Glucagon-like peptide 1-based therapies for the treatment of type 2 diabetes mellitus

INTRODUCTION — Glucagon-like peptide 1 (GLP-1)-based therapies (eg, GLP-1 receptor agonists, dual-acting GLP-1 and glucose-dependent insulinotropic polypeptide [GIP] receptor agonists, dipeptidyl peptidase 4 [DPP-4] inhibitors) affect glucose control through several mechanisms, including enhancement of glucose-dependent insulin secretion, slowed gastric emptying, and reduction of postprandial glucagon and food intake (table 1). These agents do not usually cause hypoglycemia in the absence of therapies that otherwise cause hypoglycemia.

This topic will review the mechanism of action and therapeutic utility of GLP-1-based therapies for the treatment of type 2 diabetes mellitus. The role of GLP-1 in the treatment of type 1 diabetes has been investigated but is not well defined [1-3]. We do not use GLP-1-based therapies in patients with type 1 diabetes specifically for glycemic management; this discussion will be limited to its use in type 2 diabetes. GLP-1 receptor agonists are also used for weight loss, but their role in weight loss in persons without diabetes is covered separately. (See "Obesity in adults: Drug therapy".)

DPP-4 inhibitors increase endogenous GLP-1 via inhibition of DPP-4. These agents, as well as a general discussion of the initial management and the management of persistent hyperglycemia in adults with type 2 diabetes, are also presented separately.

●(See "Dipeptidyl peptidase 4 (DPP-4) inhibitors for the treatment of type 2 diabetes mellitus".)

●(See "Initial management of hyperglycemia in adults with type 2 diabetes mellitus".)

●(See "Management of persistent hyperglycemia in type 2 diabetes mellitus".)

●(Related Pathway(s): Diabetes: Initial therapy for non-pregnant adults with type 2 DM.)

●(Related Pathway(s): Diabetes: Medication selection for non-pregnant adults with type 2 DM and persistent hyperglycemia despite monotherapy.)

GASTROINTESTINAL PEPTIDES — Glucose homeostasis is dependent upon a complex interplay of several hormones: insulin and amylin, produced by pancreatic beta cells; glucagon, produced by pancreatic alpha cells; and gastrointestinal peptides, including glucagon-like peptide 1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP; formerly called gastric inhibitory polypeptide). GLP-1 and GIP are "incretin" hormones that link the absorption of nutrients from the gastrointestinal tract with pancreatic hormone secretion. They are released in the setting of a meal, after the ingestion and absorption of glucose, protein, and fat (figure 1) [4,5] and provide one of the physiologic connections between eating and insulin release. Abnormal regulation of these peptides may contribute to the development of type 2 diabetes.

●GLP-1 – GLP-1 is produced from the proglucagon gene in L cells of the small intestine. It binds to a specific GLP-1 receptor, which is expressed in various tissues, including pancreatic beta cells, pancreatic ducts, gastric mucosa, kidney, lung, heart, skin, immune cells, and the hypothalamus [4,6]. GLP-1 exerts its main effect by stimulating glucose-dependent insulin release from the pancreatic islets [4]. It has also been shown to slow gastric emptying [7], inhibit inappropriate post-meal glucagon release [8,9], and reduce food intake (table 1 and figure 1) [9]. In patients with type 2 diabetes, there is an impaired insulin response to GLP-1, possibly related to a reduction in postprandial GLP-1 secretion (figure 2A-C) [10] or to other mechanisms [11,12].

Although GLP-1 has been shown to promote beta-cell replication and mass in animal models of prediabetes and diabetes, these findings have not been replicated in humans [13-16].

GLP-1 exhibits a short half-life of one to two minutes due to N-terminal degradation by the enzyme dipeptidyl peptidase 4 (DPP-4). Synthetic GLP-1 receptor agonists are variably resistant to degradation by the enzyme DPP-4, and therefore have a longer half-life, facilitating clinical use. Longer-acting GLP-1 receptor agonists can be administered once daily or once weekly. Like native GLP-1, all synthetic GLP-1 receptor agonists bind to the GLP-1 receptor and stimulate glucose-dependent insulin release from the pancreatic islets as their primary glucose-lowering effect. (See 'Administration' below and 'Glycemic efficacy' below.)

●GIP – GIP is produced in the K cells of the small intestine. It binds to a specific GIP receptor, which is expressed in various tissues, including pancreatic beta cells, pancreatic alpha cells, subcutaneous and visceral adipose tissue, bone, and heart. In the postprandial state, GIP is cosecreted with GLP-1, and they may interact in an additive fashion to potentiate glucose-induced insulin secretion (figure 1) [5]. However, GIP exhibits different effects from GLP-1 on glucagon secretion. In the euglycemic or hypoglycemic states, GIP enhances glucagon activity (table 1) [17,18].

A synthetic dual-acting GIP and GLP-1 receptor agonist (tirzepatide) is available for the treatment of hyperglycemia in patients with type 2 diabetes [19]. The extent to which GIP receptor activation contributes to the therapeutic effects of tirzepatide is uncertain and the subject of ongoing investigation [20-22]. Tirzepatide has a half-life of five days, allowing for once-weekly administration. (See 'Glycemic efficacy' below and 'Weight loss' below.)

SUGGESTED APPROACH TO THE USE OF GLP-1 RECEPTOR AGONIST-BASED THERAPIES

Patient selection

●Glucagon-like peptide 1 (GLP-1) receptor agonists – GLP-1 receptor agonists are particularly appropriate for use in combination with metformin (and/or another oral agent) in certain clinical settings that include the following [23,24]:

•Presence of atherosclerotic cardiovascular disease (ASCVD).

•Glycated hemoglobin (A1C) well above goal (eg, ≥1.5 percent above target).

•Primary treatment goals of body weight loss or avoidance of hypoglycemia.

•Presence of chronic kidney disease – In such patients, sodium-glucose cotransporter 2 [SGLT2] inhibitors are generally preferred, but a GLP-1 receptor agonist may be used if SGLT2 inhibitors are contraindicated or if additional glucose lowering is needed. (See "Management of hyperglycemia in patients with type 2 diabetes and advanced chronic kidney disease or end-stage kidney disease".)

In these settings, GLP-1 receptor agonists may also be used in combination with basal insulin (with or without metformin). Cost and gastrointestinal side effects may be barriers to use of GLP-1-based therapies. (See 'Administration' below and "Management of persistent hyperglycemia in type 2 diabetes mellitus", section on 'Our approach' and "Initial management of hyperglycemia in adults with type 2 diabetes mellitus", section on 'Choice of initial therapy'.)

●Dual-acting GLP-1 and GIP receptor agonists – Data are insufficient for tirzepatide use in patients with ASCVD. Tirzepatide is an option for improving glycemia in patients with type 2 diabetes and without ASCVD, particularly when weight loss is an important consideration or if A1C is well above target. (See 'Glycemic efficacy' below and 'Weight loss' below and 'Cardiovascular effects' below.)

Contraindications and precautions — GLP-1 receptor agonist-based therapies should not be used in patients with:

●A history of pancreatitis. Postmarketing reports have noted cases of hemorrhagic and nonhemorrhagic pancreatitis, and all GLP-1 receptor agonists include a warning regarding pancreatitis. They should be stopped immediately and not restarted. (See 'Pancreas' below.)

●Type 1 diabetes. Some of the salutary effects of these agents are independent of islet cell function (eg, decreased glucagon, weight loss, cardiovascular and kidney protection) and might benefit specific individuals with type 1 diabetes [1-3,25,26]. Until further data are available, however, we do not use GLP-1-based therapies in patients with type 1 diabetes specifically for glycemic management. (See "Management of blood glucose in adults with type 1 diabetes mellitus", section on 'Adjunctive therapy not recommended'.)

In addition:

●All GLP-1-based therapies slow gastric emptying [27,28]. Exenatide (short-acting) and lixisenatide should not be used in patients with gastrointestinal disease (eg, gastroparesis). Long-acting GLP-1 receptor agonists (liraglutide, dulaglutide, exenatide once weekly, tirzepatide, and semaglutide) should be used with caution in those with gastroparesis.

●Liraglutide, dulaglutide, exenatide once weekly, semaglutide (injectable or oral), and tirzepatide should not be used in patients with a personal or family history of medullary thyroid cancer or multiple endocrine neoplasia 2A or 2B. Most experts would not prescribe any GLP-1-based therapy in this population.

●Exenatide (twice daily) should not be used in patients with creatinine clearance <30 mL/min.

●Exenatide (once-weekly formulation) should not be used in patients with estimated glomerular filtration rate (eGFR) <45 mL/min/1.73 m².

●Lixisenatide should not be used in patients with eGFR <30 mL/min/1.73 m².

●Liraglutide and dulaglutide should be used with caution in patients with kidney impairment.

Choice of therapy — When a decision has been made to use GLP-1 receptor agonist-based therapies, our selection of a particular agent is guided by the presence of underlying patient comorbidities, in particular ASCVD, as well as by glycemic efficacy.

●With clinical ASCVD – In patients with clinical ASCVD (eg, prior myocardial infarction, stroke), we suggest liraglutide, semaglutide (subcutaneous), or dulaglutide, based on the respective cardiovascular outcomes study results. (See 'Cardiovascular effects' below.)

The progression of retinopathy seen in the subcutaneous semaglutide study is likely a consequence of rapid glycemic lowering (similar to that seen in other settings) rather than a direct effect of the drug (see 'Microvascular outcomes' below). If subcutaneous semaglutide is prescribed to a patient with a history of diabetic retinopathy, consideration should be given to slower titration to avoid rapid declines in A1C and retinal screening within six months of drug initiation to detect progression of retinopathy. The caution regarding rapid lowering of glycemia and risk of retinopathy applies to all glucose-lowering medications.

●Without ASCVD – In patients without ASCVD, we prefer long- over short-acting GLP-1-based therapies due to patient convenience and greater glycemic efficacy [29]. For patients in whom weight loss is a primary consideration, subcutaneous semaglutide or tirzepatide is preferred (see 'Weight loss' below). Among the longer-acting agents (liraglutide, exenatide once weekly, dulaglutide, subcutaneous semaglutide, tirzepatide), the need for reconstitution (subcutaneous preparations), patient preference, and payer coverage are also important considerations.

No comparative trials have evaluated the effects of different GLP-1-based therapies on patient-important, long-term outcomes such as microvascular complications, health-related quality of life, or mortality. A number of comparative trials have included glycemia as the primary outcome, and some have included weight loss as a secondary outcome [29-36].

●Shorter acting versus longer acting – In trials comparing exenatide administered twice daily with exenatide once weekly, liraglutide once daily, or dulaglutide once weekly, the reduction in A1C with the longer-acting (daily or weekly) GLP-1 receptor agonists was significantly greater (treatment difference -0.3 to -0.7 percent) [30-32,37,38].

●Longer acting – Among the longer-acting GLP-1-based therapies, small differences in glycemic efficacy favor tirzepatide over subcutaneous semaglutide (1 mg) [39], liraglutide or subcutaneous semaglutide over exenatide once weekly [33,40], and subcutaneous semaglutide over dulaglutide [41] or liraglutide [42]. Glycemic management appears to be similar with liraglutide and dulaglutide [43] and with oral semaglutide and liraglutide [36]. (See 'Glycemic efficacy' below.)

In these trials, weight loss was generally better with subcutaneous semaglutide (-6 kg) than once-weekly exenatide (-2 kg), dulaglutide (-3 kg), and 1.2 mg liraglutide (-2 kg), as well as with 1.8 mg liraglutide (-3.5 kg) compared with once-weekly exenatide (-2.5 kg) and dulaglutide (-3 kg) [29,33,40-43]. Tirzepatide resulted in greater weight loss than subcutaneous semaglutide (1 mg) [39]. (See 'Weight loss' below.)

Pretreatment evaluation — Prior to initiation of GLP-1-based therapy, we perform the following assessments:

●Clinical history – We assess all patients for a personal or family history of medullary thyroid cancer or multiple endocrine neoplasia types 2A or 2B. We also ask about a prior diagnosis of gastroparesis or symptoms that suggest this condition. The diagnostic evaluation for suspected gastroparesis is reviewed separately. (See "Gastroparesis: Etiology, clinical manifestations, and diagnosis", section on 'Evaluation'.)

●Physical examination – We perform a physical examination including assessment of thyroid size and the presence of palpable nodules. We also evaluate for other stigmata of multiple endocrine neoplasia (eg, mucosal neuroma). (See "Clinical manifestations and diagnosis of multiple endocrine neoplasia type 2", section on 'Clinical features'.)

●Retinal examination – We obtain a baseline retinal examination if not performed within the prior 12 months, particularly before initiating subcutaneous semaglutide or tirzepatide or in the presence of symptoms suggesting new or worsening retinopathy.

●Serum creatinine – We measure serum creatinine to calculate eGFR if a recent value is not available or any suspicion exists for worsening or advanced kidney disease.

Administration — Most GLP-1 receptor agonists are initiated at a low dose and then slowly advanced (table 2) to avoid adverse gastrointestinal side effects, which are relatively common, usually affecting from 15 to 45 percent of patients. Gastrointestinal side effects may be attenuated somewhat with longer-acting agents, although high-quality comparative studies have not been performed. There may also be individual variation in gastrointestinal tolerance among the long-acting agents, although there is limited experience with switching from one long-acting agent to another. (See 'Gastrointestinal' below.)

●Combination with oral agents – GLP-1 receptor agonists can be combined with metformin and most other oral agents. They should not be combined with DPP-4 inhibitors, as there do not appear to be additive effects on glucose lowering [44]. There are few trials directly evaluating the combination of GLP-1 receptor agonists with SGLT2 inhibitors, and the published trials are generally short-term with A1C as the primary outcome [45,46]. In some of the GLP-1 receptor agonist cardiovascular outcomes trials, a small proportion of the participants were taking SGLT2 inhibitors at baseline (eg, 15 percent), and the point estimate for ASCVD benefit was not different compared with those not taking SGLT2 inhibitors [47]. Some guidelines suggest combining SGLT2 inhibitors and GLP-1-based therapies [24]. Primary trial evidence is lacking to support additive benefits of these agents for cardiovascular or kidney protection. In individuals with ASCVD or kidney disease who are not meeting glycemic goals with an agent from either class, combination therapy may be considered using a shared decision-making approach [48,49]. (See "Management of persistent hyperglycemia in type 2 diabetes mellitus", section on 'Dual agent failure'.)

●Combination with insulin – GLP-1 receptor agonists may be combined with insulin. When used in combination with basal insulin, patients using GLP-1 receptor agonists compared with placebo achieved glycemic targets at reduced insulin doses and less hypoglycemia or weight gain but more gastrointestinal side effects [50-52]. GLP-1 receptor agonists are available in combination with long-acting insulin. Limited data support the use of GLP-1 receptor agonists in combination with prandial insulin [53,54].

●Risk of hypoglycemia – The risk of hypoglycemia is small when a GLP-1 receptor agonist is used in combination with metformin [55]. Hypoglycemic events may occur, however, when GLP-1 receptor agonists are given in conjunction with diabetes medications known to cause hypoglycemia (eg, basal insulin, sulfonylureas, meglitinides). For the majority of patients in whom the addition of GLP-1 receptor agonists is prompted by poor glycemic control, a reduction in the dose of basal insulin, sulfonylureas, and meglitinides is not typically necessary, although all patients should be informed of the possibility of hypoglycemia.

●Use in chronic kidney disease – Experience is limited with most GLP-1 receptor agonists in patients with severe kidney impairment (eGFR 15 to 29 mL/min/1.73 m²) [53,56,57]. (See "Management of hyperglycemia in patients with type 2 diabetes and advanced chronic kidney disease or end-stage kidney disease".)

•Long-acting agents – In liraglutide, dulaglutide, and semaglutide trials, the presence of mild to moderate or moderate to severe kidney impairment did not affect treatment outcomes [53,58-65]. These agents are not excreted by the kidneys, and dose reductions with impaired kidney function are not necessary [57,66,67]. They may be used in chronic kidney disease stage 4, but monitoring kidney function and providing patient education to discontinue with any signs and symptoms of dehydration due to nausea or satiety is warranted to reduce the risk of acute kidney injury (AKI).

•Short-acting agents – In the lixisenatide trial, the presence of mild (eGFR 60 to 89 mL/min/1.73 m²) or moderate (eGFR 30 to 59 mL/min/1.73 m²) kidney impairment did not affect treatment outcomes [68]. There are few data in patients with eGFR 15 to 29 mL/min/1.73 m². Lixisenatide is presumed to be eliminated by the kidneys, and exposure is increased in these patients [69]. If used in this setting, monitor closely for gastrointestinal adverse effects, which may increase risk of AKI. The single ingredient lixisenatide injection is no longer available in the United States or Canada but may be available in a few other areas.

Exenatide once weekly should not be used in patients with eGFR <45 mL/min/1.73 m². Although some data show similar safety and efficacy in patients with an eGFR 30 to <60 mL/min/1.73 m² compared with those with an eGFR ≥60 mL/min/1.73 m² [70,71], we prefer to use liraglutide, dulaglutide, or semaglutide when eGFR is between 30 and 45 mL/min/1.73 m².

Exenatide twice daily should not be used in patients with creatinine clearance <30 mL/min. For patients with moderate kidney impairment (creatinine clearance 30 to 50 mL/min), monitoring of serum creatinine is warranted when initiating therapy and after the usual dose increase from 5 to 10 mcg [72]. (See 'Kidney' below.)

Monitoring — Glycemic indices (A1C, fasting blood glucose) and kidney function are routinely monitored in all patients with type 2 diabetes. (See "Overview of general medical care in nonpregnant adults with diabetes mellitus", section on 'Glycemic management' and "Overview of general medical care in nonpregnant adults with diabetes mellitus", section on 'Diabetes-related complications'.)

●Glycemic indices – A1C is generally measured every three to six months.

●Kidney function – In individuals with known CKD or those who experience gastrointestinal symptoms that increase risk for dehydration (eg, vomiting, diarrhea), serum creatinine should be monitored within four weeks of initiating therapy and two to three months after increasing the dose. Serum creatinine is typically measured at least annually in most patients with type 2 diabetes.

●Retinal examination – For patients with a history of diabetic retinopathy prescribed subcutaneous semaglutide, slowly titrate the dose (to avoid rapid declines in A1C) and perform retinal screening within six months to detect progression of retinopathy. (See 'Microvascular outcomes' below.)

●Hypersensitivity reactions – Hypersensitivity reactions are uncommon. However, we generally use an alternative, non-GLP-1 receptor agonist glucose-lowering agent in a person with a history of a hypersensitivity reaction to any GLP-1 receptor agonist. (See 'Hypersensitivity reactions' below.)

CLINICAL OUTCOMES

Glycemic efficacy

●GLP-1 receptor agonists – Short-acting glucagon-like peptide 1 (GLP-1) receptor agonists (exenatide twice daily and lixisenatide) provide short-lived GLP-1 receptor activation, whereas long-acting agents (liraglutide, exenatide once weekly, dulaglutide, semaglutide) activate the GLP-1 receptor continuously at their recommended dose. Compared with longer-acting GLP-1 receptor agonists, the shorter-acting agents tend to have a more pronounced effect on postprandial hyperglycemia and gastric emptying and less effect on fasting glucose [73,74].

All GLP-1 receptor agonists are very effective in reducing A1C, as illustrated by the following meta-analyses:

•In a meta-analysis of 34 randomized trials comparing GLP-1 receptor agonists (exenatide, liraglutide, albiglutide, taspoglutide, lixisenatide, dulaglutide) with placebo or another GLP-1 receptor agonist, in patients with type 2 diabetes and suboptimal control on oral agents (typically metformin), all GLP-1 receptor agonists reduced A1C compared with placebo (range -0.55 to -1.38 percentage points) [38,75]. Longer-acting GLP-1 receptor agonists reduced A1C more than shorter-acting ones, but with considerable drug-specific differences in head-to-head studies. (See 'Choice of therapy' above.)

•In meta-analyses of the usually short-term (26-week), pharmaceutical company-supported studies, GLP-1 receptor agonist therapy in patients with baseline A1C levels of 8 to 8.5 percent lowered A1C more (by 0.2 to 0.8 percentage points) than the active comparators (eg, sitagliptin, pioglitazone, daily exenatide, basal insulin glargine) [75,76]. The titration algorithm for basal glargine was consistent with the standard clinical approach (eg, initial dose 10 units with weekly titration to fasting glucose goal of 72 to 99 mg/dL).

•In a meta-analysis of trials comparing the glycemic efficacy of a GLP-1 receptor agonist with basal insulin, the reductions in A1C with liraglutide or exenatide twice daily did not differ from A1C reduction with basal insulin [77]. Exenatide once weekly and dulaglutide reduced A1C modestly more (approximately 0.3 percentage points) than basal insulin, and injectable semaglutide reduced A1C by 0.8 percentage points more than glargine [78]. However, the comparison with insulin therapy is particularly problematic as the intensity of insulin titration in the comparison groups was not rigorously enforced.

In a subsequently published comparative effectiveness trial (GRADE) with a mean follow-up of five years in 5047 patients with type 2 diabetes on metformin monotherapy, the cumulative incidence of A1C ≥7 percent was lower for patients randomly assigned to liraglutide (68 percent) or glargine (67 percent) as add-on treatment than for those who received glimepiride (72 percent) or sitagliptin (77 percent) [55]. (See "Management of persistent hyperglycemia in type 2 diabetes mellitus", section on 'Our approach'.)

●Dual-acting GLP-1 and GIP receptor agonists – Tirzepatide is a dual glucose-dependent insulinotropic polypeptide (GIP) and GLP-1 receptor agonist. It appears to have remarkable glycemic (and weight-reducing) efficacy compared with either agent alone [79]. It has been studied for use as monotherapy in patients inadequately treated with diet and exercise [80], as well as in combination with other agents, including metformin, sulfonylureas, and insulin [39,81-83]. As examples,

•In a 40-week trial comparing tirzepatide with semaglutide (each administered once weekly by subcutaneous injection) in 1878 patients with type 2 diabetes who were not reaching glycemic goals with metformin monotherapy, the reduction in A1C was superior with tirzepatide (-2 to -2.3 percentage points versus -1.86 percentage points with semaglutide) [39]. In a prespecified subgroup analysis of patients with A1C >8.5 percent, the reductions in A1C were -3.22 versus -2.68 percentage points, respectively.

•In a 52-week trial comparing once-weekly subcutaneous tirzepatide with daily subcutaneous insulin glargine in 1995 people with type 2 diabetes (mean A1C 8.52 percent), body mass index (BMI) ≥25 kg/m², and high cardiovascular risk, the mean reduction in A1C with tirzepatide 10 and 15 mg was greater than with glargine (-2.43 and -2.58 percentage points, respectively, versus -1.44 percentage points with glargine [mean difference for 10 mg dose, -0.99 percentage points, 97.5% CI -1.13 to -0.86]) [81]. The majority of patients were treated with metformin (95 percent), whereas sulfonylureas were used in 54 percent and sodium-glucose cotransporter 2 (SGLT2) inhibitors in 25 percent.

The proportion of patients with hypoglycemia (glucose <54 mg/dL) was lower with tirzepatide (6 to 9 versus 16 percent with glargine).

•In a 40-week trial comparing tirzepatide with insulin glargine in 917 individuals with type 2 diabetes (mean A1C 8.7 percent, mean BMI 27.9 kg/m²), the mean reduction in A1C was greater with tirzepatide (-2.24, -2.44, and -2.49 percentage points with 5, 10, and 15 mg of tirzepatide, respectively, versus -0.95 percentage points with glargine) [84]. Treatment was added to background therapy with metformin with or without a sulfonylurea. No severe hypoglycemic events occurred during the study.

•In a 52-week trial comparing once-weekly subcutaneous tirzepatide with three-times daily prandial insulin lispro in 1428 individuals with type 2 diabetes (mean A1C 8.8 percent), the mean reduction in A1C was greater with tirzepatide (-2.1 percentage points with 5, 10, or 15 mg of tirzepatide [pooled cohort] versus -1.1 percentage points with lispro) [83]. Treatment was added to background therapy with basal insulin, with or without up to two oral glucose-lowering medications. A higher percentage of participants in the tirzepatide group achieved an A1C <7 percent compared with those in the lispro group (68 versus 36 percent, respectively). Fewer episodes of severe hypoglycemia occurred with tirzepatide (17 events) than with lispro (89 events).

●Triple-acting GLP-1, GIP, and glucagon receptor agonists (in development, not commercially available) – Retatrutide is a triple GLP-1, GIP, and glucagon receptor agonist with robust glucose-lowering efficacy in early trials. In a trial in 281 adults with type 2 diabetes (mean A1C 8.3 percent), participants were randomly assigned to once-weekly injections of retatrutide (maintenance dose of 0.5, 4, 8, or 12 mg; with or without escalation to doses >0.5 mg), placebo, or dulaglutide (1.5 mg) [85]. After 24 weeks, retatrutide 8 or 12 mg weekly led to greater mean reduction in A1C than placebo or dulaglutide (-1.88 to -2.02 percentage points with retatrutide versus -0.01 percentage points with placebo and -1.41 percentage points with dulaglutide). Reductions in A1C were sustained through 36 weeks of retatrutide treatment.

●Small molecule oral GLP-1 receptor agonists (in development, not commercially available) – Unlike oral semaglutide, nonpeptide oral GLP-1 receptor agonists (danuglipron, orforglipron) do not have to be taken in a fasting state. Their glucose-lowering efficacy has been evaluated in short-term trials [86,87]. For example, in a trial in 383 adults with type 2 diabetes (mean A1C 8.1 percent), participants were randomly assigned to orforglipron (3, 12, 24, 36, or 45 mg daily), placebo, or once-weekly dulaglutide (1.5 mg) [87]. After 26 weeks, orforglipron doses ≥12 mg daily led to greater mean reduction in A1C than placebo or dulaglutide (-1.79 to -2.1 percentage points with orforglipron versus -0.43 percentage points with placebo and -1.1 percentage points with dulaglutide).

Weight loss — Weight loss is common with GLP-1 receptor agonist-based therapies [75,76,88-90]. Weight loss may be due, in part, to the effects of GLP-1 on slowed gastric emptying and their well-recognized side effects of nausea and vomiting. However, slowed gastric emptying is attenuated over time, at least in longer-acting GLP-1 receptor agonists, and these agents are known to increase satiety through effects on the appetite centers in the brain [30,91,92]. (See 'Gastrointestinal peptides' above.)

●GLP-1 receptor agonists – In a meta-analysis of 34 trials comparing GLP-1 receptor agonists (albiglutide, dulaglutide, exenatide, liraglutide, lixisenatide, and taspoglutide) with placebo or another GLP-1 receptor agonist in patients with type 2 diabetes and suboptimal control on oral agents (typically metformin), all approved GLP-1 receptor agonists reduced weight compared with placebo with little difference between individual agents [38].

In the comparative effectiveness trial (GRADE, mean follow-up of five years in patients with type 2 diabetes on metformin monotherapy), the incidence of body weight gain ≥10 percent was lower for patients randomly assigned to liraglutide (6.1 percent) as add-on therapy than for those who received glargine (13.1 percent), glimepiride (12.1 percent), or sitagliptin (9.1 percent) [55]. Mean body weight loss was greater in the liraglutide group (3.5 kg) than in the sitagliptin (2.0 kg), glimepiride (0.73 kg), or glargine (0.61 kg) groups.

In trials designed specifically to evaluate weight loss in patients with type 2 diabetes, liraglutide and semaglutide reduced weight compared with placebo [89,90,93]. As examples:

•In a 56-week trial, comparing once-daily subcutaneous liraglutide (3 or 1.8 mg) with placebo in 846 patients with type 2 diabetes (mean A1C 7.9 percent) and obesity (mean weight 106 kg), significant weight loss occurred in the liraglutide groups (-6.4 kg [-6 percent] and -5 kg [-4.7 percent] compared with -2.2 kg [-2 percent] in the placebo group; mean difference liraglutide 3 mg compared with placebo -4 percent, 95% CI -5.1 to -2.9) [89].

•In a 68-week trial comparing once-weekly subcutaneous semaglutide (2.4 [investigational dose] or 1 mg [standard dose]) with placebo in 1210 patients with type 2 diabetes (mean A1C 8.1 percent) and obesity (mean weight 99.8 kg), significant weight loss occurred in the semaglutide groups (-9.7 kg [-9.6 percent] and -6.9 kg [-7 percent]) compared with placebo (-3.5 kg [-3.4 percent], mean difference semaglutide 2.4 mg compared with placebo -6.21 percent, 95% CI -7.28 to -5.15) [93].

In both trials, treatment with the GLP-1 receptor agonist was associated with better glycemic control, a reduction in the use of oral hypoglycemic agents, and a reduction in systolic blood pressure. The side effects were similar to those found in previous studies of GLP-1 receptor agonist therapy in diabetes with a three- to sixfold increase in gastrointestinal side effects. (See 'Gastrointestinal' below.)

The role of GLP-1 as a weight loss agent in patients without diabetes is reviewed separately. (See "Obesity in adults: Drug therapy".)

●Dual-acting GLP-1 and GIP receptor agonist – Dual-acting therapy appears to result in greater weight reduction than GLP-1 receptor agonists [94].

•In a 40-week trial comparing tirzepatide with semaglutide (each administered once weekly by subcutaneous injection), described above, the mean reduction in body weight was greater with tirzepatide (-7.6 kg, -9.3 kg, and -11.2 kg for 5, 10, and 15 mg of tirzepatide, respectively, versus -5.7 kg with semaglutide) [39].

•In a 52-week trial comparing tirzepatide with insulin glargine, described above, patients in the tirzepatide groups lost weight (7 to 11.7 kg), whereas weight increased slightly in the glargine group (1.9 kg) [81].

•In a 40-week trial comparing tirzepatide with insulin glargine, described above, tirzepatide led to body weight loss (mean reduction -5, -7, and -7.2 kg for 5, 10, and 15 mg of tirzepatide, respectively), whereas body weight increased with insulin glargine (1.5 kg) [84].

●Triple-acting GLP-1, GIP, and glucagon receptor agonists (in development, not commercially available) – Triple-acting therapy also leads to substantial body weight reduction. In the trial that compared retatrutide with placebo and dulaglutide, described above, retatrutide 12 mg weekly led to greatest mean reduction in body weight over 36 weeks of treatment (-17.2 versus -3.3 kg with placebo and -2 kg with dulaglutide) [85].

●Small molecule oral GLP-1 receptor agonists (in development, not commercially available) – Like other GLP-1 receptor agonists, nonpeptide oral agents lead to body weight loss [86,87]. As an example, in the 26-week trial that compared orforglipron with placebo and dulaglutide, described above, orforglipron ≥12 mg daily conferred greater mean reduction in body weight than placebo or dulaglutide (-6.5 to -10.1 kg with orforglipron versus -2.2 kg with placebo and -3.9 kg with dulaglutide) [87].

Cardiovascular effects — The cardiovascular studies to date (with the possible exception of dulaglutide studies) primarily have been carried out in very high-risk populations to increase the hazard rate for major cardiovascular disease (CVD) events and complete the studies in a relatively brief period of time. Therefore, there are few data on cardiovascular safety or putative benefits in lower-risk patients. Of note, the comparative effectiveness GRADE study was carried out in a cohort with generally low CVD risk [95].

●Atherosclerotic CVD (ASCVD) – In patients with type 2 diabetes and CVD, there was a reduction in ASCVD outcomes with the following GLP-1 receptor agonists when compared with placebo (table 2):

•Liraglutide [60]

•Semaglutide once weekly [59]

•Dulaglutide [64]

•Albiglutide (withdrawn from the market for commercial reasons) [96]

•Efpeglenatide (investigational) [47]

Lixisenatide, once-weekly exenatide, and oral semaglutide did not increase or decrease CVD outcomes [70,97]. Differences in CVD outcomes in studies conducted thus far may be related to intrinsic properties of available agents (such as pharmacokinetics and glucose-lowering efficacy) or may be related to differences in patient selection and study design [98,99].

●Stroke – In a meta-analysis of trials comparing a GLP-1 receptor agonist (lixisenatide, once-weekly exenatide, albiglutide, liraglutide, semaglutide) with placebo in people with diabetes and established CVD, GLP-1 receptor agonists reduced the risk of fatal or nonfatal stroke (26 versus 29 per 1000 persons; odds ratio [OR] 0.87, 95% CI 0.77-0.98) [100]. A subsequent meta-analysis found that GLP-1 receptor agonist use reduced risk of ischemic but not hemorrhagic stroke compared with placebo or active comparator (eg, insulin glargine, glimepiride, sitagliptin, or a sodium-glucose cotransporter 2 [SGLT2] inhibitor) [101].

In a network meta-analysis of 209 trials of drug therapies for type 2 diabetes, only GLP-1 receptor agonists reduced the risk of nonfatal stroke [102]. Individual trial data also support a protective effect of pioglitazone for stroke reduction, particularly for decreasing risk of recurrent stroke (see "Thiazolidinediones in the treatment of type 2 diabetes mellitus", section on 'Atherosclerotic cardiovascular events').

●Heart failure – Liraglutide had no effect on heart failure outcomes in patients with diabetes and established heart failure [103]. In a meta-analysis of trials comparing a GLP-1 receptor agonist (lixisenatide, once-weekly exenatide, albiglutide, liraglutide, semaglutide) with placebo in people with diabetes and established CVD, GLP-1 receptor agonists did not reduce the risk of hospitalization for heart failure (38 versus 40 per 1000 persons; OR 0.95, 95% CI 0.85-1.06) [100]. In a subsequent meta-analysis of trials comparing a GLP-1 receptor agonist with placebo in patients with type 2 diabetes and heart failure, GLP-1 receptor agonists similarly did not reduce hospitalization for heart failure, nor did they improve left ventricular ejection fraction [104]. However, compared with placebo, GLP-1 receptor agonists led to greater increase in the six-minute walk test distance.

●Cardiovascular mortality – GLP-1 receptor agonists reduce cardiovascular mortality in individuals with type 2 diabetes and existing CVD or high CVD risk [102]. In a meta-analysis of trials comparing a GLP-1 receptor agonist (lixisenatide, once-weekly exenatide, albiglutide, liraglutide, semaglutide) with placebo in people with diabetes and established CVD, GLP-1 receptor agonists reduced the risk of cardiovascular mortality (39 versus 44 events per 1000 persons; OR 0.87, 95% CI 0.79-0.95) [100].

The trials are reviewed below:

●Liraglutide – In the liraglutide trial (LEADER), 9340 patients with type 2 diabetes (mean A1C 8.7 percent) and at least one coexisting cardiovascular condition (approximately 80 percent had prior myocardial infarction, stroke, or kidney failure) if ≥50 years, or at least one cardiovascular risk factor (eg, hypertension, microalbuminuria) if ≥60 years, were randomly assigned to subcutaneous liraglutide or placebo [60]. Most patients were on combination therapy, taking either metformin (76 percent), sulfonylureas (50 percent), and/or insulin (44 percent).

After a median follow-up of 3.8 years, the primary endpoint (time to first occurrence of a composite endpoint [death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke]) occurred in fewer patients in the liraglutide group (13 versus 14.9 percent, hazard ratio [HR] 0.87, 95% CI 0.78-0.97). There were fewer add-on therapies for diabetes medications, lipid-lowering medications, and diuretics in patients in the liraglutide group than in those in the placebo group.

In a separate trial of liraglutide versus placebo in 300 patients (59 percent with type 2 diabetes) with established heart failure and reduced left ventricular ejection fraction who were recently hospitalized, liraglutide had no significant effect on the composite outcome (time to death, time to rehospitalization for heart failure, and time-averaged proportional change in N-terminal pro-B-type natriuretic peptide level) [103]. In a prespecified subgroup analysis, there was no effect of liraglutide compared with placebo on heart failure outcomes in the subset of patients with diabetes.

In the GRADE trial (patients with type 2 diabetes and low baseline prevalence of CVD), the incidence of any CVD (composite of major adverse cardiovascular events [MACE], hospitalization for heart failure or unstable angina, or any arterial revascularization) over a mean five-year follow-up was numerically lower for patients randomly assigned to liraglutide as add-on treatment to metformin (6.6 percent) than for patients assigned to glargine (9 percent), glimepiride (9.2 percent), or sitagliptin (9.6 percent) [95]. The rate of any CVD was lower for liraglutide than for all other treatments combined (HR 0.7, 95% CI 0.6-0.9). However, the rates of the individual outcomes of MACE, hospitalization for heart failure, and both cardiovascular and all-cause mortality were not significantly different between the liraglutide group and the other three treatment groups.

●Semaglutide – The subcutaneous preparation of semaglutide has been shown to reduce major adverse cardiovascular outcomes (driven by a reduction in nonfatal stroke). The small reduction in the occurrence of major adverse cardiovascular outcomes with oral semaglutide did not reach statistical significance, though a significant reduction in cardiovascular mortality (an individual component of the composite outcome) was seen.

•Injectable – In the subcutaneously administered semaglutide trial, 3297 patients with type 2 diabetes (mean A1C 8.7 percent) and established CVD, heart failure, or chronic kidney disease if ≥50 years of age (83 percent), or at least one cardiovascular risk factor if age ≥60 years (17 percent), were randomly assigned to semaglutide (0.5 or 1 mg subcutaneously once weekly) or placebo [59]. Most patients were taking combination therapy with either metformin (73 percent), insulin (58 percent), and/or sulfonylureas (43 percent). Cardiovascular medications included antihypertensives (93 percent), lipid-lowering drugs (76 percent), and antithrombotics (76 percent), and they were prescribed evenly to both groups.

After a median follow-up of two years, the primary endpoint (a composite of first occurrence of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke) occurred in fewer patients in the semaglutide group (6.6 versus 8.9 percent in the placebo group; HR 0.74, 95% CI 0.58-0.95). Among the individual components of the composite outcome, the occurrence of nonfatal stroke was significantly lower in the semaglutide group (1.6 versus 2.7 percent), whereas the reduction in nonfatal myocardial infarction (2.9 versus 3.9 percent) was not significantly different and the risk of cardiovascular death (2.7 versus 2.8 percent) was similar.

Diabetic retinopathy complications occurred more frequently in the semaglutide group. (See 'Microvascular outcomes' below.)

•Oral – In the orally administered semaglutide trial, 3183 patients with type 2 diabetes (mean A1C 8.2 percent) with established CVD or chronic kidney disease if ≥50 years of age (85 percent), or with at least one cardiovascular risk factor if ≥60 years (15 percent), were randomly assigned to once-daily oral semaglutide (target dose, 14 mg) or placebo, in addition to their other diabetes medications (predominantly metformin or insulin) [105].

After a median follow-up of 15.9 months, the primary endpoint (a composite of the first occurrence of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke) was not significantly different between treatment groups (3.8 versus 4.8 percent, HR 0.79, 95% CI 0.57-1.11). Among the individual components of the composite outcome, the occurrence of death from cardiovascular causes was lower in the oral semaglutide group (0.9 versus 1.9 percent, HR 0.49, 95% CI 0.27-0.92), whereas the difference in nonfatal myocardial infarction (2.3 versus 1.9 percent) and nonfatal stroke (0.8 versus 1 percent) were not statistically significant.

No reported increase in retinopathy was observed in patients receiving oral semaglutide (7.1 versus 6.3 percent in the placebo group); however, patients with proliferative retinopathy or actively treated macular edema were excluded from study participation. (See 'Microvascular outcomes' below.)

●Dulaglutide – In the dulaglutide trial (REWIND), 9901 patients with type 2 diabetes (mean A1C 7.2 percent) ≥50 years with established CVD (31.5 percent) or CVD risk factors were randomly assigned to either weekly subcutaneous dulaglutide (1.5 mg) or placebo [64]. Most patients were taking combination therapy, either metformin (81 percent), sulfonylureas (46 percent), and/or insulin (24 percent). After a median follow-up of 5.4 years, the primary endpoint (time to first occurrence of a composite endpoint [death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke]) occurred in fewer patients in the dulaglutide group (12 versus 13.4 percent, HR 0.88, 95% CI 0.79-0.99). Among the individual components of the composite outcome, the occurrence of nonfatal stroke was significantly lower in the dulaglutide group.

●Lixisenatide – In the lixisenatide trial, 6068 patients with type 2 diabetes and either a myocardial infarction or hospitalization for unstable angina in the past 180 days were randomly assigned to receive subcutaneous lixisenatide or placebo in addition to other diabetes medications (predominantly metformin, insulin, and sulfonylureas) [97]. After a median follow-up of 25 months, the primary endpoint (a composite endpoint of cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, or hospitalization for unstable angina) occurred in a similar proportion of patients (13.4 and 13.2 percent in the lixisenatide and placebo groups, respectively; HR 1.02, 95% CI 0.89-1.17). There was no significant difference in any of the individual components of the composite endpoint. There was no significant difference in the rate of hospitalization for heart failure (approximately 4 percent in each group).

●Exenatide once weekly – In a noninferiority trial, 14,752 patients with type 2 diabetes (73.1 percent had previous CVD) were randomly assigned to receive subcutaneous exenatide or placebo once weekly [70]. After a median follow-up of 3.2 years, the primary endpoint (a composite of first occurrence of death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke) was not significantly different between treatment groups (11.4 versus 12.2 percent with placebo, HR 0.91, 95% CI 0.83-1.0). There was no significant difference in the rate of hospitalization for heart failure (approximately 3 percent in each group). An important limitation of the trial was a high rate of discontinuation of the treatment regimen (approximately 40 percent in each group).

●Efpeglenatide – In the efpeglenatide trial (once-weekly subcutaneous injection), 4076 patients with type 2 diabetes and either CVD or chronic kidney disease (plus at least one other cardiovascular risk factor) were randomly assigned to receive weekly subcutaneous efpeglenatide or placebo [47]. After a median follow-up of 1.8 years, the primary endpoint (a composite of nonfatal myocardial infarction, nonfatal stroke, or death from cardiovascular or unknown cause) occurred in fewer patients in the efpeglenatide group (7 versus 9.2 percent, HR 0.73, 95% CI 0.58-0.92). There were no significant differences in any of the individual components of the composite endpoint.

●Tirzepatide – Cardiovascular outcomes have only been measured as part of a safety assessment. Tirzepatide does not increase the risk of major cardiovascular events [81,106]. As an example, in the trial described above comparing tirzepatide with insulin glargine in patients at high cardiovascular risk (see 'Glycemic efficacy' above), the composite cardiovascular endpoint (cardiovascular death, myocardial infarction, stroke, hospitalization for unstable angina) occurred in a similar proportion of patients in the two treatment groups (5 to 6 percent) [81]. In a meta-analysis of seven phase II and III trials (7215 participants at low, medium, or high cardiovascular risk) comparing tirzepatide with placebo or an active comparator, there was no increase in the composite cardiovascular endpoint with tirzepatide (HR 0.80, 95% CI 0.57-1.11) [106]. Trials specifically designed to evaluate cardiovascular benefit are ongoing [107].

Microvascular outcomes — There are no trials evaluating microvascular disease as the primary outcome in patients taking GLP-1 receptor agonists [108]. In trials designed to assess cardiovascular outcomes in patients with or at high risk for CVD, liraglutide, semaglutide, dulaglutide, and efpeglenatide (investigational) reduced nephropathy outcomes, whereas there was an increase in retinopathy outcomes with injectable semaglutide (table 2). In a trial designed to assess glycemic control in patients with moderate to severe chronic kidney disease, dulaglutide attenuated progression of kidney disease. The trials are reviewed below:

●Liraglutide – In the LEADER trial described above (9340 patients with type 2 diabetes and at least one coexisting cardiovascular condition, median follow-up of 3.8 years) [60], the secondary endpoint (a composite of new-onset persistent macroalbuminuria, persistent doubling of the serum creatinine level, end-stage kidney disease, or death due to kidney disease) occurred in fewer patients taking liraglutide (5.7 versus 7.2 percent with placebo, HR 0.78, 95% CI 0.67-0.92) [58]. The results were driven by a lower incidence of new-onset, persistent macroalbuminuria. There was no significant effect on the incidence of the other three components of the composite outcome.

In the GRADE trial (5047 patients with type 2 diabetes and low baseline prevalence of CVD or kidney disease, median follow-up five years), patients randomly assigned to liraglutide had similar rates of moderately or severely increased albuminuria and impairment of kidney function (eGFR <60 mL/min/1.73 m²) as those assigned to glargine, glimepiride, or sitagliptin [95]. The rate of peripheral neuropathy also was similar across groups.

There were few retinal outcomes based on participant self-report, defined as the need for laser therapy or intravitreal injections or the development of blindness, in this trial.

●Semaglutide – In the subcutaneously administered semaglutide trial described above (3297 patients with established CVD, heart failure, or chronic kidney disease or age ≥60 years with at least one cardiovascular risk factor, median follow-up two years), diabetic retinopathy complications occurred more frequently in the semaglutide group (3 versus 1.8 percent in the placebo group, HR 1.76, 95% CI 1.11-2.78), particularly among patients with existing retinopathy [59]. The higher rate of retinopathy complications was unexpected and may be a consequence of rapid glycemic control similar to that seen in other settings [109]. New or worsening nephropathy occurred less frequently (3.8 versus 6.1 percent) and was driven by a lower incidence of persistent macroalbuminuria.

No reported increase in retinopathy was observed in patients receiving oral semaglutide (7.1 versus 6.3 percent in the placebo group); however, patients with proliferative retinopathy or actively treated macular edema were excluded from study participation [105].

●Lixisenatide – In the lixisenatide trial (6068 patients with type 2 diabetes and either a myocardial infarction or hospitalization for unstable angina in the past 180 days, median follow-up 25 months), changes in the urinary albumin-to-creatinine ratio were evaluated [97]. Although the percentage change in the ratio was modestly better with lixisenatide than placebo, the median values at baseline and follow-up were similar in the two groups.

●Dulaglutide – In the dulaglutide trial (REWIND, 9901 patients with diabetes and CVD or risk for CVD, median follow-up 5.4 years), there was a reduction in the composite clinical microvascular outcome (first occurrence of either a retinal [photocoagulation, anti-vascular endothelial growth factor therapy, or vitrectomy] or kidney [development of urinary albumin-to-creatinine ratio >33.9 mg/mmol, sustained ≥30 percent decline in estimated glomerular filtration rate (eGFR), or chronic kidney replacement therapy] outcome) in the dulaglutide group (18.4 versus 20.6 percent, HR 0.87, 95% CI 0.79-0.95), primarily driven by significantly fewer composite kidney outcomes [64]. In a subsequent exploratory analysis of the secondary kidney outcomes, there was a significant reduction in the development of new macroalbuminuria (8.9 versus 11.3 percent, HR 0.77, 95% CI 0.68-0.87) [110].

In a 52-week, open-label trial of weekly dulaglutide (1.5 or 0.75 mg) or daily insulin glargine, both in combination with prandial insulin lispro, in 577 patients with type 2 diabetes and moderate to severe chronic kidney disease (mean eGFR 38.3 mL/min/1.73 m², 30 percent had eGFR between 15 and 30 mL/min/1.73 m²), dulaglutide slowed progression of kidney disease and prevented worsening of albuminuria [53]. The reduction in A1C was similar in the dulaglutide and glargine groups.

●Efpeglenatide – In the efpeglenatide trial, 4076 patients with type 2 diabetes and either CVD or chronic kidney disease (plus at least one other cardiovascular risk factor) were randomly assigned to receive weekly subcutaneous efpeglenatide or placebo [47]. After a median follow-up of 1.8 years, the secondary endpoint (a composite of a decrease in kidney function or macroalbuminuria) occurred in fewer patients in the efpeglenatide group (13 versus 18.4 percent, HR 0.68, 95% CI 0.57-0.79).

It is important to note that these trials were not specifically designed and were of relatively short duration to assess microvascular outcomes. In addition, the presence of baseline retinopathy or neuropathy was not consistently and systematically evaluated. Trials with primary microvascular outcomes and in patients who are not at high cardiovascular risk are required in order to better understand the microvascular effects of GLP-1 receptor agonists. The mechanism of these effects also needs to be better understood as the separation in A1C was relatively small and over a relatively brief period of time to affect microvascular disease.

All-cause mortality — GLP-1 receptor agonists decrease overall mortality in people with diabetes and established CVD [102]. As an example, in a meta-analysis of seven trials comparing GLP-1 receptor agonists (lixisenatide, exenatide, albiglutide, liraglutide, semaglutide) with placebo in patients with diabetes and CVD, GLP-1 receptor agonists reduced the risk of all-cause mortality (60 versus 68 events per 1000 persons, OR 0.88, 95% CI 0.82-0.95) [100].

ADVERSE EFFECTS — The following precautions and adverse effects pertain to glucagon-like peptide 1 (GLP-1) receptor agonists, used alone or in combination with a glucose-dependent insulinotropic polypeptide (GIP) receptor agonist. The long-term safety of GLP-1 receptor agonists has not been established, as the majority of clinical trials are less than four years in duration.

Gastrointestinal — The side effects of GLP-1-based therapies are predominantly gastrointestinal, particularly nausea, vomiting, and diarrhea, which are frequent [111]. They occur consistently in trials in 10 to 50 percent of patients [76]. In a network meta-analysis of 236 clinical trials, GLP-1 receptor agonists compared with oral agents were associated with greater adverse events leading to treatment discontinuation [112].

When used for body weight reduction, GLP-1-based therapies have been associated with more severe gastrointestinal risks, including obstruction and symptomatic gastroparesis [113]. Anesthesia guidelines recommend holding these therapies prior to elective intubation for presumed risk of aspiration. (See "Anesthesia for patients with diabetes mellitus and/ or hyperglycemia", section on 'Medication regimen' and "Rapid sequence induction and intubation (RSII) for anesthesia", section on 'Patients taking GLP-1 receptor agonists'.)

Nausea is the most frequent adverse event with exenatide once weekly, but it has been reported less frequently with once-weekly than with twice-daily administration (26 versus 50 percent) and also less frequently than with liraglutide (9 versus 21 percent) [32,33]. Gastrointestinal adverse effects with exenatide once weekly may be increased in patients with an estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m² [71].

Subcutaneous and oral semaglutide are also associated with gastrointestinal side effects. In one trial, nausea, vomiting, and diarrhea occurred in 15, 9, and 12.3 percent, respectively, of patients taking semaglutide (14 mg orally daily) compared with 6.9, 4.1, and 7.9 percent, respectively, of patients taking sitagliptin (100 mg daily) [114]. In a trial comparing tirzepatide with semaglutide, gastrointestinal adverse effects were similar in the two groups (nausea 17.4 to 22.1 percent, diarrhea 11.5 to 16.4 percent, decreased appetite 5.3 to 8.9 percent) [39].

Nausea may wane with duration of therapy and can be reduced with dose titration [111,115].

Pancreas — Acute pancreatitis has been reported in association with GLP-1 receptor agonist treatment [113,116-119]. There are insufficient data to know if there is a causal relationship. Pancreatitis should be considered in patients with persistent severe abdominal pain (with or without nausea), and GLP-1 receptor agonists should be discontinued in such patients. If pancreatitis is confirmed, it should not be restarted. In addition, GLP-1 receptor agonists should not be initiated in a patient with a history of pancreatitis.

In a population-based case-control study using a large insurance database, treatment with GLP-1-based therapy (sitagliptin and exenatide) was associated with an increased risk of hospitalization for acute pancreatitis (adjusted odds ratio [OR] 2.07, 95% CI 1.36-3.13) [120]. In contrast, retrospective cohort studies [121-123] and meta-analyses of randomized trials [124-126] did not identify an increased risk. In population-based cohort studies, there was no difference in the risk of pancreatitis in patients taking GLP-1-based therapies compared with sulfonylureas (1.45 and 1.47 per 1000 patients per year, respectively) [127] or other oral agents [128]. Overall, the incidence of pancreatitis is low (16 cases among 14,562 patients enrolled in GLP-1 receptor agonist randomized trials) [124].

In some trials, GLP-1 receptor agonists increased pancreatic enzymes (amylase and lipase) from baseline levels, although often remaining within the normal range [111,129]. In one analysis, lipase and amylase levels increased above the upper limit of normal in the liraglutide and placebo groups (51 and 32 percent of participants, respectively, for lipase and 29 and 23 percent, respectively, for amylase) [130]. These elevations did not predict risk of subsequent acute pancreatitis. The diagnosis of acute pancreatitis should not be made solely on the basis of an elevation in pancreatic enzymes. (See "Clinical manifestations and diagnosis of acute pancreatitis", section on 'Diagnosis'.)

There have also been case reports of an increased risk of subclinical pancreatic inflammation, pancreatic cancer, and neuroendocrine tumors in exenatide users [118,131-133]. A causal relationship has not been established. After a review of available data, the US Food and Drug Administration (FDA) and the European Medicines Agency agreed that there was insufficient evidence to confirm an increased risk of pancreatic cancer with use of GLP-1-based therapies [134-136]. However, concerns remain [137], and monitoring for and reporting of pancreatic adverse effects will continue [134,136,138].

Gallbladder and biliary diseases — GLP-1 receptor agonist therapy has been associated with increased risk of gallbladder and biliary diseases including cholelithiasis and cholecystitis. In one meta-analysis of 76 trials, participants randomly assigned to GLP-1 receptor agonist treatment had an increased risk of the composite outcome of gallbladder or biliary diseases (event rate 1.58 versus 1.19 percent, relative risk [RR] 1.37, 95% CI 1.23-1.52) [139]. Use of GLP-1 receptor agonists specifically for weight loss, higher doses, and longer duration of treatment were all associated with greater risk. Elevated risk of acute cholecystitis with GLP-1 receptor agonist treatment has further been supported by a subsequently published postmarketing surveillance report [140].

Hypersensitivity reactions

●Angioedema/anaphylaxis – Rare cases of angioedema and anaphylaxis have been reported with GLP-1 receptor agonists, including semaglutide, liraglutide, dulaglutide, exenatide, and lixisenatide [141-145]. In a case report, a patient with hypersensitivity reactions to both exenatide and lixisenatide did not have a reaction to liraglutide [146], suggesting that immunogenicity varies among the agents. However, we generally use an alternative glucose-lowering agent in a person with a history of a hypersensitivity reaction to any GLP-1 receptor agonist.

●Injection site reactions – In studies comparing insulin administration with once-weekly GLP-1 receptor agonists, local site reactions were more common with GLP-1 receptor agonists, particularly albiglutide and exenatide once weekly (approximately 10 percent), compared with 1 to 5 percent with insulin [147,148]. In comparison trials, injection site reactions were significantly more common with exenatide once weekly compared with exenatide twice daily [30,111] and more common with exenatide once weekly [33] or albiglutide [34] than liraglutide. Reactions noted with exenatide once weekly include abscess, cellulitis, and necrosis, with or without subcutaneous nodules [149].

●Immunogenicity – Antibodies to GLP-1 receptor agonists may develop. In the majority of patients, the titer of antibodies decreases over time and does not affect glycemic control. However, some patients develop high titers of antibodies that may attenuate the glycemic response [150]. In a meta-analysis of 17 trials, the proportion of patients with antibodies against GLP-1 was higher in the albiglutide group compared with placebo (6.4 percent albiglutide 30 mg weekly versus 2 percent with placebo) [76]. In addition, up to 50 percent of patients developed low levels of anti-exenatide antibodies, with no relation to glycemic control or safety parameters.

Kidney — There have been case reports of acute kidney failure or impaired kidney function in patients using exenatide twice daily, typically in the setting of severe gastrointestinal adverse effects resulting in dehydration [111,151-153]. In a report of four patients, the time between initiation of exenatide and diagnosis of acute kidney failure ranged from two to nine months [154]. All four patients presented with nausea, vomiting, and/or decreased fluid intake, and all were receiving angiotensin-converting enzyme (ACE) inhibitors and diuretics, which can contribute to the decline in kidney function. None of the patients were taking nonsteroidal antiinflammatory drugs (NSAIDs). After a dose reduction or withdrawal of exenatide, recovery of kidney function was incomplete in three of the four patients. Kidney biopsy in one patient showed ischemic glomeruli with moderate to severe interstitial fibrosis, tubular atrophy, and early diabetic nephropathy. The relationship between these findings and exenatide could not be determined.

Acute kidney injury (AKI) after taking other GLP-1 receptor agonists has been infrequently reported [111,151,155,156]. Kidney function should be monitored in patients with severe gastrointestinal adverse effects [111,151]. (See 'Monitoring' above.)

Thrombocytopenia — In case reports, exenatide has been associated with drug-induced immune thrombocytopenia, with detection of immunoglobulin G (IgG) antibody that reacts with platelets only when exenatide is present [157]. Serious bleeding may occur. Exenatide should be discontinued immediately and should not be restarted. However, prolonged thrombocytopenia may occur after discontinuation of exenatide owing to the long half-life (median two weeks) of the sustained-release formulation [158]. A warning is included in exenatide labeling, but routine monitoring of platelet counts has not been recommended.

Other — In rodent studies, liraglutide and dulaglutide were associated with benign and malignant thyroid C cell tumors [159,160]. In addition, stimulation of calcitonin release was reported in rats and mice exposed to exenatide and liraglutide [160,161]. This effect is mediated by the GLP-1 receptor [160].

It is unclear whether any effect is present in humans because humans have far fewer C cells than rats, and expression of the GLP-1 receptor in human C cells is very low [160]. There were no changes in calcitonin levels in short-term human studies, but medullary thyroid carcinoma may take years to develop, and its low prevalence complicates any quantification of risk [160,162]. One nested case-control study found a modestly increased risk of both medullary and all thyroid cancer among individuals with type 2 diabetes prescribed a GLP-1 receptor agonist as second-line therapy [163], but this analysis did not control for key risk factors including body mass index (BMI), personal history of thyroid disease, or family history of thyroid cancer. Further, the increased risk was identified only among individuals with one to three years of GLP-1 receptor agonist use, suggesting the influence of detection bias rather than a direct role in tumorigenesis [164]. In addition, criteria for a presumed diagnosis of medullary thyroid cancer included surrogate serum markers rather than tissue pathology.

The potential effect of long-acting GLP-1 receptor agonists and mimetics on thyroid C cells in humans requires further investigation. Until such data are available, liraglutide, exenatide once weekly, and semaglutide (oral and injectable) are not recommended for use in patients with a personal or family history of medullary thyroid cancer or multiple endocrine neoplasia 2A or 2B [119,165].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Diabetes mellitus in adults".)

SUMMARY AND RECOMMENDATIONS

●Gastrointestinal peptides – Glucagon-like peptide 1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP) are incretin hormones that are released after the ingestion of a meal (figure 1). They stimulate glucose-dependent insulin release from the pancreatic islets. They also slow gastric emptying, regulate postprandial glucagon, and reduce food intake (table 1). Synthetic GLP-1 receptor agonists are variably resistant to degradation by the enzyme dipeptidyl peptidase 4 (DPP-4), and therefore have a longer half-life, facilitating clinical use. (See 'Gastrointestinal peptides' above.)

●Patient selection – GLP-1-based therapies are appropriate to use in combination with metformin (and/or another oral agent), particularly in patients with existing atherosclerotic cardiovascular disease (ASCVD), when weight loss or avoidance of hypoglycemia is a primary consideration, when glycated hemoglobin (A1C) is well above goal (eg, >8.5 percent), and, and/or when cost is not a major barrier. In these settings, GLP-1 receptor agonists may also be used in combination with basal insulin (with or without metformin). (See 'Patient selection' above and "Management of persistent hyperglycemia in type 2 diabetes mellitus", section on 'Our approach'.)

●Choice of therapy

•With clinical ASCVD – When a decision has been made to use a GLP-1 receptor agonist in a patient with clinical ASCVD, we suggest liraglutide, subcutaneous semaglutide, or dulaglutide (Grade 2B) based on the respective cardiovascular outcomes study results. (See 'Choice of therapy' above and 'Cardiovascular effects' above.)

•Without clinical ASCVD – For patients without clinical ASCVD, we prefer long-acting (liraglutide, subcutaneous semaglutide, dulaglutide, tirzepatide, or once-weekly exenatide) rather than short-acting GLP-1 receptor agonists (table 2). This is predominantly due to patient convenience and better glycemic efficacy. Among the long-acting agents, efficacy for glucose and body weight lowering, patient preference, and payer coverage are important considerations in selecting an agent. (See 'Choice of therapy' above.)

●Administration – Most GLP-1 receptor agonist-based therapies are initiated at a low dose and then slowly advanced to avoid adverse gastrointestinal side effects (table 2). GLP-1 receptor agonist-based therapies can be combined with metformin and most other oral agents. They should not be combined with DPP-4 inhibitors, as there do not appear to be additive effects on glucose lowering. When used in combination with basal insulin, patients using GLP-1 receptor agonist-based therapies compared with placebo achieved glycemic targets at reduced insulin doses and less hypoglycemia or weight gain but more gastrointestinal side effects. (See 'Administration' above.)

●Clinical outcomes – GLP-1 receptor agonists reduce A1C by approximately 1 to 2 percentage points. They lead to weight loss, which varies with the individual drug. The dual GIP and GLP-1 receptor agonist tirzepatide appears to have better glycemic and weight-reducing efficacy compared with either class of agent alone. (See 'Glycemic efficacy' above and 'Weight loss' above.)

Dulaglutide, efpeglenatide, liraglutide, and subcutaneous semaglutide are effective in reducing cardiovascular disease (CVD) in patients with existing ASCVD (table 2). In trials designed to assess cardiovascular outcomes in patients with or at high risk for CVD, liraglutide, semaglutide, dulaglutide, and efpeglenatide (investigational) reduced nephropathy outcomes, whereas there was an increase in retinopathy outcomes with injectable semaglutide. The higher rate of retinopathy complications was unexpected and is likely a consequence of rapid glycemic lowering similar to that seen in other settings. (See 'Cardiovascular effects' above and 'Microvascular outcomes' above and 'Monitoring' above.)

●Adverse effects – The side effects of GLP-1 receptor agonist-based therapies are predominantly gastrointestinal, particularly nausea, vomiting, and diarrhea, and occur consistently in trials in 10 to 50 percent of patients. The risk of hypoglycemia is small. Hypoglycemic events may occur, however, when GLP-1 receptor-based therapies are given in conjunction with diabetes medications known to cause hypoglycemia (eg, insulin, sulfonylureas, glinides). (See 'Adverse effects' above.)