Exenatide: pharmacokinetics, clinical use, and future directions

ABSTRACT

Introduction: The first-in-class glucagon-like peptide-1 receptor agonist (GLP-1RA) exenatide, which was initially approved in 2005, is available in twice-daily (BID) and once-weekly (QW) formulations. Clinical trial data suggest both formulations are effective and safe for patients with type 2 diabetes (T2D), both as monotherapy and as part of combination therapy. Since exenatide was approved, several other GLP-1RAs have become available for clinical use.

Areas covered: Many ongoing clinical trials involving exenatide BID and exenatide QW are investigating new indications (exenatide BID) and new end points and combination therapies (exenatide QW). This review provides an overview of the delivery and pharmacokinetics of both formulations of exenatide, reviews existing data in T2D, and summarizes ongoing investigations.

Expert opinion: Exenatide BID and QW have substantial clinical benefits. Comparisons with other GLP-1RAs demonstrate some differences in efficacy and safety profiles that make assessment of benefit:risk ratios complex. Head-to-head comparisons of QW GLP-1RA formulations may assist in the ranking of GLP-1RAs according to efficacy and safety. Results on the impact of exenatide QW on cardiovascular outcomes are eagerly awaited. The potential clinical utility of exenatide BID in other indications will clarify whether exenatide holds clinical promise in diagnoses other than T2D.

KEYWORDS:

• Exenatide
• extended-release
• glucagon-like peptide-1 receptor agonist
• glycemic control
• type 2 diabetes

1. Introduction

Exenatide is a first-in-class glucagon-like peptide-1 receptor agonist (GLP-1RA), initially approved for treatment of type 2 diabetes (T2D) in 2005. In patients with T2D, GLP-1RAs improve glycemic control through mechanisms including glucose-dependent insulinotropic and glucagonostatic effects, delayed gastric emptying, and decreased food intake [1,2]. Exenatide is available in a short-acting formulation for administration twice daily (BID) before meals and a long-acting formulation, containing the same parent drug as exenatide BID, for administration once weekly (QW) [3,4]. Alternative formulations of exenatide are currently being investigated, including a QW suspension for autoinjection.

In clinical trials of patients with T2D, exenatide BID reduced postprandial glucose and glycated hemoglobin (HbA1c), promoted weight loss, and decreased markers of cardiovascular (CV) disease risk [5–11]. Exenatide QW also reduced HbA1c, fasting glucose (FG), weight, and CV risk factors in patients with T2D in randomized clinical trials [12–19], with benefit observed in patients followed for up to 7 years in open-label extensions [20–22]. Several studies are ongoing to explore the efficacy of exenatide QW in new combination therapies for the treatment of T2D, in special patient populations, and for patient benefits beyond glycemic control.

This review presents an overview of the delivery and pharmacokinetics of exenatide, reviews existing data in patients with T2D, and summarizes ongoing investigations of exenatide.

2. Pharmacokinetics of exenatide

The pharmacokinetics of exenatide BID are dose proportional, with maximum serum concentrations after a single subcutaneous dose of 2.5 or 5 µg of 56 and 85 pg/mL, respectively, and the area under the concentration–time curve of 159 and 340 pg∙h/mL [23]. The time to reach maximum serum concentrations after exenatide BID administration is ~2 h [23].

Exenatide QW is the same active drug as exenatide BID dispersed in microspheres made of the medical grade, biodegradable polymer poly-(D,L-lactide-co-glycolide) [24]. Unlike exenatide BID, which is immediately available and has an elimination half-life of 2.4 h [4,24], the peptide dispersed in the microspheres is released via biodegradation in three stages: initial release (first 48 h), diffusion (~2 weeks), and erosion release (~7 weeks) (Figure 1(a)) [24]. Therapeutic delivery from microspheres is unique as, over time, the exenatide dose in a given week is derived from multiple previous injections undergoing different phases of microsphere dissolution (Figure 1(b)) [25]. With simultaneous release from multiple microspheres at different stages of dissolution, exenatide is released at a constant rate, without peaks or troughs in drug concentration [24].

Figure 1. a) Mechanism of exenatide once weekly release from poly-(D,L-lactide-co-glycolide) microspheres [24]. Reproduced with permission from DeYoung MB, et al. Diabetes Technol Ther. 2011;13:1145–1154, Published by Mary Ann Liebert, Inc., New Rochelle, NY; b) A theoretical model of the plasma concentrations of exenatide after multiple subcutaneous doses of exenatide once weekly. The concentration profile was produced by adding single-dose concentration profiles together at weekly intervals. Reprinted from Clin Ther, Vol. 38, Brunton S, Davidson JA, Exenatide once weekly: a review of pharmacology and treatment considerations in type 2 diabetes, 582–594, Copyright (2016), with permission from Elsevier [33]; c) Schematic of the exenatide once-weekly dual-chamber pen [29]. Republished with permission of J Diabetes Sci Technol, from Evaluation of the dual-chamber pen design for the injection of exenatide once weekly for the treatment of type 2 diabetes, LaRue S, Malloy J, vol. 9, 2015; permission conveyed through Copyright Clearance Center, Inc.

Clinical studies investigating the pharmacokinetics of exenatide QW show that plasma drug concentrations (50 pg/mL) are sufficient for FG reduction after 2 weeks of exenatide 2 mg QW. Steady state occurs after 6–7 weeks, and the compound can be measured in plasma at least 8.5 weeks after a single injection [25,26]. Plasma levels of exenatide QW gradually decrease over 3 months upon treatment discontinuation [25]. Additional studies of exenatide QW show that the pharmacokinetics of exenatide QW are similar among Asian and Caucasian populations [27,28].

The exenatide-containing microspheres require suspension in an aqueous diluent immediately before injection, involving several preparation steps [3]. Recently, a dual-chamber pen was developed for administration of exenatide QW, aiming to simplify microsphere reconstitution, increase usability, and reduce user error (Figure 1(c)) [29,30].

3. Effect of exenatide on clinical and safety outcomes in patients with T2D

In the US, exenatide BID and QW are indicated for use in patients with T2D to improve glycemic control as an adjunct to diet and exercise [3,4,24]. In Europe, exenatide BID and QW are indicated for use in combination with metformin and/or a sulfonylurea, or metformin and/or a thiazolidinedione, in adults with T2D with inadequate glycemic control on maximally tolerated doses of those oral agents [31,32]. In addition, exenatide BID is indicated for use in these patients as an adjunct to basal insulin ± metformin and/or pioglitazone [31]. Phase 1 and 2 studies of exenatide QW in patients with T2D showed that treatment resulted in reductions in HbA1c, FG, postprandial hyperglycemia, and weight [25,26,28].

The DURATION (Diabetes Therapy Utilization: Researching Changes in A1C, Weight, and Other Factors Through Intervention With Exenatide Once Weekly) trials are phase 3 studies investigating the efficacy and safety of exenatide QW versus active controls (Table 1). The core studies (DURATION-1–6) were of 24–30-weeks duration [12–14,16–18]. Long-term extension studies were also undertaken for DURATION-1, -2, and -3; to date, 7-year data were reported for DURATION-1 [22], and 1-year and 3-year data for DURATION-2 [19] and DURATION-3 [20], respectively. Due to the inherent limitations of open-label extension studies (including open-label design creating potential bias, completer populations possibly representing only responders to therapy, and lack of control group) [34,35], these results should be interpreted with caution.

Table 1. Phase 3 DURATION trials and extension studies.

Study (acronym)	Study design (duration)	Treatment (n)	Background therapy	Change from baseline
				HbA1c, %	FG, mmol/L	Body weight, kg
Drucker et al. 2008 (DURATION-1) [17]	Multicenter, open-label, parallel, randomized in pts with T2D (30 wks)	ExQW 2 mg (148)	Monotherapy	ExQW: –1.9	ExQW: –2.3	ExQW: –3.7
		ExBID 5–10 μg BID (147)	or adjunct to OADs	ExBID: –1.5**	ExBID: –1.4***	ExBID: –3.6
Buse et al. 2010 (DURATION-1 extension) [14]	Extension of DURATION-1 (52 wks)	After initial 30 wks, switch from ExBID to ExQW (130) or continued ExQW (128)	Monotherapy or adjunct to OADs	ExQW: –2.0	ExQW: –2.6^a	ExQW: –4.1
				ExBID→ExQW: –2.0	ExBID→ExQW: –2.4^a	ExBID→ExQW: –4.5
MacConell et al. 2013 (DURATION-1 extension) [36]	Extension of DURATION-1 (3 yrs)	All patients received ExQW after initial 30 wks (194)	Monotherapy or adjunct to OADs	ExQW: –1.6	ExQW: −1.8	ExQW: –2.3
Wysham et al. 2015 (DURATION-1 extension) [37]	Extension of DURATION-1 (5 yrs)	All patients received ExQW after initial 30 wks (153)	Monotherapy or adjunct to OADs	ExQW: –1.6	ExQW: –1.6^a	ExQW: –3.0
Bergenstal et al. 2010 (DURATION-2) [12]	Double-blind, multicenter, parallel, randomized in pts with T2D (26 wks)	ExQW 2 mg (160)	MET	ExQW: –1.5	ExQW: –1.8	ExQW: –2.3
		SITA 100 mg OD (166)		SITA: –0.9***	SITA: –0.9**	SITA: –0.8***
		PIO 45 mg OD (165)		PIO: –1.2*	PIO: –1.5	PIO: +2.8***
Diamant et al. 2010 (DURATION-3) [16]	Multicenter, open-label, parallel, randomized in pts with T2D (26 wks)	ExQW 2 mg (233)	OADs	ExQW: –1.5	ExQW: –2.1	ExQW: –2.6
		INS glargine^b (223)		INS: –1.3*	INS: –2.8***	INS: +1.4***
Diamant et al. 2012 (DURATION-3 extension) [38]	Extension of DURATION-3 (84 wks)	ExQW 2 mg (233)	OADs	ExQW: –1.2	ExQW: –2.4	ExQW: –2.1
		INS glargine^b (223)		INS: –1.0*	INS: –3.0**	INS: +2.4***
Diamant et al. 2014 (DURATION-3 extension) [20]	Extension of DURATION-3 (156 wks)	ExQW 2 mg (233)	OADs	ExQW: –1.01	ExQW: –1.73	ExQW: –2.49
		INS glargine^b (223)		INS: –0.81*	INS: –2.65***	INS: +2.01***
Russell-Jones et al. 2012 (DURATION-4) [18]	Double-blind, multicenter, parallel, randomized in treatment-naïve pts with T2D (26 wks)	ExQW 2 mg (248)	None	ExQW: –1.53	ExQW: –2.3	ExQW: –2.0
		MET 2000 mg/day (246)		MET: –1.48	MET: –2.0	MET: –2.0
		PIO 45 mg OD (163)		PIO: –1.63	PIO: –2.6	PIO: +1.5***
		SITA 100 mg OD (163)		SITA: –1.15***	SITA: –1.1***	SITA: –0.8***
Blevins et al. 2011 (DURATION-5) [13]	Multicenter, open-label, randomized in pts with T2D (24 wks)	ExQW 2 mg (129)	Monotherapy or adjunct to OADs	ExQW: –1.6	ExQW: –1.9	ExQW: –2.3
		ExBID 5–10 μg (123)		ExBID: –0.9***	ExBID: –0.7***	ExBID: –1.4
Buse et al. 2013 (DURATION-6) [15]	Multicenter, open-label, parallel, randomized in pts with T2D (26 wks)	ExQW 2 mg (461)	OADs	ExQW: –1.28	ExQW: –1.76	ExQW: –2.68
		LIRA 1.8 mg OD (450)		LIRA: –1.48**	LIRA: –2.12*	LIRA: –3.57***
Frìas et al. 2016 (DURATION-8) [39]	Multicenter, double-blind, randomized in pts with T2D (28 wks)	ExQW 2 mg + DAPA 10 mg OD (231)	MET	ExQW + DAPA: –2.0**,^‡	ExQW + DAPA: –2.30**,^‡	ExQW + DAPA: –3.41**,^†
		ExQW 2 mg (231)		ExQW: –1.6	ExQW: –1.19	ExQW: –1.54
		DAPA 10 mg OD (233)		DAPA: –1.4	DAPA: –1.46	DAPA: –2.19

^aConverted from mg/dL to mmol/L using the conversion factor 0.055499.

^bInsulin glargine was titrated to achieve an FG of 4.0–5.5 mmol/L.

*p < .05 vs. ExQW; **p < .01 vs. ExQW; ***p ≤ .001 vs. ExQW; ^†p < .01 vs. DAPA; ^‡p ≤ .001 vs. DAPA.

DAPA: dapagliflozin; ExBID: exenatide twice daily; ExQW: exenatide once weekly; FG: fasting glucose; HbA1c: glycated hemoglobin; INS: insulin; LIRA: liraglutide; MET: metformin; OADs: oral antidiabetes drugs; OD: once daily; PIO: pioglitazone; pts: patients; SITA: sitagliptin; T2D: type 2 diabetes; wks: weeks; yrs: years.

3.1. Glycated hemoglobin

The DURATION studies have shown that exenatide QW results in HbA1c reductions of 1.3–1.9% over 24–30 weeks [12,13,15–18]. Over 26 weeks, treatment with pioglitazone, sitagliptin, metformin, liraglutide, and insulin glargine resulted in significant HbA1c reductions from baseline, ranging from 0.9–1.6% in patients enrolled in the DURATION studies [12,15,16,18]. HbA1c improvements with exenatide QW were generally similar to or greater than comparators [12,13,15–18], except in the DURATION-6 study, in which the HbA1c reduction from baseline was greater with liraglutide (–1.5% vs. –1.3%, respectively; p = .02), with the between-group difference not meeting criteria for noninferiority [15]. DURATION-1 and -5 investigated treatment with exenatide QW versus exenatide BID in which exenatide QW recipients had a significantly greater HbA1c reduction from baseline versus exenatide BID recipients (p < .0025; Table 1) [13,17].

The DURATION extension studies showed that HbA1c improvements were maintained long-term in patients continuing on therapy [20,21,37,40]. After 3 years’ follow-up, patients receiving exenatide QW in the DURATION-3 study (n = 146 vs. n = 209 completing the core study) had a mean reduction from baseline of −1.01% (Figure 2) [20]. In the DURATION-1 extension, the 5-, 6-, and 7-year least-squares mean reductions from baseline were −1.6% (Figure 3), −1.6%, and –1.5% [21,22,37]; although fewer patients completed 5, 6, and 7 years of treatment compared with the number initially enrolled in the extension study (n = 153, 136, and 122 vs. n = 258), baseline characteristics of the completer populations in the 5-, 6-, and 7-year reports were similar to the intent-to-treat population in the core DURATION-1 study.

Figure 2. Effect of exenatide once weekly on glycemic end points and body weight after 3 years of treatment. a) HbA1c; b) fasting serum glucose; c) self-monitored blood glucose; d) body weight [20]. HbA1c: glycated hemoglobin. Republished with permission of Lancet Diabetes Endocrinol, from Exenatide once weekly versus insulin glargine for type 2 diabetes (DURATION-3): 3-year results of an open-label randomised trial, Diamant M, et al. vol. 2, 2014.

Figure 3. Effect of exenatide once weekly on glycemic end points and body weight after 5 years of treatment. a) HbA1c; b) FPG; c) body weight [37]. *p < .05 for change from baseline. Conversion factor for FPG: 1 mg/dL = 0.0555 mmol/L. FPG: fasting plasma glucose; HbA1c: glycated hemoglobin; LS: least-squares; SE: standard error. Reprinted from Mayo Clin Proc, Vol. 90, Wysham CH, et al. Five-year efficacy and safety data of exenatide once weekly: long-term results from the DURATION-1 randomized clinical trial, 356–365, Copyright (2015), with permission from Elsevier.

3.2. Fasting glucose

Treatment with exenatide QW also resulted in FG reductions of 1.8–2.3 mmol/L in the core DURATION studies [12,13,15–18]. Generally, FG reductions from baseline with exenatide QW after 24–30 weeks were significantly greater than those seen with exenatide BID and sitagliptin, and similar to pioglitazone and metformin; both insulin and liraglutide resulted in greater FG reductions than exenatide QW (p < .05; Table 1). FG reductions were also sustained long-term in the DURATION extension studies [21,22,37]. Exenatide QW-treated patients in the DURATION-3 study had a mean FG reduction from baseline of −1.7 mmol/L at 3 years (Figure 2) [20], while FG reductions at 5, 6, and 7 years in the DURATION-1 study were −1.6 mmol/L (p < .05 vs. baseline; Figure 3) [37], −1.6 mmol/L [21], and –1.3 mmol/L [22].

3.3. Weight loss

Patients in the core DURATION studies who received exenatide QW for 24−30 weeks had a mean weight loss of 2.0–3.7 kg [12,13,15–18]. Exenatide QW-treated patients tended to have similar weight loss at the end of 24−30 weeks compared with patients receiving exenatide BID or metformin, and greater weight loss than with sitagliptin, pioglitazone, and insulin. In contrast, liraglutide-treated patients lost more weight than exenatide QW-treated patients (p = .0005; Table 1) [15]. The average 3-year weight loss with exenatide QW in DURATION-3 was –2.49 kg (Figure 2) [20]. Similar results were seen in the long-term DURATION-1 study, with 5-, 6-, and 7-year mean weight loss with exenatide QW in the completer population (n = 153, n = 136, and n = 122) of –3.0 kg (p < .05 vs. baseline; Figure 3) [37], −4.2 kg [21], and –3.9 kg [22], respectively.

Exenatide BID was more effective than metformin at reducing weight (–5.80 vs. –3.81 kg; p < .01) and HbA1c (–2.10% vs. –1.66%; p = .045) in 59 obese patients with T2D treated for 26 weeks [41]. Exenatide BID also produced greater reductions in weight (–6.16 vs. –3.97 kg; p = .003) and HbA1c (–1.21% vs. –0.73%; p < .0001) than placebo when combined with a lifestyle modification program in a 24-week study in 194 overweight or obese patients with T2D on stable metformin or sulfonylurea therapy [42].

3.4. Safety of exenatide

In rat studies, treatment with long-acting GLP-1RAs resulted in secretion of calcitonin, C-cell proliferation, and formation of C-cell thyroid tumors [43]. Although 2-year exposure to liraglutide in humans did not result in increased calcitonin levels [43], US Food and Drug Administration (FDA) Adverse Event Reporting System data suggest that the number of thyroid cancer events reported is increased in patients treated with exenatide versus rosiglitazone (odds ratio: >3.0; p-value not reported), although this is based on spontaneous adverse event (AE) reports and therefore may be affected by reporting bias [44]. However, exenatide QW is contraindicated in patients with a personal or family history of medullary thyroid carcinoma and in those with medullary endocrine neoplasia syndrome type 2 [3].

Due to postmarketing reports of pancreatitis and pancreatic cancer in patients using incretin-based therapies, the US FDA and the European Medicines Agency (EMA) assessed the pancreatic safety of these agents [45]. The conclusion was that a causal relationship between incretin-based therapies and pancreatitis and pancreatic cancer could not be supported by current data [45]. Nevertheless, the prescribing information for exenatide QW contains warnings regarding the risk of pancreatitis [3]. Additional warnings and precautions concern the risk of hypoglycemia when exenatide QW is used in combination with a sulfonylurea, that exenatide QW is not recommended in patients with severe gastrointestinal disease, and the potential for hypersensitivity reactions and injection-site reactions [3].

The most frequent AEs in exenatide BID-treated patients are transient nausea, vomiting, diarrhea, nasopharyngitis, and headache [46]. The overall safety of exenatide BID was investigated in an integrated analysis of 19 randomized controlled clinical trials, representing more than 1500 patient-years of exposure, which showed that exenatide BID was safe and well tolerated in patients with T2D, with an incidence of AEs similar to that of comparators, except for gastrointestinal-related AEs [46]. Gastrointestinal events were the most common AEs according to system organ class and occurred in 51% of exenatide BID recipients and 21% of comparator recipients. Nausea, vomiting, and diarrhea occurred significantly more often among exenatide BID recipients than comparator-treated patients [46]; generally, nausea and vomiting events among exenatide BID-treated patients were dose-dependent and typically resolved within 1−2 days. Major hypoglycemia was rare, and >90% of hypoglycemic episodes were associated with concomitant sulfonylurea or insulin use regardless of whether patients were receiving exenatide or a comparator [46]. AEs related to renal impairment and pancreatitis were rare and similar between groups; while thyroid neoplasms were also rare, exenatide was associated with a higher exposure-adjusted incidence rate of thyroid neoplasms than pooled comparators (risk difference, 0.3; 95% CI: 0.01−0.53) [46].

A similar analysis of eight randomized controlled trials, representing more than 2100 patient-years of exposure, examined the safety and tolerability of exenatide QW [47]. The analysis showed that exenatide QW is also well tolerated, with a similar overall tolerability profile to non-GLP-1RA drug classes and/or liraglutide, except for the incidence of gastrointestinal AEs, injection-site–related AEs, and hypoglycemia [47]. The frequency of gastrointestinal AEs ranged from 23–46% [47]; exenatide QW-treated patients had a lower frequency of transient nausea and vomiting than exenatide BID-treated patients but a higher frequency than those receiving non-GLP-1RAs. In DURATION-6, the exenatide QW group reported less nausea, vomiting, and diarrhea than the liraglutide group (9%, 4%, and 6% vs. 21%, 11%, and 13%, respectively) [47]. Subcutaneous nodules may occur at the injection site in exenatide QW-treated patients when the microspheres produce an inflammatory foreign body reaction [3,24]. Injection-site reactions were more frequent with exenatide QW than with exenatide BID or non-GLP-1RAs (20% vs. 8.0% and 8.0%, respectively) and more frequent with exenatide QW than with liraglutide (16.0% vs. 3.0%) [47]. The incidence of hypoglycemia was low in all treatment groups in the absence of concomitant sulfonylurea use, but minor hypoglycemia was more frequent in the non-GLP-1RA group than in the exenatide QW and BID groups with and without existing sulfonylurea treatment [47]. There was no difference between the GLP-1RAs exenatide QW and liraglutide for rates of minor hypoglycemia, which were low in the absence of sulfonylureas for both agents. AEs related to renal failure, thyroid neoplasms, and pancreatitis were rare; there was one report of pancreatic neoplasm in an exenatide BID-treated patient [47].

Development of antipeptide antibodies is a common problem with peptide therapies, although the presence of antibodies may or may not have clinical consequences [48]. In an analysis of 3763 patients receiving exenatide BID and 781 receiving exenatide QW, 36.7% of patients were anti-exenatide antibody positive after 30-weeks treatment with exenatide BID in uncontrolled studies, a proportion that decreased over time (24.7% and 16.9% positive at 52 weeks and 3 years, respectively) [48]. In controlled exenatide BID trials, 34.9% of patients were antibody positive at study end (12−52 weeks). Antibody titers in exenatide QW-treated patients peaked at a similar time to those in exenatide BID recipients (6−22 weeks); 56.8% and 45.4% of exenatide QW recipients were antibody positive at weeks 24−30 and 52, respectively [48]. In all cases, most patients exhibited low antibody titers (≤125). No relationship was found between change in HbA1c and antibody titer in exenatide BID-treated patients, and only half of the patients who withdrew from these studies because of loss of glycemic control were antibody positive [48]. For exenatide QW, while there was no correlation between HbA1c reductions in antibody-negative patients and those with a low antibody titer (–1.6% and –1.5%, respectively), HbA1c reductions were less marked in patients with high antibody titers versus those who were antibody negative or had low titers (mean reduction, –0.6%; r = 0.22284; p < .0001); however, reductions in HbA1c were considered clinically meaningful in this subgroup [48]. Antibody-positive patients were more likely to have injection-site reactions than antibody-negative patients, although such events were usually transient, mild, and resolved without treatment interruption. The incidence of antibody development with liraglutide has been shown to be lower than that seen with exenatide (~8–9%), and the presence of anti-liraglutide antibodies did not affect liraglutide efficacy [49]. When patients were switched from exenatide to liraglutide in the LEAD-6 extension phase, the presence of anti-exenatide antibodies did not reduce the efficacy of subsequent liraglutide treatment [49].

4. Combination therapy with exenatide

Management of hyperglycemia with the use of multiple glucose-lowering therapies of complementary mechanisms of action is the cornerstone of effective diabetes management [50,51]. The prescribing information for exenatide BID includes recommendations for its use in combination with various agents, including metformin and insulin [4], and the DURATION trials investigated exenatide QW in a variety of situations: as monotherapy, in combination with metformin therapy, or as an adjunct to certain other oral glucose-lowering therapies (Table 1).

The efficacy of exenatide QW alone and in combination with various other glucose-lowering agents means that it can be used in a variety of patients; this is reflected in the treatment algorithm outlined in the 2015 position statement from the American Diabetes Association/European Association for the Study of Diabetes [51], where GLP-1RAs are recommended for use in combination with metformin, as part of triple therapy with metformin and either a sulfonylurea, thiazolidinedione, or insulin, or in combination with metformin and basal insulin. The prescribing information for exenatide QW states that it should not be used as first-line therapy in patients with T2D inadequately controlled on diet and exercise; rather, exenatide QW is indicated for use in patients inadequately controlled on existing oral glucose-lowering therapies (excluding dipeptidyl peptidase-4 inhibitors and sodium-glucose cotransporter 2 [SGLT2] inhibitors, primarily due to lack of data on these combinations) [3].

5. Efficacy of novel formulations of exenatide

Alternative exenatide formulations that may improve ease of delivery are under investigation or recently investigated (Table 2). The DURATION-NEO studies are investigating the effect of 28 weeks of treatment with exenatide QW suspension for autoinjection on HbA1c in patients with T2D uncontrolled on oral glucose-lowering medications versus exenatide BID (DURATION-NEO-1) and versus sitagliptin or placebo in patients with T2D uncontrolled on metformin (DURATION-NEO-2). Results for DURATION-NEO-1 show that HbA1c reduction from baseline was greater with exenatide QW suspension for autoinjection than with exenatide BID (−1.39% vs. −1.02%, respectively; p = .007), and there was similar weight loss between groups after 28 weeks (−1.49 vs. −1.89 kg) [52]. Available results for DURATION-NEO-2 show a significantly greater HbA1c reduction from baseline with exenatide QW suspension for autoinjection than with sitagliptin (−1.13% vs. −0.75%; p = .021) and a similar degree of weight loss between groups after 28 weeks (−1.12 vs. −1.19 kg) [53].

Table 2. Trials of alternative formulations of exenatide.^a

ClinicalTrials.gov identifier (acronym)	Study design (duration)	Patient population (background glucose-lowering therapy)	Treatment(s)	Expected enrollment	Primary end point	Study status	Available data
Exenatide once-weekly suspension for autoinjection
NCT01652716 (DURATION-NEO-1)	Multicenter, open-label, parallel, randomized (28 wks)	T2D; ≥18 yrs (lifestyle modification; MET, SU, or PIO)	ExQW autoinjector ExBID	377	HbA1c	Completed	Wysham et al. 2014 [52]
NCT01652729 (DURATION-NEO-2)	Multicenter, open-label, parallel, randomized (28 wks)	T2D; ≥18 yrs (MET)	ExQW 2 mg autoinjector SITA 100 mg OD PBO	365	HbA1c	Completed	ClinicalTrials.gov record [53]
Exenatide once-monthly suspension
NCT01104701 (BCB111)	Multicenter, open-label, randomized (20 wks)	T2D; ≥18 yrs (lifestyle modification; MET, PIO, or MET/PIO)	ExQW ExQM	121	HbA1c	Completed	Wysham et al. 2016 [54]
Exenatide continuous sub-dermal delivery (ITCA 650)
NCT00943917	Multicenter, open-label, randomized (24 wks + 24-wk extension)	T2D; ≥18 yrs (MET)	ITCA 650 20 μg/day ITCA 650 40 μg/day ExBID 5–10 μg BID	155	HbA1c	Completed	Henry et al. 2013 [55,56]
NCT01798264	Multicenter, open-label, randomized (4 wks)	T2D; ≥30 yrs (MET, TZD, or MET/TZD)	ITCA 650 40 or 80 μg/day	44	Tolerability	Completed	Henry et al. 2013 [57]
NCT01455870	Multicenter, randomized, double-blind (52 wks)	T2D; ≥18 yrs (MET)	ITCA 650 60 μg/day SITA 100 mg/day	535	HbA1c	Completed	ClinicalTrials.gov record [58]
NCT01455857	Multicenter, randomized, double-blind, placebo-controlled (39 wks)	T2D; ≥18 yrs (lifestyle modification, MET, SU, TZD, or MET/SU, MET/TZD, or SU/TZD)	ITCA 650 40 μg/day ITCA 650 60 μg/day ITCA PBO	460	HbA1c	Completed	ClinicalTrials.gov record [59]
NCT01785771	Multicenter, open-label (39 wks + 26-wk extension)	T2D; ≥18 yrs; HbA1c >10% and ≤12% (lifestyle modification ± MET, SU, TZD, or a combination)	ITCA 650 20–60 μg/day	100	HbA1c	Ongoing, not recruiting	ClinicalTrials.gov record [60]
NCT01455896	Multicenter, randomized, double-blind (104 wks)	T2D; ≥40 yrs (NA)	ITCA 650 60 μg/day ITCA PBO	4000	Time to first occurrence of any MACE component	Ongoing, not recruiting	ClinicalTrials.gov record [61]
NCT02638805	Multicenter, open-label, parallel, randomized (26 wks)	T2D; 18–80 yrs; switching from liraglutide	ITCA 650 20/60 µg/day ITCA 60 µg/day	126	Safety: nausea/vomiting events	Recruiting	ClinicalTrials.gov record [62]

^aStatus of all trials from ClinicalTrials.gov correct as of 28 September 2016.

ExBID: exenatide twice daily; ExQW: exenatide once weekly; ExQM: exenatide once monthly; HbA1c: glycated hemoglobin; MACE: major adverse cardiovascular event; MET: metformin; NA: not available; OD: once daily; PBO: placebo; PIO: pioglitazone; SITA: sitagliptin; SU: sulfonylurea; T2D: type 2 diabetes; TZD: thiazolidinedione; wks: weeks; yrs: years.

The efficacy and safety of an exenatide once-monthly suspension were evaluated in a phase 2 study in 121 patients with T2D [54]. After 20 weeks, reductions in HbA1c were –1.07%, –1.31%, and –1.45%, respectively, with exenatide once-monthly suspension doses of 5, 8, and 11 mg, compared with –1.54% for exenatide QW. Weight loss for the four treatment groups was –1.10, –0.41, –1.14, and –1.36 kg, respectively.

Several trials investigating the subcutaneous exenatide delivery system ITCA 650 are completed or ongoing (Table 2). Results published thus far suggest that continuous, controlled subcutaneous delivery of exenatide reduces HbA1c, plasma glucose, and body weight [55,57]. An extension study showed that reductions in HbA1c, FG, and body weight are maintained with long-term ITCA 650 therapy [56]. Other studies showed that 39 weeks of subcutaneous delivery of exenatide 40 or 60 μg/day via ITCA 650 in patients with HbA1c of 7.5−10% on diet and exercise or oral glucose-lowering drugs reduced HbA1c from baseline by 1.1% and 1.2%, respectively (p = .001 vs. placebo); among patients not receiving sulfonylureas, HbA1c was reduced from baseline by 1.7% [63]. Patients with a baseline HbA1c of >10% and ≤12% had reductions from baseline of −2.8% (p < .001) after 39-weeks treatment with ITCA 650, and 22% of these patients achieved HbA1c <7% [64].

6. Future directions

6.1. Other outcomes in patients with T2D

6.1.1. Patients with T2D and increased CV risk

The ongoing EXenatide Study of Cardiovascular Event Lowering (EXSCEL; NCT01144338) trial is investigating the effect of adding exenatide QW to usual care versus usual care alone on the incidence of major adverse CV events (Table 3) [65]. EXSCEL has enrolled 14,753 patients with T2D, of which 10,784 (73%) had prior CV events [66].

Table 3. Ongoing and recently completed trials with exenatide once weekly and twice daily.^a

ClinicalTrials.gov identifier (acronym)	Sponsor	Study design (duration)	Patient population (background glucose-lowering therapy)	Treatment(s)	Expected enrollment	Primary end point	Study status	Expected completion date
ExQW
T1D
NCT01928329	Yale University	Phase 2, multicenter, randomized, double-blind, parallel (6 mos)	T1D; 18–60 yrs (INS)	ExQW 2 mg PBO	120	HbA1c	Recruiting	April 2018
New combinations, new outcomes, new patient subpopulations, and long-term safety
New combinations
NCT02229383	AstraZeneca	Phase 3, multicenter, randomized, double-blind, parallel (28 wks)	T2D; ≥18 yrs (INS glargine; INS glargine/MET, INS glargine/MET/SU)	ExQW 2 mg	808	HbA1c	Completed	N/A
NCT02229396	AstraZeneca	Phase 3, multicenter, randomized, double-blind, parallel (28 wks then 24-wk extension)	T2D; ≥18 yrs (MET)	ExQW 2 mg + DAPA 10 mg/day ExQW 2 mg DAPA 10 mg/day	695	HbA1c	Active, no longer recruiting	December 2017
NCT02811484	Weill Medical College, Cornell University	Phase 4, randomized, double-blind (24 wks)	Overweight/obese with T2D on insulin; 18–75 yrs (INS)	ExQW 2 mg + DAPA 10 mg/day ExQW 2 mg INS	75	HbA1c	Recruiting	December 2018
New outcomes
NCT02288273	AstraZeneca	Phase 4, multicenter, randomized, double-blind, parallel (10 wks)	T2D; 12–75 yrs (MET)	ExQW 2 mg PBO	239	24-h mean weighted glucose	Completed	N/A
NCT01144338 (EXSCEL)	AstraZeneca	Phase 4, multicenter, randomized, double-blind, parallel (NR)	T2D; ≥18 yrs (OAD(s) ± INS or INS alone)	ExQW 2 mg PBO	14,000	Composite end point of CV death, nonfatal MI, or nonfatal stroke	Active, no longer recruiting	April 2018
NCT02162550	Phoenix VA Health Care System	Phase 4, randomized, double-blind, parallel (18 mos)	T2D; 21–75 yrs (OAD(s) ± INS)	ExQW 2 mg PBO	148	Carotid plaque volume	Recruiting	June 2018
NCT02380521	University of Palermo	Phase 4, open-label (8 mos)	T2D; 18–80 yrs (MET)	ExQW 2 mg	60	Subclinical atherosclerosis	Recruiting	November 2016
New patient populations
NCT02251431	Baylor Research Institute	Phase 3, double-blind, parallel, randomized (38 wks)	T2D; type 4 cardiorenal syndrome; ≥18 yrs (OADs NR; ± INS)	ExQW 2 mg PBO	104	Myocardial injury, kidney injury, cardiac fibrosis, left ventricular strain	Recruiting	June 2018
NCT02496611	University of Minnesota – Clinical and Translational Science Institute	Phase 2, double-blind, parallel, randomized (52 wks)	Severely obese; 12–17 yrs (N/A)	ExQW (dose NR) PBO	100	Weight loss maintenance	Recruiting	July 2020
NCT02635386	Woman’s Hospital Baton Rouge, LA	Phase 3, single-blind, randomized, parallel (24 wks)	Obese, polycystic ovary syndrome; 18–45 yrs (N/A)	ExQW 2 mg DAPA 10 mg ExQW 2 mg + DAPA 10 mg DAPA 10 mg + MET ER 2000 mg	135	Insulin sensitivity/insulin action	Recruiting	September 2018
NCT01794429	Bjørn H. Ebdrup	Phase 3, double-blind, randomized, parallel (3 mos)	Drug-induced obesity: obese pts with a schizophrenic spectrum diagnosis on antipsychotic medication (N/A)	ExQW (dose NR) PBO	45	Weight loss	Completed	N/A
NCT02417142	University of Massachusetts, Worcester	Phase 4, double-blind, randomized (24 wks)	Schizophrenia, schizoaffective disorder (N/A)	ExQW 2 mg PBO	70	Negative symptoms	Recruiting	September 2017
NCT02664441	Seattle Children’s Hospital	Phase 3, double-blind, randomized, parallel, multicenter (36 wks)	Hypothalamic obesity with history of craniopharyngioma; 10–25 yrs (NR)	ExQW (dose NR) PBO	48	BMI	Recruiting	July 2019
NCT02847403	Azienda Ospedaliero–Universitaria di Roma	Phase 3, open-label, randomized, parallel (32 wks)	Dysglycemia or prediabetes with mild cognitive impairment (NR)	ExQW 2 mg PBO	40	Cognitive function (ADAS-cog)	Recruiting	July 2018
Safety
NCT01940770	AstraZeneca	Observational, multicenter (≥3 yrs)	T2D; Japanese (lifestyle modification; SU, MET, TZDs alone or in combination)	ExQW 2 mg	1100	Tolerability	Ongoing, no longer recruiting	April 2020
NCT02533453	AstraZeneca	Phase 4, open-label, multicenter (12–24 wks)	T2D; Korean (MET, SU, TZD, MET/SU, MET/TZD)	ExQW 2 mg	115	Proportion of pts with AEs and SAEs; nature, incidence, and severity of AEs	Ongoing, no longer recruiting	January 2017
ExBID
T1D
NCT01269034	Albert Einstein College of Medicine of Yeshiva University	Phase 4, open-label, randomized, crossover (NR)	New-onset T1D	ExBID 1.25 μg INS ExBID 1.25 μg + INS	20	Postprandial hyperglycemia	Recruiting	December 2017
T2D
NCT02455076	Emory University	Phase 4, open-label, randomized, parallel (10 days)	T2D inpatients (lifestyle modification; OADs; low-dose INS)	ExBID 5 μg INS glargine OADs Rapid-acting INS	150	HbA1c, blood glucose	Recruiting	December 2016
NCT02449603	Xijing Hospital	Phase 4, multicenter, open-label, randomized, parallel (16 wks)	T2D; ≥18 yrs; Chinese (MET)	ExBID 5–10 μg INS aspart 30	150	Glycemic excursions	Recruiting	September 2017
NCT02042664	Assistance Publique Hopitaux De Marseille	Phase 3, open-label, parallel (3 yrs)	T2D; obese (BMI ≥30 kg/m^2); ≥18 yrs (NR)	ExBID (dose NR) MET (dose NR)	44	Intracardiac triglycerides	Completed	N/A
NCT02118376	Shanghai 10th People’s Hospital	Open-label (3 mos)	T2D; obese; 18–65 yrs (NR)	ExBID 5 μg	20	Fat distribution	Unknown	December 2014
NCT02303730	Fudan University	Phase 4, multicenter, open-label, randomized, parallel (24 wks)	T2D; NAFLD; 18–70 yrs (NR)	ExBID 5 μg (4 wks) then 10 μg (20 wks) INS glargine	76	Liver fat content	Recruiting	May 2017
NCT02690883	Nanfang Hospital, Southern Medical University	Phase 4, multicenter, open-label, randomized, parallel (24 wks)	T2D, diabetic nephropathy; 18–70 yrs (NR)	ExBID 5 μg then 10 μg INS lispro	90	UAER	Not yet recruiting	October 2017
Combination therapy
NCT02467920	Huazhong University of Science and Technology	Phase 4, open-label, randomized, parallel (24 wks)	T2D (INS + MET)	INS glargine + ExBID (dose NR) Aspart 30	350	HbA1c	Recruiting	December 2017
NCT01107717	The University of Texas Health Science Center at San Antonio	Open-label, randomized, parallel (3 yrs)	T2D (none or MET <3 mos)	MET + PIO + ExBID MET, GLYB, INS glargine	600	HbA1c	Recruiting	December 2017
Prediabetes
NCT02104739	The University of Texas Health Science Center, Houston	Phase 4, double-blind, randomized, crossover single-dose study (+ 6-wk nonrandomized extension)	Obese (BMI 30–35 kg/m^2); prediabetes; 30–70 yrs (N/A)	ExBID 10 μg SAXA 5 mg PIO 45 mg PBO Extension: ExQW 2 mg	50	Free fatty acids	Recruiting	March 2018
Obesity
NCT01590433	Eleftheria Maratos-Flier	Single-blind, randomized, parallel (52 wks)	Overweight/obese (BMI 28–48 kg/m^2); female; 18–70 yrs (N/A)	ExBID 5 μg then 10 μg PBO	300	Weight	Recruiting	September 2016
NCT01484873	Vanderbilt University	Phase 2, open-label (52 wks)	Hypothalamic obesity (BMI >30 kg/m^2); 18–40 yrs (N/A)	ExBID 5 μg (4 wks) then 10 μg (46 wks)	10	Weight	Completed	N/A
NCT02860923(CRANIOEXE)	University Hospital, Bordeaux	Phase 3, multicenter, double-blind, randomized, parallel (6 mos)	Hypothalamic obesity after cranio-pharyngioma therapy; 18–75 yrs (N/A)	ExBID 5 μg then 10 μg PBO	50	Weight	Not yet recruiting	April 2017
NCT00856609	National Institute of Diabetes and Digestive and Kidney Diseases	Phase 2, double-blind, randomized, parallel (24 wks)	Obese (BMI >30 kg/m^2); 18–55 yrs (N/A)	ExBID 10 µg PBO	180	Weight, energy expenditure, and food intake	Completed	N/A
NCT00845507	University of Cincinnati	Phase 4, double-blind, randomized, parallel (28 days)	Drug-induced obesity (BMI >30 kg/m^2); 18–55 yrs (N/A)	ExBID 5 μg then 10 μg PBO	60	Weight	Completed	N/A
NCT01783717	Huashan Hospital	Open-label (12 wks)	Overweight/obese (BMI ≥28 kg/m^2) or T2D due to hypothalamic damage (NR)	ExBID 5 μg (4 wks) then 10 μg (8 wks)	10	Weight, BMI, HbA1c, glucose, β-cell function, insulin sensitivity	Unknown	July 2015
Cardiac patients
NCT01969149 (ExSTRESS)	Centre Hospitalier Universitaire de Besancon	Phase 2/3, multicenter, open-label, randomized, parallel (NR)	Prevention of stress hyperglycemia in cardiac surgery; ≥18 yrs (NR)^b	EXE IV 0.05 μg/min bolus then 0.025 μg/min INS	544	Proportion of pts with >50% of time in glycemic target range (100–140 mg/dL)	Recruiting	December 2017
NCT00766857	VU University Medical Center	Phase 4, open-label, randomized, parallel (26 wks)	T2D; congestive heart failure (MET)	ExBID 5 μg then 10 μg INS	27	Global cardiac function (LVEF)	Completed	N/A
NCT02432976 (ExeQOL)	Centre Hospitalier Universitaire de Besancon	Phase 2/3, open-label, randomized, parallel (3 mos)	Management of postoperative stress hyperglycemia after CABG (NR)^b	EXE IV 0.05 μg/min bolus then 0.025 μg/min INS	544	Variation in SF-36 score	Recruiting	December 2017
NCT02344641	Chinese Academy of Medical Sciences & Fuwai Hospital	Open-label, randomized, parallel (4 wks)	T2D; heart failure (NR)	EXE 5 μg Control	234	Plasma NT-proBNP	Not yet recruiting	December 2016
NCT01938235 (EMPRES)	University Heath Network, Toronto	Phase 2, double-blind, randomized, parallel (NR)	STEMI with emergent PCI	EXE IV 1.5 µg, then 1.2 µg/h, then 1.9 µg/h PBO	198	Infarct size	Recruiting	January 2018
NCT02404376 (COMBAT-MI)	Hospital Universitari Vall d’Hebron Research Institute	Phase 3, double-blind, randomized (2–7 days)	STEMI (NR)	EXE IV (dose NR) + RIC EXE RIC	360	Myocardial infarct size; MI	Recruiting	September 2018

^aStatus of all trials correct as of 28 September 2016.

^bPatients with and without T2D enrolled; only patients with T2D who did not require insulin were included.

ADAS-cog: Alzheimer’s Disease Assessment Scale-cognitive; AEs: adverse events; BMI: body mass index; CABG: coronary artery bypass graft; CV: cardiovascular; DAPA: dapagliflozin; ExBID: exenatide twice daily; ER: extended release; EXE: exenatide; ExQW: exenatide once weekly; GLYB: glyburide; HbA1c: glycated hemoglobin; INS: insulin; IV: intravenous; LVEF: left ventricular ejection fraction; MET: metformin; MI: myocardial infarction; mos: months; N/A: not applicable; NAFLD: nonalcoholic fatty liver disease; NR: not reported; NT-proBNP: N-terminal of the prohormone brain natriuretic peptide; OAD(s): oral antidiabetes drug(s); PBO: placebo; PCI: percutaneous coronary intervention; PIO: pioglitazone; pts: patients; RIC: remote ischemic conditioning; SAEs: serious adverse events; SAXA: saxagliptin; SF-36: 36-item short-form health survey; STEMI: ST-elevation myocardial infarction; SU: sulfonylurea; T1D: type 1 diabetes; T2D: type 2 diabetes; TZD: thiazolidinedione; wks: weeks; UAER: urinary albumin excretion rate; yrs: years.

The median baseline HbA1c was 8.0% and median diabetes duration was 12 years. Patients were receiving up to three oral glucose-lowering therapies or insulin, alone or in combination with up to two oral glucose-lowering therapies; the most common oral glucose-lowering therapy was metformin (77% of patients), and 46% were receiving insulin at baseline. The primary end point is a composite of CV death, nonfatal myocardial infarction (MI), and nonfatal stroke. EXSCEL is expected to be completed before 2018.

6.1.2. Patients with T2D and nonalcoholic fatty liver disease

Exenatide QW assists in fat loss, and some studies have investigated the use of GLP-1RAs in patients with nonalcoholic fatty liver disease (NAFLD). In a study of obese patients with T2D and hepatic steatosis receiving either exenatide BID (n = 19) or liraglutide (n = 6), intrahepatic lipids decreased by 42% after GLP-1RA treatment, which was accompanied by a reduction in the liver enzymes alanine transaminase (ALT; p < .05) and γ-glutamyl transferase (γ-GT; p < .001) [67], indicating a reduction in liver damage.

Studies of exenatide BID have shown similar significant reductions from baseline in liver enzymes, including aspartate transaminase, ALT, and γ-GT, versus comparators (p < .05). These include a study of 117 patients with poorly controlled T2D and NAFLD [68] and a study of 60 obese patients with T2D and NAFLD [69].

In a pooled analysis of data from 675 patients with T2D who received exenatide QW for 48−52 weeks, exenatide QW treatment for 52 weeks resulted in significant ALT reductions from baseline (−4.3 U/L; p < .0001); 34% of patients with abnormal ALT concentrations at baseline had normal concentrations after 52-weeks treatment [70].

One phase 4 study comparing the effect of exenatide BID and insulin glargine on liver fat content in patients with T2D and NAFLD is currently recruiting patients (NCT02303730; Table 3). At the time of writing, no trials of exenatide QW in patients with T2D and NAFLD are underway.

6.2. Use of exenatide in populations other than T2D

With the efficacy and tolerability of exenatide as a glucose-lowering agent for the treatment of established T2D confirmed, investigations into other therapeutic uses of exenatide have been performed or are underway. To date, the majority of published studies in these wider indications are investigating exenatide BID; however, there are ongoing studies investigating the use of exenatide QW in many of these indications (Table 3).

6.2.1. Patients with type 1 diabetes

Early studies found that exenatide may affect glucose levels in patients with type 1 diabetes (T1D) via its effects on glucagon and gastric emptying rather than insulin secretion [71]; these results suggest that the augmentation of the incretin response by exenatide may be useful in patients with T1D [72]. Indeed, published studies suggest that exenatide BID improves the daily glycemic profiles of patients with T1D and reduces the requirement for insulin. In patients with new-onset T1D, addition of exenatide BID or sitagliptin to insulin therapy resulted in a significant decrease in insulin requirement from baseline (a reduction of 39.2 and 23.7 units, respectively; both p = .03 vs. insulin alone) without increasing endogenous insulin production [72]. In this study, the number of hypoglycemic events was similar among patients receiving insulin alone, insulin plus exenatide BID, and insulin plus sitagliptin (4, 3, and 3, respectively), and there were no instances of ketoacidosis during the follow-up period. In a crossover study with no placebo group, exenatide 10 μg administered four times daily for 6 months in patients with long-term T1D (n = 20; mean disease duration, 21.3 years) resulted in mean weight loss of 4.2 kg (p = .0003), a significant reduction in postprandial insulin requirement versus baseline (0.18 vs. 0.26 units/kg/day; p = .006), lower postprandial glucose levels (7.52 vs. 7.91 mmol/L; p = .0005), and an increase in insulin sensitivity (7.15 vs. 5.21 mg/m²/min per mcU/mL; p = .0039) [73].

In 17 patients with T1D with and without residual insulin production, exenatide BID treatment before a glucose challenge resulted in a 33% reduction in glucose excursions following a liquid meal, delayed gastric emptying (absorption of acetaminophen reduced from 2058 to 686 μg/mL with exenatide administration; p = .017), and suppressed glucagon levels (without treatment: 15,909 ng/L; with treatment: 12,124 ng/L; p = .0015); insulin secretory responses were not affected (p = 1.0) [74], but, overall, these results suggest a reduced glycemic response to food in patients with T1D receiving exenatide BID. A retrospective, observational study of 11 patients with T1D receiving continuous insulin infusions who were treated with exenatide QW for 3 months had significant reductions from baseline in HbA1c (−0.6%; p = .013), body weight (−3.7%; p = .008), and total daily and bolus insulin doses (−12.9% and −7.8 units, respectively; p ≤ .015) [75].

A phase 4, single-dose, open-label, crossover study of the tolerability of exenatide BID in adolescents with T1D (NCT01269034) is currently recruiting patients (Table 3). Recruitment is underway for a phase 2, double-blind, placebo-controlled study investigating HbA1c reductions with exenatide QW in patients with T1D aged 18–60 years (NCT01928329; Table 3). Of note, liraglutide is not being pursued for use in patients with T1D due to a statistically significant increase in the incidence of confirmed symptomatic hypoglycemia with liraglutide versus placebo in the ADJUNCT ONE study [76].

6.2.2. Patients with prediabetes/impaired glucose tolerance

Early glucose-lowering treatment of patients with prediabetes or impaired glucose tolerance (IGT) may delay the development of T2D [77]; however, in this context, it is important to consider that patients receiving ongoing glucose-lowering therapy may have developed T2D but maintained glycemic control due to treatment. In patients with prediabetes/IGT, exenatide BID treatment improved glycemic parameters, improved glucose tolerance, and delayed the development of diabetes [77]. In 105 patients with IGT and/or impaired fasting glucose (IFG) receiving either lifestyle modification, pioglitazone plus metformin, or pioglitazone, metformin, and exenatide BID, there were FG reductions from baseline (all p < .001) with pioglitazone plus metformin and triple therapy (–0.39 vs. –0.61 mmol/L), decreases in plasma glucose area under the concentration–time curve (–12.0% vs. −21.2%), and improvements in insulin sensitivity (42% vs. 52%) and β-cell function (50% vs. 109%) [77]. Patients receiving triple therapy had significant weight loss after 5.5 months (−3.7 kg; p < .001); pioglitazone plus metformin recipients had non-significant weight loss after 6.9 months (−1.4 kg; p = .05) [77].

A randomized, open-label study of 3 months treatment with exenatide BID versus metformin in obese nondiabetic patients with IFG, IGT, or HbA1c ≥5.7% showed that exenatide BID had similar effects to those of metformin on inflammatory markers, endothelial function, oxidative stress, and vascular activation but a significantly greater effect on triglyceride levels (–0.29 vs. –0.03 mmol/L, respectively; p = .032) [78]. A single exenatide dose resulted in a trend toward increased postprandial endothelial function versus placebo in 16 patients with IGT in another study (p = .1) [79].

One phase 4 study is investigating the use of exenatide BID in patients with prediabetes (NCT02104739; Table 3); this study is currently recruiting and is expected to finish in March 2018.

6.2.3. Obesity

Data suggest that endogenous glucagon-like peptide-1 (GLP-1) plays an important role in regulating body weight (predominantly by mediating satiety) and administration of native GLP-1 as well as GLP-1 analogs to overweight or obese patients with T2D results in weight loss [80,81]. Thus, GLP-1RAs such as exenatide could also have potential as weight-loss therapy in nondiabetic patients with obesity [80]. Indeed, a high dose of the GLP-1RA liraglutide has been approved by the FDA as a weight management therapy in overweight or obese individuals [82].

Exenatide BID has been shown to promote weight loss in patients without T2D. A randomized, double-blind, placebo-controlled study in obese women (n = 41) showed that exenatide BID recipients had significant weight loss versus placebo recipients (–2.49 vs. +0.43 kg, respectively; p < .01), which was evident after 2 weeks of treatment [83]. Retrospective analysis of this population showed that patients could be stratified according to response, into high (>5% weight loss, n = 11), moderate (<5% weight loss, n = 14), and nonresponders (no weight loss or weight gain, n = 12). High responders could be identified by their response within 4 weeks of the start of exenatide therapy [83]. In 152 adult obese patients with IGT or IFG undertaking a structured program of diet and exercise, those receiving exenatide BID for 24 weeks lost significantly more weight than placebo recipients (–5.1 vs. –1.6 kg, respectively; p < .001) [84].

In a pilot crossover study of 11 severely obese children and adolescents aged 9−16 years, 3 months of exenatide BID plus lifestyle modification resulted in significantly greater reductions in body mass index (between-group difference, −1.7 kg/m²; p = .01) and body weight (between-group difference, −3.9 kg; p = .02) versus lifestyle modifications alone [85]. Another 3-month, multicenter, randomized, double-blind, placebo-controlled study of 26 severely obese adolescents aged 12−19 years showed that, compared with placebo, exenatide BID recipients had significantly greater reductions from baseline in body mass index (between-group difference, −1.13 kg/m²; p = .02) and body weight (between-group difference, −3.26 kg; p = .02) [86].

Several ongoing studies of exenatide BID with weight loss and/or weight-related end points in obese patients without T2D will conclude in the next few years (Table 3). One trial not yet recruiting will investigate the effect of exenatide QW on weight loss maintenance in severely obese adolescents (NCT02496611; Table 3). Currently, one ongoing trial is investigating the effect of exenatide BID on fat distribution in obese patients with T2D (NCT02118376; Table 3).

6.2.4. Patients with renal impairment

Patients with impaired renal function may accumulate drugs such as exenatide that are renally excreted [87], which may potentially lead to an increase in drug-related toxicity. Exenatide pharmacokinetics studies have shown that the half-life of exenatide is substantially increased and exenatide clearance is significantly (p < .001) reduced in patients with end-stage renal disease [87]. The pharmacokinetics of exenatide BID in patients with mild-to-moderate renal impairment were similar to those of healthy controls, and the authors concluded that no dose adjustment was required for these patients [87]. Further investigations into the impact of exenatide on kidney function and kidney-related AEs suggest no adverse effect. In one real-world study, patients with T2D and varying levels of renal impairment receiving exenatide BID for 1 year did not have changes in renal function or albuminuria different from those seen in patients receiving insulin [88]. In a post hoc analysis of pooled clinical trial data, neither exenatide BID nor exenatide QW was associated with deleterious effects on renal function or kidney-related AEs [89]. In a study of patients aged ≥75 years with renal impairment, maximum plasma exenatide concentrations and exenatide exposure after single doses of exenatide 5 or 10 μg were greater than in younger controls (two-thirds of whom had normal renal function), but between-group differences were not statistically significant [90]. Furthermore, elderly patients receiving exenatide did not experience hypoglycemia or serious AEs [90]. More elderly patients than control patients reported nausea and vomiting upon receiving exenatide 10 μg (p-value not reported), but no other notable differences in the AE profile were evident [90].

Postmarketing reports among exenatide BID-treated patients include instances of increased serum creatinine, renal impairment, acute renal failure, and worsened chronic renal failure [91], although no causal relationship has been established. A retrospective analysis of a medical and pharmacy claims database showed that, although patients with diabetes had a higher risk of acute renal failure than patients without diabetes, there was no evidence that exenatide contributed to that increased risk [92]. However, the prescribing information for both exenatide BID and exenatide QW advises caution when exenatide is used in patients with moderate renal impairment or in those who have undergone renal transplantation, and states that exenatide should not be used in patients with severe renal impairment [3,4].

One study of exenatide QW in patients with T2D and type 4 cardiorenal syndrome (where dysfunction in the heart or kidneys induces dysfunction in the other) is ongoing (NCT02251431; Table 3), with results expected in February 2017.

6.2.5. Patients with increased CV risk

The potential for improved cardiac outcomes in humans was initially proposed based on early animal studies that showed that exenatide had a protective effect against MI in rats [93]. In a subsequent study in patients with ST-elevation MI undergoing percutaneous coronary intervention, administration of intravenous exenatide at the time of reperfusion resulted in an increase in myocardial salvage [94]; a post hoc analysis of this study showed that patients with short system delay who received intravenous exenatide had a 30% decrease in final infarct size, suggesting a potential cardioprotective effect in this patient subgroup [95]. Additional ongoing studies are investigating the impact of exenatide BID when administered to patients with heart failure, after cardiac surgery, and following MI (Table 3).

6.3. Safety and tolerability

Several ongoing studies with a focus on the safety and tolerability of both exenatide BID and QW are ongoing (Table 3). These studies include investigations into the tolerability of exenatide BID in adolescents with T1D aged 12−18 years; a 3-year study of the safety of exenatide QW in Japanese patients with T2D; and the nature of AEs seen with exenatide QW in Korean patients with T2D (Table 3).

6.4. New combination therapy with exenatide

Studies investigating the combination of exenatide QW with insulin or an SGLT2 inhibitor (dapagliflozin) are currently underway (Table 3). The rationale behind the combination of exenatide QW and dapagliflozin is complementary mechanisms, with the glucagonostatic effect of exenatide potentially counteracting the glucagonotropic effect of dapagliflozin, which may improve the glucose-lowering effect of SGLT2 inhibition; both of these noninsulin agents lower HbA1c, weight, and blood pressure, but by different mechanisms that may be complementary. In contrast, the combination of exenatide QW with insulin may represent a QW alternative for patients needing better glucose control despite therapy with basal insulin.

7. Conclusions

Exenatide QW and BID are effective therapies for glycemic control in patients with T2D, both as monotherapy and as a component of combination therapy. Exenatide BID is being explored for use in other indications, such as T1D, prediabetes/IGT, obesity, cardiac disorders, and NAFLD. Clinical trials of exenatide QW are also ongoing and include studies into new end points, such as CV outcomes (EXSCEL), new combinations, and long-term safety. In addition, new formulations of and devices containing exenatide are being developed, such as the dual-chamber pen and the QW suspension for autoinjection. The results of in-progress trials of exenatide are awaited with interest.

8. Expert opinion

Exenatide BID was the first of the GLP-1RA class to reach the markets in 2006. Exenatide QW was approved by the EMA and the US FDA in 2011 and 2012, respectively. Until the approvals of albiglutide and dulaglutide in 2014, exenatide QW was the only QW GLP-1RA accessible to clinicians and patients with T2D. So, given the rising competition, is exenatide still relevant today?

The studies reviewed here have demonstrated that exenatide BID and QW, both as monotherapy and added on to different background therapies, have substantial clinical effects in terms of glycemic control and weight. Long-term extensions with up to 7 years of follow-up originating from the DURATION clinical study program have demonstrated sustained beneficial glycemic and weight-reducing effects over time (DURATION-1, -2, and -3) [19,20,22]. However, these long-term results should be interpreted with caution due to loss of patients and the open-label designs of these extension studies. When evaluating the treatment efficacy of exenatide, it seems most relevant to investigate comparisons with other drugs from the same class.

Head-to-head trials with exenatide and other GLP-1RAs so far include DURATION-6, LEAD-6, GetGoal-X, and AWARD-1. In addition, the DURATION-1 and -5 studies were designed as comparative studies between exenatide QW and exenatide BID. The DURATION-6 trial demonstrated significantly larger reductions in HbA1c (0.2%), FG (0.4 mmol/L), and weight (0.9 kg) after 26-weeks treatment with liraglutide 1.8 mg once daily (QD) compared with exenatide 2.0 mg QW [15]. The LEAD-6 trial compared exenatide 10 µg BID with liraglutide 1.8 mg QD added on to metformin ± sulfonylurea and reported superior effects of liraglutide with respect to HbA1c (0.3%) and FG (1.0 mmol/L), whereas similar reductions in body weight were observed [96]. As expected, short-acting exenatide BID elicited greater reductions in postprandial glucose excursions following breakfast and dinner compared with long-acting liraglutide, most likely due to rapid tachyphylaxis of the initial liraglutide-mediated slowing of gastric emptying. The GetGoal-X trial assessed the efficacy of the two short-acting GLP-1RAs exenatide 10 µg BID and lixisenatide 20 µg QD added on to metformin [97]. Exenatide showed significantly greater reductions in HbA1c (0.2%) and body weight (1.0 kg) compared with lixisenatide, while equal reductions in FG were reported. In the AWARD-1 study, exenatide 10 µg BID was compared with two different dulaglutide doses (0.75 mg and 1.5 mg QW) [98]. Both dosing regimens of dulaglutide demonstrated superior effects on HbA1c (0.3% and 0.5%, respectively), while similar reductions in body weight were observed with exenatide and dulaglutide 1.5 mg. Unfortunately, no head-to-head studies comparing different QW GLP-1RAs have been conducted to date. In the DURATION-1 and -5 studies, exenatide QW was associated with significantly greater HbA1c reductions (0.4% and 0.7%, respectively) compared with exenatide BID. Both studies demonstrated similar reductions in body weight between the exenatide QW and BID treatment groups [13,17].

Overall, exenatide QW and BID have been demonstrated to be well tolerated. Nonetheless, the most common short-term AEs were the well-known gastrointestinal AEs, which, along with injection-site-related AEs, were reported more often after treatment with exenatide compared with non-GLP-1RA drug classes. In DURATION-6, the exenatide QW group reported fewer gastrointestinal AEs than the liraglutide-treated group, whereas subcutaneous nodules were more frequent with exenatide QW [15]. In contrast, the LEAD-6 trial reported very similar rates of these various AEs following treatment with exenatide BID and liraglutide QD [96]. Exenatide BID was associated with higher frequencies of nausea, vomiting, and diarrhea, but fewer injection-site reactions than lixisenatide QD in the GetGoal-X trial [97]. Lower rates of gastrointestinal AEs and similar frequencies of injection-site reactions were reported with exenatide BID compared with dulaglutide 1.5 mg QW in the AWARD-1 trial [98]. Finally, in both DURATION-1 and -5, fewer cases of nausea and vomiting were observed with QW dosing, whereas more injection-site reactions were reported in patients treated with exenatide QW compared with exenatide BID [13,17].

Findings from the DURATION-6 and LEAD-6 studies point to a potential benefit of liraglutide compared with exenatide in terms of efficacy. However, in DURATION-6, liraglutide was also associated with more gastrointestinal AEs; thus, the differences in efficacy and AE frequency may represent a dosing issue. The possibility of less frequent (QW) administration of exenatide is a potential advantage of this drug compared with liraglutide.

Similar to liraglutide, the minor distinctions in efficacy between exenatide and both lixisenatide and dulaglutide were in both cases counterbalanced to some extent by frequencies of AEs, which complicates the benefit:risk assessment. A head-to-head trial including the long-acting GLP-1RAs for QW administration has much potential to help elucidate the potential clinical ranking of these compounds with respect to glycemia, weight, and AEs.

The potential CV benefit is also important. The recently published results from the LEADER trial demonstrated that liraglutide elicited a beneficial effect regarding the composite end point of CV death, nonfatal MI, or nonfatal stroke [99]. Results from the EXSCEL trial are expected with enthusiasm in order to determine whether exenatide will also establish beneficial effects on CV outcomes in patients with T2D. With its 14,753 patients, 27% of whom had no prior history of CV events [66], this trial will also address the potential role of exenatide QW in the primary prevention of CV events in patients with T2D.

While exenatide was the first GLP-1RA for diabetes, in 2015, liraglutide 3.0 mg QD was the first GLP-1RA to obtain an indication other than T2D–namely obesity. The ongoing assessment of the potential clinical utility of exenatide BID in conditions such as T1D, IGT, obesity, and NAFLD will clarify whether exenatide, like liraglutide, holds clinical promise in diagnoses other than T2D.

Article highlights

• In patients with type 2 diabetes (T2D), the glucagon-like peptide-1 receptor agonist (GLP-1RA) class of glucose-lowering medications acts to improve glycemic control through mechanisms including glucose-dependent insulinotropic and glucagonostatic effects, delayed gastric emptying, and decreased food intake.

• The GLP-1RA exenatide is available in twice-daily (BID) and once-weekly (QW) formulations. Clinical trial data show both formulations improve glycemic control, promote weight loss, reduce cardiovascular (CV) risk markers, and are well tolerated, both as monotherapy and as part of combination therapy for T2D.

• There are numerous ongoing clinical trials involving exenatide BID and exenatide QW. Exenatide BID is being explored for use in other indications, such as type 1 diabetes (T1D), prediabetes/impaired glucose tolerance (IGT), obesity, cardiac disorders, and nonalcoholic fatty liver disease.

• Exenatide QW is being evaluated for potential effects on CV end points and as part of new combination therapies. New devices and formulations of exenatide that aim to improve administration and convenience are also being investigated.

• Available data suggest that treatment with exenatide may improve glycemic end points in patients with T1D and in individuals with IGT, and may promote weight loss in obese patients with or without diabetes. Results on the impact of exenatide QW on CV outcomes are eagerly awaited.

This box summarizes key points contained in the article.

Declaration of interest

FK Knop has received fees for consultancy or being part of an advisory board from AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly & Co., Gilead Sciences, Novo Nordisk, Ono Pharmaceuticals, Sanofi-Aventis, and Zealand Pharma, has received fees for lectures from AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly & Co., Gilead Sciences, Merck Sharp & Dohme, Novo Nordisk, Ono Pharmaceuticals, Sanofi-Aventis, and Zealand Pharma, and has received research support from AstraZeneca and Sanofi-Aventis. A Brønden has received research support from the Danish Diabetes Academy supported by the Novo Nordisk Foundation. T Vilsbøll has received fees for consultancy or being part of an advisory board from Amgen, AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly & Co., Merck Sharp & Dohme, Novo Nordisk, and Sanofi-Aventis, has received fees for lectures from AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly & Co., Merck Sharp & Dohme, Novartis, and Novo Nordisk, and has received research support from Merck Sharp & Dohme and Novo Nordisk. Writing assistance, provided by Sheridan Henness, PhD, of inScience Communications, Springer Healthcare, was utilized in the production of this manuscript and funded by AstraZeneca. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Acknowledgments

The manuscript was reviewed by Mary Beth DeYoung, PhD, of AstraZeneca LP.

Additional information

Funding

Medical writing support for this manuscript was funded by AstraZeneca.