Semaglutide: First Global Approval
Abstract
Novo Nordisk has developed a subcutaneous formulation of semaglutide (Ozempic®), a modified human glucagon-like peptide-1 (GLP-1) analogue, for the treat- ment of type 2 diabetes mellitus. It has been developed using Novo Nordisk’s proprietary protein-acylation tech-
nology, and is administered using an injection device. Semaglutide lowers blood glucose by stimulating the release of insulin and also lowers body weight. Once- weekly subcutaneous semaglutide has recently been approved in the US, Puerto Rico and Canada, and has received a positive opinion in the EU for the treatment of patients with type 2 diabetes. It will be launched as the Ozempic® Pen, a pre-filled device. Semaglutide is also under regulatory review in Japan and Switzerland for the treatment of type 2 diabetes. Clinical development for
obesity, non-alcoholic steatohepatitis and non-alcoholic fatty liver disease is underway worldwide. This article summarizes the milestones in the development of semaglutide leading to this first approval for type 2 diabetes.
1 Introduction
Patients with type 2 diabetes have an attenuated incretin effect (defined as augmentation of insulin secretion in response to high glucose levels after a meal), possibly because of reduced levels of the glucagon-like peptide-1 (GLP-1) hormone [1]. GLP-1 improves beta-cell function and insulin secretion, and suppresses glucagon release and hepatic glucose output. In addition, GLP-1 decreases the rate of gastric emptying and acid secretion, resulting in reduced appetite and contributing to weight loss. These effects of GLP-1 have potential clinical value in the treatment of type 2 diabetes and obesity; however, native GLP-1 has a very short half-life, as it is rapidly degraded by proteases [e.g. dipeptidyl peptidase (DPP)-4] [1]. This led to the development of GLP-1 analogues, which are structurally similar to GLP-1 but are resistant to proteolytic degradation. Semaglutide (Ozempic®) is one such modified human GLP-1 analogue, which has been developed by Novo Nordisk for the treatment of type 2 diabetes mellitus.
In December 2017, the US FDA approved subcutaneous once-weekly semaglutide injection as an adjunct to diet and exercise to improve glycaemic control in adults with type 2 diabetes mellitus. It will be launched in the Ozempic® Pen,
the latest generation of Novo Noridsk prefilled devices [2, 3]. The approval of semaglutide was based on results from the SUSTAIN clinical trial programme (Sect. 2.3). It is recommended that semaglutide be initiated at 0.25 mg once weekly and after 4 weeks, the dosage should be increased to 0.5 mg once weekly [4]. If additional gly- caemic control is needed after at least 4 weeks, the dosage may be increased to 1 mg (maximum dose) once weekly [4].
In December 2017, semaglutide also received a positive opinion from the EU Committee for Medicinal Products for Human Use (CHMP) for the treatment of type 2 diabetes. It is indicated for use in adults with insufficiently controlled type 2 diabetes mellitus as an adjunct to diet and exercise [5]. Semaglutide can be used as monotherapy when met- formin is considered inappropriate (due to intolerance or contraindications) or in combination with other medicinal products for the treatment of diabetes [5]. Health Canada in January 2018 approved semaglutide as an adjunct to diet and exercise in adults with type 2 diabetes, when met- formin is not tolerated or contraindicated [6].
Novo Nordisk is also developing an oral formulation of semaglutide for the treatment of type-2 diabetes [7]. The company was developing a transdermal micro-needle patch formulation of the drug in collaboration with Zosano Pharma; however, development was discontinued at the preclinical stage after termination of the collaboration agreement between the two companies [8]. Novo Nordisk has patent protection for semaglutide, a type-2 diabetes drug, with an expiration date in 2031 (Novo Nordisk Form 20-F, February 2017).
Clinical development of semaglutide for obesity, non- alcoholic steatohepatitis and non-alcoholic fatty liver dis- ease is also underway worldwide. This article summarizes the milestones in the development of semaglutide leading to this first approval for type 2 diabetes.
2 Scientific Summary
2.1 Pharmacodynamics
Semaglutide is a GLP-1 analogue with 94% sequence homology to human GLP-1 [4]. Compared with human GLP-1, semaglutide has two amino acid substitutions and is derivatized at lysine 26 with a hydrophilic spacer and a C18 fatty di-acid, which result in increased albumin affinity [4, 9]. In addition, semaglutide is modified at position 8 to provide stabilization against degradation by the enzyme DPP-4. These modifications result in a long plasma half- life, which allows for once-weekly administration of semaglutide. On the other hand, the GLP-1 receptor affinity of semaglutide is decreased three-fold compared with that of liraglutide (0.38 ± 0.06 nM) [4, 9].
Semaglutide selectively binds to and activates the GLP- 1 receptor (the target of na¨ıve GLP-1), thus acting as a GLP-1 receptor agonist [4]. It reduces blood glucose by stimulating insulin secretion and lowering glucagon secretion in a glucose-dependent manner. In patients with type 2 diabetes, 12 weeks’ treatment with semaglutide 1 mg reduced the fasting glucose level by 22%, 2 h post- prandial level by 36% and 24 h glucose level by 22% (values are absolute changes from baseline relative to placebo) [4, 10]. Both first- and second-phase insulin secretion was increased with semaglutide in patients with type 2 diabetes, and the fasting glucagon level (8%), post prandial glucagon response (14–15%) and 24 h glucagon level (12%) was reduced relative to placebo [4, 10]. The insulin secretion rate with semaglutide in patients with type 2 diabetes was similar to that in healthy subjects [4, 10]. Semaglutide also delayed gastric emptying to a minor extent in the early postprandial phase, thus reducing the rate at which glucose appears in the circulation postpran- dially [4]. In patients with type 2 diabetes, semaglutide, compared with placebo, did not alter the counter regulatory responses of increased glucagon during induced hypogly- caemia and did not impair the decrease of C-peptide compared with placebo [4, 10].
In diabetes-induced obese rats, once-weekly adminis- tration of semaglutide 1 or 4 lg/kg/day for 11 weeks, resulted in a dose-dependent decrease in bodyweight, pri- marily because of the loss of fat mass [11]. In mouse models of atherosclerosis, semaglutide protected against atherosclerosis independently of bodyweight, probably via its anti-inflammatory effects [12].Semaglutide at a dose 1.5 times the maximum recom- mended dose did not prolong QTc intervals to a clinically relevant extent in a thorough QTc trial [4, 13].There is an increased risk of hypoglycaemia when semaglutide is coadministered with insulin secretagogues (e.g. sulfonylureas) or insulin, which may be reduced by decreasing the dose of insulin or insulin secretagogues [4].
2.2 Pharmacokinetics
In patients with type 2 diabetes, semaglutide exposure increases in a dose-proportional manner after administra- tion of 0.5 and 1 mg once weekly dosages, and steady state exposure is achieved after 4–5 weeks of once weekly administration [4]. A pharmacokinetic population analysis estimated that the mean steady-state concentrations of semaglutide after 0.5 and 1 mg once weekly administration were approximately 65.0 and 123.0 ng/mL, respectively [4]. The maximum bioavailability of semaglutide is 89% and the maximum drug concentration is reached 1–3 days post dose [4]. Subcutaneous administration of semaglutide in the abdomen, thigh or upper arm results in similar exposure to the drug. Following subcutaneous administra- tion in patients with type 2 diabetes, the mean apparent volume of distribution of semaglutide is & 12.5 L. Semaglutide is extensively ([99%) plasma albumin bound [4].
After a 0.5 mg subcutaneous dose in healthy adult men, semaglutide was eliminated largely by metabolism via proteolytic cleavage of the peptide backbone and sequen- tial beta-oxidation of the fatty acid side-chain; metabolism was not confined to specific organs [14]. Semaglutide-re- lated material was excreted largely via the urine and faeces, with & 3% of the administered dose excreted as intact semaglutide in the urine [14]. In patients with type 2 dia- betes, the apparent clearance of semaglutide is & 0.05 L/h [4]. Semaglutide has an elimination half-life of & 1 week and, therefore, will be present in the circulation for & 5 weeks after the last dose [4].
Age, sex, race and ethnicity [4], as well as renal impairment [15] or hepatic impairment [16] do not affect the pharmacokinetics of semaglutide to a clinically relevant extent. In clinical trials, semaglutide doses of 0.5 and 1 mg provided adequate systemic exposure to the drug across a bodyweight range of 40–198 kg [4]. Semaglutide has very low potential to induce or inhibit CYP enzymes or to inhibit drug transporters, according to in vitro studies [4]. As semaglutide causes a delay of gastric emptying, there is a potential for it to impact the absorption of oral medica- tions administered concomitantly [4]. However, no clini- cally relevant drug-drug interactions were seen when semaglutide was coadministered with metformin, warfarin, digoxin or atorvastatin [17], as well as with the oral con- traceptives ethinylesteradiol or levonorgestrel [18]. Nev- ertheless, caution is advised when coadministering semaglutide with oral medications [4].
2.3 Therapeutic Trials
2.3.1 Type 2 Diabetes
2.3.1.1 As Monotherapy Semaglutide 0.5 or 1.0 mg once weekly monotherapy significantly improved glycaemic control and reduced bodyweight in treatment-naive patients with type 2 diabetes, according to results from the 30-week, phase 3a, placebo-controlled SUSTAIN 1 study (NCT02054897) [19]. At week 30, glycated haemoglobin (HbA1c) levels (primary endpoint) were significantly (p\0.0001) reduced with semaglutide 0.5 or 1.0 mg compared with placebo (mean change from baseline – 1.45 and – 1.55 vs. – 0.02; baseline 8.05%). Mean bodyweight (confirmatory secondary endpoint) was also significantly (p\0.0001) reduced with semaglutide 0.5 or 1.0 mg rela- tive to placebo (mean change from baseline – 3.73 and – 4.53 vs. – 0.98 kg, respectively; baseline 91.93 kg) [19]. In addition, significantly greater (p\0.0001) proportions of patients receiving semaglutide 0.5 and 1.0 mg compared with placebo achieved HbA1c targets of B 6.5% (59 and 60 vs. 13%) and\7% (74 and 72 vs. 25%), and significant (p\0.01) improvements from baseline were seen with semaglutide 0.5 and 1.0 mg relative to placebo in other secondary endpoints, including fasting plasma glucose (FPG) levels, body mass index (BMI) and mean waist circumference (WC) [19]. Patients eligible for this ran- domised, double-blind, multicentre study were aged C 18 years, had type 2 diabetes (HbA1c level of 7–10%; mean disease duration was 4.2 years) and had been treated with diet and exercise alone for C 30 days before screening. Following randomization, patients received once-weekly semaglutide 0.5 mg (n = 128) or 1.0 mg (n = 130), or placebo (n = 129) for 30 weeks [19].
In the 30-week, phase 3 SUSTAIN study (NCT02254291) in Japanese patients, semaglutide 0.5 and 1.0 mg once weekly monotherapy (n = 103 and 102) was significantly (p\0.0001) more effective than oral sita- gliptin 100 mg once daily (n = 103) in improving gly- caemic control (mean change from baseline in HbA1c – 1.9 and – 2.2 vs. – 0.7%; baseline 8.1%) and reducing body weight (mean change from baseline – 2.2 and – 3.9 vs. no change; 69.3 kg) in patients with type 2 diabetes [20]. Efficacy was assessed as secondary outcomes, with safety the primary outcome of the study [20]. Other sec- ondary outcomes were also significantly (p\0.0001) improved with semaglutide 0.5 and 1.0 mg relative to sitagliptin, including the proportion of patients achieving HbA1c targets of B 6.5% (71 and 87 vs. 16%) and\7% (84 and 95 vs. 35%), and the mean change from baseline in BMI and WC [20]. This randomized, open-label, multi- centre study included patients aged C 20 years who were treated with diet and exercise plus oral antidiabetic drug (OAD) monotherapy if their HbA1c levels were 6.5–9.5%, or those treated with diet and exercise only if their HbA1c levels were 7.0–10.5% for C 30 days before screening; the mean disease duration was 8.0 years [20].
2.3.1.2 As Add-On Therapy In the 56-week, phase 3a, SUSTAIN 2 trial (NCT01930188), add-on therapy with semaglutide 0.5 or 1 mg (n = 409 per group) once weekly was more effective than add-on oral sitagliptin 100 mg once daily (n = 407) in improving glycaemic control and reducing body weight in patients with type 2 diabetes who had inadequate glycaemic control with stable dosages of metformin, thiazolidinediones (pioglitazone or rosiglita- zone) or a combination of both [21]. At week 56, add-on semaglutide 0.5 or 1.0 mg was noninferior and superior (both p\0.0001) to add-on sitagliptin in terms of the mean change from baseline in HbA1c levels (primary endpoint; – 1.3 and – 1.6 vs. – 0.5%; baseline 8.1%) [21]. Mean bodyweight (confirmatory secondary endpoint) was also significantly (p\0.0001) reduced with semaglutide 0.5 and 1.0 mg relative to sitagliptin (mean change from baseline – 4.3 and – 6.1 vs. – 1.9 kg; baseline 89.5 kg) [21]. Sig- nificantly (p\0.0001) greater proportions of semaglutide 0.5 and 1.0 mg than sitagliptin recipients achieved HbA1c targets of B 6.5% (53 and 66 vs. 20%) and\7% (69 and 78 vs. 36%), and significantly (p\0.0001) greater improve- ments from baseline were seen with semaglutide than sitagliptin in other secondary endpoints, including FPG levels, BMI and WC [21]. This randomized, double-blind, multicentre study included patients aged C 18 years (C 20 years in Japan) with type 2 diabetes (mean disease duration 6.6 years) and inadequate glycaemic control (HbA1c 7–10%) while receiving stable dosages of met- formin, pioglitazone, rosiglitazone or metformin in com- bination with either thiazolidinedione for 90 days before screening [21].
The 56-week, phase 3a, open-label SUSTAIN 3 study (NCT01885208) showed that add-on therapy with semaglutide 1.0 mg once weekly (n = 404) was more effective than add-on oral exenatide extended-release (ER) 2.0 mg once weekly (n = 405) in improving glycaemic control and reducing body weight in patients with type 2 diabetes who had inadequate glycaemic control despite treatment with one or two OADs [22]. At week 56, add-on semaglutide was noninferior and superior (both p\0.0001) to add-on exenatide ER in terms of the mean change from baseline in HbA1c levels (confirmatory primary endpoint; – 1.5 vs. – 0.9%; baseline 8.3%) [22]. Semaglutide was also significantly (p\0.0001) more effective than exe- natide ER in reducing bodyweight (secondary endpoint; mean change from baseline – 5.6 vs. – 1.9 kg; baseline 95.8 kg) [22]. In addition, significantly greater proportions of semaglutide than exenatide ER recipients achieved HbA1c targets of B 6.5% (47 vs. 22%) and\7% (67 vs. 40%), and other secondary endpoints, including FPG, BMI and WC, improved from baseline to a significantly (p\0.0001) greater extent with semaglutide than exenatide ER [22]. Patients eligible for the study were aged C 18 years, had type 2 diabetes (HbA1c 7.0–10.5%; mean dis- ease duration was 9.2 years) and were receiving stable dosages of one or two OADs (metformin, and/or thiazolidinediones, and/or sulfonylureas) for C 90 days before screening [22].
The 30-week, phase 3 SUSTAIN 4 study (NCT02128932) showed that semaglutide 0.5 or 1.0 mg once weekly as add-on therapy (n = 362 and 360 per group) was noninferior to add-on therapy with insulin glargine once weekly (mean dose 29.2 IU/day at week 30; n = 360) in terms of glycaemic control in insulin-na¨ıve patients with type 2 diabetes (mean disease duration 8.6 years) who had inadequate glycaemic control with stable dosages of metformin ± sulfonylureas [23]. Pre- specified hierarchical testing showed that at week 30, semaglutide 0.5 or 1.0 mg significantly (p\0.0001) improved HbA1c levels relative to insulin glargine (pri- mary endpoint; mean change from baseline – 1.21 and – 1.64 vs. – 0.83%; baseline 8.2%) and significantly (p\0.0001) reduced mean bodyweight (confirmatory sec- ondary endpoint; mean change from baseline – 3.47 and – 5.17 vs. 1.15; baseline 93.5 kg) [23]. Significantly (p\0.0001) greater proportions of semaglutide than insu- lin glargine recipients achieved HbA1c targets of B 6.5% (37 and 54 vs. 18%) and\7% (57 and 73 vs. 38%), and other secondary endpoints including, FPG, BMI and WC, improved from baseline to a significantly (p\0.001) greater extent with semaglutide than insulin glargine [23]. Patients eligible for the randomized, open-label study were aged C 18 years, were insulin-naive and had inadequately controlled glycaemia (HbA1c 7.0–10.0%) despite treat- ment with stable dosages of metformin alone or in com- bination with a sulfonylurea for C 90 days before screening [23].
Add-on semaglutide 0.5 and 1.0 mg once weekly (n = 133 and 132 per group) provided better glycaemic control and weight loss than add-on placebo (n = 131) in patients with type 2 diabetes (mean disease duration 13.3 years) who had inadequate glycaemic control despite ongoing treatment with basal insulin ± metformin, according to results of the 30-week, phase 3a, double-blind SUSTAIN 5 study (NCT02305381) [24]. At week 30, HbA1c levels were significantly reduced with semaglutide 0.5 and 1.0 mg once weekly compared with placebo (pri- mary endpoint; mean change from baseline – 1.4 and – 1.8 vs. 0.1%; baseline 8.4%) [24]. Weight loss at week 30 was also significantly greater in the semaglutide 0.5 and 1.0 mg groups than in the placebo group (mean change from baseline – 3.7 and – 6.4 vs. – 1.4 kg; baseline 91.7 kg) [24].
An earlier 12-week, randomized, double-blind, dose- finding, phase 2 study (NCT00696657; n = 415) had shown that add-on semaglutide once weekly (0.1–0.8 mg fixed-dose, or 0.8 or 1.6 mg after dose-escalation) dose- dependently reduced HbA1c levels (mean change from baseline – 0.6 to – 1.7%; baseline 8.1%) and bodyweight (mean change from baseline – 0.8 to – 4.8 kg; baseline 87.5 kg) in patients with type 2 diabetes who were inade- quately controlled (HbA1c 7–10%) with diet and exercise (alone or in combination with metformin) [25].
2.3.1.3 Cardiovascular Safety Trial In patients with type 2 diabetes who are at high cardiovascular (CV) risk, semaglutide 0.5 or 1.0 mg once weekly was noninferior to placebo for CV outcomes, according to results from the large (n = 3297), 104-week, phase 3, randomized, double- blind, placebo-controlled, multicentre SUSTAIN 6 study (NCT01720446) [26]. Eligible patients had type 2 diabetes (HbA1c C 7%) and had not been treated with an antihy- perglycemic drug or had been treated with no more than two OADs, with or without basal or premixed insulin. The median observation duration for the study was 2.1 years [26]. The mean disease duration was 13.9 years and 83.0% of patients at baseline had established CV disease (CVD) [26]. Patients were randomized to receive semaglutide 0.5 mg or 1.0 mg (n = 826 and 822) or volume-matched placebo 0.5 or 1.0 mg (n = 824 and 825) [26].
The primary composite outcome was the first occurrence of death from CV causes, nonfatal myocardial infarction (MI; including silent) or nonfatal stroke. Over a planned observation period of 109 months (104-week treatment period and 5-week follow-up period), the composite pri- mary outcome occurred in 6.6% of patients in the semaglutide group and 8.9% of patients in the pooled placebo group [hazard ratio (HR) 0.74; 95% CI 0.58–0.95], demonstrating noninferiority (p\0.001) of semaglutide to placebo (the upper limit of the 95% CI of the HR was less than the noninferiority margin of 1.8) [26]. Subsequent superiority testing of the primary composite outcome indicated significant (p = 0.02) benefit of semaglutide over placebo (superiority testing was not predefined or adjusted for multiplicity). Nonfatal stroke occurred in fewer semaglutide than placebo recipients (1.6 vs. 2.7%; p = 0.04), while there was no significant difference between semaglutide and placebo groups for the rates of nonfatal MI (2.9 vs. 3.9%) or the rates of death from any cause (3.8 vs. 3.6%) or CV causes (2.7 vs. 2.8%). The rate of new or worsening nephropathy (3.8 vs. 6.1%) was also lower in the semaglutide group, but the rate of retinopathy complications (vitreous haemorrhage, blindness or condi- tions requiring treatment with an intravitreal agent or photocoagulation) was significantly higher with semaglu- tide (3.0 vs. 1.8%; p = 0.02). In addition, by week 104, patients in the semaglutide 0.5 or 1.0 mg once weekly groups had significantly (p\0.0001) lower HbA1c levels (mean change from baseline – 1.1 and – 1.4 vs. – 0.4 and
– 0.4% in the placebo groups; baseline 8.7%) and greater weight loss (- 3.6 and – 4.9 vs. – 0.7 and – 0.5 kg; baseline 92.1 kg) than patients in the placebo groups [26].
2.3.2 Obesity
In April 2017, Novo Nordisk completed a phase 2 trial assessing the safety and efficacy of once-daily subcuta- neous semaglutide in obese patients without diabetes mellitus (NCT02453711). Relative change in body weight at 52 weeks was the primary endpoint. The randomised, double-blind trial was initiated in October 2015 and enrolled 957 patients in the US, Australia, Belgium, Canada, Germany, Israel, Russia and the UK. Novo Nor- disk announced initial results from the phase 2 trial in June 2017, indicating up to 13.8% weight loss with semaglutide therapy [27]. Detailed results are awaited with interest.
2.4 Adverse Events
Subcutaneous semaglutide 0.5 and 1.0 mg once weekly was generally well tolerated in clinical trials in patients with type 2 diabetes [4]. Data were pooled from two pla- cebo-controlled semaglutide trials (a monotherapy trial and a trial in combination with basal insulin) in patients with type 2 diabetes (n = 521) who received semaglutide 0.5 or 1.0 mg once weekly for a mean exposure period of 32.9 weeks. The most common adverse reactions with semaglutide 0.5 and 1.0 mg (n = 260 and 261) compared with placebo (n = 262) were nausea (15.8 and 20.3 vs. 6.1%), vomiting (5.0 and 9.2 vs. 2.3%), diarrhoea (8.5 and 8.8 vs. 1.9%), abdominal pain (7.3 and 5.7 vs. 4.6%) and constipation (5.0 and 3.1 vs. 1.5%). Gastrointestinal (GI) adverse reactions, overall, occurred approximately twice as frequently with semaglutide 0.5 and 1.0 mg than placebo (32.7 and 36.4 vs. 15.3%); the proportion of patients dis- continuing treatment because of GI adverse reactions was also higher in the semaglutide groups than the placebo group (3.1 and 3.8 vs. 0.4%). In addition to these GI dis- orders, semaglutide 0.5 and 1.0 mg was associated with dyspepsia (3.5 and 2.7 vs. 1.9% with placebo), eructation (2.7 and 1.1 vs. 0%), flatulence (0.4 and 1.5 vs. 0.8%), gastroesophageal reflux disease (1.9 and 1.5 vs. 0%) and gastritis (0.8 and 0.4 vs. 0.8%) [4]. Hypoglycaemia occurred infrequently with semaglutide 0.5 and 1.0 mg monotherapy or placebo (n = 127, 130 and 129, respectively), with no patient in any treatment group reporting severe hypoglycaemia; symptomatic hypogly- caemia occurred in 1.6 and 3.8 versus 0% of patients, and severe or blood glucose confirmed symptomatic hypogly- caemia was reported in and 0 and 0 versus 1.6% of patients [4]. Documented symptomatic hypoglycaemia was defined as B70 mg/mL glucose threshold, and severe hypogly- caemia as an episode requiring assistance of another person.
Hypoglycaemia was more frequent with semaglutide when used in combination with basal insulin or sulfony- lureas [4]. When semaglutide 0.5 and 1.0 mg was used as add-on therapy to basal insulin ± metformin versus add-on placebo ± metformin (n = 132, 131 and 132, respectively), severe hypoglycaemia occurred in 0 and 1.5 versus 0% of patients, symptomatic hypoglycaemia in 16.7 and 29.8 versus 15.2% of patients, and severe or blood glucose confirmed hypoglycaemia in 8.3 and 10.7 versus 5.3% of patients [4]. When coadministered with a sulfonylurea, severe hypoglycaemia was reported in 0.8 and 1.2% of patients receiving semaglutide 0.5 or 1.0 mg, documented symptomatic hypoglycaemia occurred in 17.3 and 24.4% of patients, respectively, and severe or blood glucose confirmed hypoglycaemia in 6.5 and 10.4% of patients [4]. In rodents, semaglutide is associated with a dose-de- pendent and treatment duration-dependent increase in the incidence of thyroid C-cell tumours (adenomas and carci- nomas) after lifetime exposure at clinically relevant plasma exposures [4]. The US prescribing information carries a boxed warning regarding the same. It is not known whether semaglutide causes thyroid C-cell tumours (including medullary thyroid carcinoma) in humans. In the USA, treatment with semaglutide is contraindicated in patients with a personal or family history of medullary thyroid carcinoma or in patients with multiple endocrine neoplasia syndrome type 2 [4].
Across placebo- and active comparator-controlled trials, 1% (32 patients) of semaglutide recipients developed anti- drug antibodies to semaglutide, of which 0.6% (19 patients) developed antibodies cross-reacting with native GLP-1 [4].In the large 104-week CV outcomes trial (SUSTAIN 6; NCT01720446), the majority of patients experienced adverse events (AEs) with semaglutide 0.5 and 1.0 mg (89.6 and 89.1 vs. 90.8 and 89.2% with corresponding placebo recipients), with most AEs of mild or moderate severity and occurring during the first 30 weeks of treat- ment [26]. Serious AEs were reported in 35.0 and 33.6% of semaglutide 0.5 and 1.0 mg recipients versus 39.9 and 36.1% of placebo recipients. AEs led to treatment discon- tinuation in 11.5 and 14.5% of semaglutide 0.5 and 1.0 mg recipients versus 5.7 and 7.6% of placebo recipients, lar- gely due to GI disorders. The most common AEs with semaglutide 0.5 and 1.0 mg were GI disorders (50.7 and
52.3 vs. 35.7 and 35.2% with placebo), with diarrhoea (17.9 and 18.4 vs. 11.9 and 10.5%), nausea (17.3 and 21.9 vs. 7.5 and 8.1%) and vomiting (10.5 and 14.8 vs. 5.2 and 4.1%) reported most frequently. In the semaglutide 0.5 and 1.0 mg groups versus placebo groups, cardiac disorders occurred in 20.9 and 18.2 versus 22.9 and 21.0% of patients; severe or symptomatic hypoglycaemic events in
23.1 and 21.7 versus 21.5 and 21.0% of patients; neoplasms in 8.0 and 10.8 versus 8.5 and 8.4% of patients; acute renal failure in 5.1 and 2.8 versus 4.1 and 4.2% of patients; and acute pancreatitis in 6 and 3 versus 3 and 9 patients [26]. Diabetic retinopathy complications occurred in signifi- cantly (p = 0.02) more patients receiving semaglutide 0.5 or 1.0 mg than patients receiving placebo in SUSTAIN 6 [50 (3.0%) vs. 29 (1.8%) patients], with the between-group treatment difference seen very early in the trial [26]. However, no imbalance in diabetic retinopathy AEs was seen between patients receiving semaglutide 0.5 and 1.0 mg and those receiving comparators in pooled data from SUSTAIN 1–5 trials and two Japanese trials (2.1 and 1.5 vs. 2.0%); all AEs were of mild or moderate severity [28]. Rapid improvement in glucose control has been associated with a temporary worsening of diabetic retinopathy [4], for example during treatment with insulin. The majority of the effect with semaglutide relative to placebo with regard to diabetic retinopathy may also be because of the magnitude and rapidity of HbA1c reduction in the first 16 weeks of therapy in patients with pre-existing diabetic retinopathy, poor glycaemic control at baseline and who were treated with insulin [28].
2.5 Ongoing Clinical Trials
In August 2017, Novo Nordisk initiated the phase 3 SUSTAIN—CHINA MRCT trial to evaluate the effect of subcutaneous semaglutide 0.5 and 1.0 mg once weekly versus sitagliptin 100 mg once daily on glycaemic control after 30 weeks of treatment (NCT03061214). The ran- domised, double-blind, double-dummy, active-controlled, multicentre and multinationall trial is enrolling approxi- mately 1050 patients in Brazil, China, Hong Kong, South Korea, South Africa, Taiwan and Ukraine.
In March 2017, Novo Nordisk initiated the phase 3 SUSTAIN 8 trial to compare the effect of subcutaneous semaglutide 1.0 mg once weekly with that of oral cana- gliflozin 300 mg once daily on glycaemic control in sub- jects with type 2 diabetes on a background treatment of metformin (NCT03136484). The randomised, double- blind, multinational trial has enrolled approximately 784 patients in the US, Argentina, Brazil, Canada, India, Ire- land, Italy, Lebanon, Malaysia, Mexico, Sweden and the UK.
In March 2017, Novo Nordisk initiated the phase 3 SUSTAIN 9 trial to compare the effect of semaglutide subcutaneous 1 mg once weekly with that of placebo (as add-on to sodium glucose co-transporter-2 inhibitor monotherapy or in combination with metformin or sul- fonylurea) on glycaemic control after 30 weeks of treatment in patients with type 2 diabetes (NCT03086330). This randomised, double-blind, placebo-controlled trial is enrolling approximately 300 patients in the US, Austria, Canada, Japan, Norway, Puerto Rico and Russia.
In June 2017, Novo Nordisk initiated the SUSTAIN 10 phase 3 trial to compare the effect of subcutaneous semaglutide 1 mg once weekly versus subcutaneous liraglutide 1.2 mg once-daily, as add-on to 1–3 oral OADs, on glycaemic control after 30 weeks of treatment in patients with type-2 diabetes (NCT03191396). This ran- domised trial has enrolled approximately 576 patients in Bulgaria, Czech Republic, Finland, France, Germany, Hungary, Italy, Poland, Sweden, Finland, Slovenia, Spain, Sweden and the UK.
In November 2016, Novo Nordisk initiated a phase 2 trial to evaluate the efficacy and safety of three doses of subcutaneous semaglutide once daily versus placebo in patients with non-alcoholic steatohepatitis (NCT02970942). The 72-week, randomised, double-blind trial is recruiting approximately 372 patients in Austria, Belgium, Bulgaria, Canada, Finland, France, Greece, Japan, Netherlands, Puerto Rico, Russia, Spain, Sweden, UK and USA.
3 Current Status
Semaglutide received its first global approval on 5 December 2017 in the USA to improve glycaemic control in adults with type 2 diabetes, and was subsequently approved by Health Canada on the 4 January 2018. It also received a positive opinion from the EU CHMP on 14 December 2017 for the treatment of patients with type 2 diabetes.