GLP-1RA can help to maintain residual β-cell mass. In T2D, SGLT-2i and GLP-1RA decrease the risk for CVD and progression of chronic kidney disease independently of glycemic control. It remains to be investigated whether this applies also to T1D. The authors received no financial support for the research, authorship, and/or publication of this article. G.D. has given lectures, attended conferences, participated in trials, and received research support from Abbott, AstraZeneca, Boehringer Ingelheim, Eli Lilly, ELPEN, MSD, Novartis, Novo Nordisk, Sanofi, and VIANEX. V.L. has given lectures, attended conferences, and participated in trials sponsored by Novo Nordisk, Sanofi-Aventis, Novartis, AstraZeneca, Boehringer Ingelheim, MSD, GSK, VIANEX, Elli Lilly, ELPEN, Mylan and Aegereon, Amryt, and Amgen. adverse event cardiovascular disease diabetic ketoacidosis endogenous glucose production glucagon-like peptide 1 GLP-1 receptor agonist glucose uptake glycated hemoglobin A1c insulin resistance insulin sensitivity sodium glucose cotransporter-2 inhibitor type 1 diabetes type 2 diabetes