Type 2 diabetes is a progressive disease associated with macrovascular and microvascular comorbidities. Worldwide, approximately 382 million people have diabetes, and this is expected to increase to 592 million by 2035 (International Diabetes Federation [IDF], 2013).
When considering the management of type 2 diabetes, various UK, European and US guidelines (NICE, 2009; Handelsman et al, 2011; Inzucchi et al, 2012; Garber et al, 2013) recommend early, ongoing dietary improvement and increased physical activity, alongside diabetes education. As diet and exercise usually become inadequate within a year, and HbA1c levels start to rise (Nathan et al, 2009), step-wise treatment intensification is needed to restore HbA1c to target levels (NICE, 2009; Inzucchi et al, 2012).
After starting oral antidiabetes drug (OAD) monotherapy, typically metformin, dual and later triple therapy may be added. The addition of insulin may be required if hyperglycaemia continues. Current guidelines stress the importance of individualising treatment, balancing glucose lowering with risk of hypoglycaemia and other adverse effects.
Traditional therapies
Traditional OADs, such as metformin, sulphonylureas (SUs) or thiazolidinediones (TZDs), and insulin can normalise blood glucose levels, but may lead to limiting adverse events (Pi-Sunyer, 2009; Pollack et al, 2010): SUs and insulin are associated with a risk of hypoglycaemia; metformin and alpha-glucosidase inhibitors may produce gastrointestinal disturbances; TZDs can lead to oedema; SUs, glinides, TZDs and insulin are associated with weight gain (Rodbard et al, 2009). Options may also be limited by disease. For example, metformin and TZDs are contraindicated in some people with heart failure and metformin in people with renal disease (Rodbard et al, 2009).
Given these limitations, the emergence of incretin therapies is encouraging. Oral dipeptidyl peptidase-4 inhibitors (DPP-4Is) and injectable glucagon-like peptide-1 receptor agonists (GLP-1RAs) stimulate insulin release and inhibit glucagon secretion in a glucose-dependent manner and are therefore associated with a low inherent risk of hypoglycaemia (Inzucchi et al, 2012). Furthermore, they demonstrate weight neutrality (DPP-4Is) or weight loss (GLP-1RAs) and data suggest modest benefits on some cardiovascular risk factors (Mudaliar and Henry, 2009; Monami et al, 2013; 2014).
GLP-1 and the incretin system in type 2 diabetes
GLP-1 is a major incretin hormone secreted by the gastrointestinal system in response to food intake (Holst et al, 2009). Insulin produced after eating is partly stimulated by this hormone (Nauck et al, 1986a; 1986b) within the “incretin response” – where insulin secretion is greater following an oral rather than intravenous glucose load, despite similar plasma glucose levels (Figure 1). GLP-1 also contributes to feelings of fullness, which reduce energy intake and promote weight loss. However, native GLP-1 has a half-life of <2 minutes (Vilsboll et al, 2003), and is therefore impractical as a therapeutic agent.
Incretin therapies: Overview and mechanism of action
DPP-4Is and GLP-1RAs both increase the effects of GLP-1. DPP-4Is inhibit the activity of the enzyme, DPP-4, which rapidly degrades GLP-1, in order to prolong the effects of native GLP-1. GLP-1RAs provide activation of the GLP-1 receptor while remaining resistant to degradation by DPP-4.
Several incretin-based therapies have been approved for use. Four GLP-1RAs (exenatide, exenatide extended release, liraglutide, and lixisenatide) are available. There are now five DPP-4Is available (sitagliptin, saxagliptin, vildagliptin, linagliptin and alogliptin).
Exenatide shows 53% amino-acid similarity to native GLP-1 (Chen and Drucker, 1997), but is resistant to breakdown by DPP-4 and therefore has a half-life of 2–4 hours, necessitating twice-daily (BD) injection. The long-acting release formulation of exenatide allows once-weekly (OW) dosing (electronic Medicines Compendium [eMC], 2014a).
Liraglutide is a human GLP-1RA with 97% homology to native GLP-1 (Knudsen et al, 2000). Liraglutide is administered by subcutaneous injection, where it self-associates into heptamers resulting in delayed absorption (Steensgaard et al, 2008). With a half-life of 11–15 hours, liraglutide is administered once daily (OD; Elbrond et al, 2002; eMC, 2014b).
Lixisenatide is a synthetic GLP-1RA, a 44-amino-acid peptide that differs from exendin-4 (a peptide produced exclusively by the salivary glands of the Gila monster [Heloderma suspectum]) by the addition of six lysine residues and the deletion of one proline at the C-terminal (Werner et al, 2010). The elimination half-life of lixisenatide is approximately 2–4 hours (Barnett, 2011) but, unlike exenatide, it is administered once-daily (eMC, 2014c) as the once-daily regimen gave the best balance of efficacy and tolerability in clinical trials (Ratner et al, 2010).
DPP-4 is a cell-surface aminopeptidase enzyme that degrades some gastrointestinal hormones, neuropeptides, cytokines and chemokines (Drucker and Nauck, 2006). The DPP-4Is are reversible, competitive inhibitors of DPP-4 and inhibit approximately 80–90% of DPP-4 activity (Forst et al, 2011; Herman et al, 2005; He et al, 2007), leading to a 2–3-fold elevation in GLP-1 (Herman et al, 2005; Mari et al, 2005). The half-life of 3–21 hours allows OD (saxagliptin, sitagliptin, linagliptin, alogliptin [eMC, 2014d,e,f,g]) or BD (vildagliptin) regimens (eMC, 2014h), administered orally.
Efficacy and safety of GLP-1RAs and DPP-4 inhibitors
Comparison of currently available GLP-1 RAs
Several trials have directly compared the two classes of incretin therapy (Table 1). At the time of writing, five such trials compared different GLP-1RAs. In the LEAD-6 study, liraglutide led to significantly greater reductions in HbA1c and fasting plasma glucose (FPG) versus exenatide, with similar reductions in weight and systolic blood pressure (SBP; Buse et al, 2009). In an extension phase, patients switched from exenatide to liraglutide experienced additional reductions in HbA1c, FPG, weight and SBP (Buse et al, 2010a). Both agents were well-tolerated, with significantly less ongoing nausea in liraglutide- (3%) versus exenatide-treated patients (9%) at 26 weeks (P<0.0001). Minor hypoglycaemia was significantly less frequent with liraglutide (1.93 versus 2.60 events/patient-year; P=0.01, Buse et al, 2009).
In a 26-week study comparing liraglutide (1.8 mg OD) with exenatide 2 mg OW (Buse et al, 2013), liraglutide led to greater reductions in HbA1c, weight and SBP, although gastrointestinal adverse events were more common in the liraglutide group (21% versus 9% in exenatide group). Hypoglycaemia rates were similar between groups (12% in the liraglutide group and 15% in the exenatide group for people taking concomitant SU, and 3% in the liraglutide group and in 4% in the exenatide group for those without concomitant SU).
In a 24-week study comparing lixisenatide to exenatide BD, both led to reductions in HbA1c and weight, with lixisenatide associated with lower levels of symptomatic hypoglycaemia (2.5 versus 7.9%; P<0.05) and nausea (24.5 versus 35.1%; P<0.05; Rosenstock et al, 2013).
Three studies (Drucker et al, 2008; Buse et al, 2010b; Blevins et al, 2011) compared formulations of exenatide (Table 1). After 30 weeks in DURATION-1, exenatide OW led to significantly greater reductions in HbA1c and FPG compared with exenatide BD, and more people reached HbA1c ≤53 mmol/mol (77% versus 61%; P=0.004). Reductions in weight were similar. The most commonly reported adverse event was nausea, occurring in fewer people taking exenatide OW (26.4% versus 34.5%). Rates of minor hypoglycaemia were very low in both groups (0% and 1.1%), occurring predominantly in people taking SUs (14.5% and 15.4%). Improvements were maintained in people continuing exenatide OW for a further 22 weeks; those who switched from BD to OW exenatide saw additional HbA1c reductions. Glycaemic and weight benefits were maintained in people continuing treatment for 2 years, and SBP and lipid profiles significantly improved. Nausea frequency decreased over time, and was generally mild.
In the 24-week DURATION-5 study, mean HbA1c reductions were less than in DURATION-1, but consistent with previous studies (Buse et al, 2004; DeFronzo et al, 2005; Heine et al, 2005; Kendall et al, 2005; Bergenstal et al, 2010; Diamant et al, 2010).
Overall, in these trials, exenatide OW and liraglutide appeared to outperform exenatide BD in terms of glucose-lowering efficacy, and gastrointestinal tolerability was better with exenatide OW than with liraglutide or exenatide BD. One point of note is that in these trials, liraglutide was used at a dose of 1.8 mg, which is not currently recommended by NICE (NICE, 2010).
The development of antibodies during treatment can be associated with reduced efficacy (Berntorp et al, 2006) or adverse events (Jahn and Schneider, 2009). Exenatide appears to be more immunogenic than liraglutide, and the limited antibody response with liraglutide does not affect glycaemic efficacy or treatment safety (Buse et al, 2011).
Comparison of currently available DPP-4Is
Two trials have compared DPP-4Is: saxagliptin 5 mg OD versus sitagliptin 100 mg OD, both in addition to metformin (Scheen et al, 2010) and 5 mg saxagliptin OD versus 100 mg sitagliptin OD versus 50 mg vildagliptin BD, each in addition to metformin and another oral hypoglycemic agent (Li et al, 2014). Their efficacy and incidences of adverse events were similar (Table 1).
Comparisons of currently available GLP-1RAs and DPP-4Is
In a comparison of exenatide OW versus sitagliptin (Bergenstal et al, 2010; Table 1), reductions in HbA1c and weight loss were greater with exenatide OW. The most frequent adverse events with exenatide OW and sitagliptin were nausea (24% and 10%) and diarrhoea (18% and 10%, respectively). During a 26-week extension, people switched from sitagliptin to exenatide OW experienced significant further improvements in HbA1c (reductions of 3 mmol/mol [0.3%]), FPG (0.7 mmol/L) and weight (−1.1 kg; Wysham et al, 2011). In another 26-week trial (DURATION-4), exenatide OW led to significantly greater changes in HbA1c and weight compared with sitagliptin 100 mg (Russell-Jones et al, 2012).
There has been one large trial comparing liraglutide with a DPP-4I (Pratley et al, 2010; Table 1). Combined with metformin, liraglutide 1.2 mg or 1.8 mg led to greater changes in HbA1c, FPG and weight loss versus sitagliptin, which were maintained at 52 weeks (Pratley et al, 2011).
Patient-reported outcome comparison studies
Patient-reported outcome (PRO) data can provide useful insight into individuals’ experiences. A comparison of liraglutide and exenatide BD showed similar, high baseline treatment satisfaction (TS) for both (Schmidt et al, 2011). At 26 weeks, Diabetes Treatment Satisfaction Questionnaire (DTSQ) scores increased more with liraglutide compared with exenatide (4.71 versus 1.66 points; P<0.0001), and liraglutide-treated patients perceived a greater reduction in hypoglycaemia on the DTSQ change version (DTSQc), but not DTSQ status version (DTSQs; DTSQc: between-treatment difference=0.48; P=0.02), and hyperglycaemia (DTSQc: between-treatment difference=0.74; P=0.001). Significantly more people in the liraglutide group (91% versus 82%; P=0.02) expressed TS (DTSQs >24). During a 26-week extension, TS increased significantly (P=0.003 at week 40) in people switching to liraglutide.
In a patient subgroup from a liraglutide versus sitagliptin trial (Davies et al, 2011), overall TS improvement was greater with liraglutide 1.8 mg versus sitagliptin (P=0.03), and similar between sitagliptin and liraglutide 1.2 mg. In a 26-week extension, a greater increase in DTSQ score was reported with liraglutide 1.8 mg versus sitagliptin (P=0.03; Pratley et al, 2011).
Among sitagliptin-treated people switched to liraglutide 1.2 mg or 1.8 mg, overall DTSQ score increased compared with baseline (liraglutide 1.2 mg group, P=0.02; Pratley et al, 2012), although there was a transient increase in gastrointestinal reactions. In this trial, participants originally randomised to receive liraglutide continued unchanged.
A comparison of PRO outcomes for exenatide, sitagliptin and pioglitazone in the DURATION-2 study found that IWQOL-Lite (Impact of Weight on Quality of Life) scores had increased significantly in both the exenatide OW and sitagliptin treatment arms (both P<0.05) and there were no statistically significant differences between the exenatide QW and sitagliptin groups in total weight-related quality of life. General health utility, measured using EQ-5D (EuroQol 5 Dimensions measure), also increased in people in the exenatide OW and sitagliptin groups (P<0.05; Best et al, 2011). All groups experienced significant improvements on the psychological well-being global scale and all six domain scores. All groups experienced significant improvements in total diabetes treatment satisfaction scores and the exenatide OW group experienced greater improvement than the sitagliptin group in treatment satisfaction total scores. In this trial, nausea and vomiting reported with exenatide did not reduce patient satisfaction.
In the DURATION-1 study (Drucker et al, 2008), people in exenatide OW and BD groups showed similar improvements in treatment satisfaction and weight-related quality of life after 30 weeks of treatment (Best et al, 2009). However, DTSQ items related to “perceived frequency of hyperglycaemia” and “willingness to continue current treatment” were significantly higher in the exenatide OW group. Participants who switched from exenatide BD to exenatide OW after 30 weeks reported further improvements in treatment satisfaction.
Incretin-based therapies appear to be well received by patients, and these data discussed above suggest that the injectable administration route is not a barrier to GLP-1RAs, as is the case for insulin.
Dosage and administration
GLP-1RAs are administered by subcutaneous injection: liraglutide OD and exenatide OW at any time of day (eMC, 2014a,b), and exenatide BD and lixisenatide OD within 60 minutes before meals (eMC, 2014c,i). Exenatide OW requires mixing and syringe preparation using a single-dose kit, so healthcare professionals should ensure that individuals receive the appropriate education. The DPP-4Is are administered orally, either OD (saxagliptin, sitagliptin, linagliptin, alogliptin) or BD (vildagliptin). They do not need to be taken with food.
Therapeutic experience is limited in special patient groups for both GLP-1RAs and DPP-4Is. In people aged >75 years, no dose adjustment is recommended, although escalation with exenatide BD should be done cautiously. For linaglitpin, clinical experience for people >80 years is lacking so caution must also be exercised. None is recommended for patients aged <18 years (see Summary of Product Characteristics [SPC] for each product). Exenatide is cleared renally (Linnebjerg et al, 2007), and hence dose escalation should be performed prudently in people with moderate renal impairment. Exenatide BD and lixisenatide are not recommended in severe renal impairment (eMC, 2014b,i); exenatide OW is not recommended in moderate or severe renal impairment (eMC, 2014a). Liraglutide is not renally excreted but, due to limited therapeutic evidence, it currently can not be recommended for use in people with moderate or severe renal impairment in the UK (eMC, 2014b) but it is approved in the US to be used with caution at all stages of renal disease (Novo Nordisk, 2013). Exenatide BD, lixisenatide and liraglutide are indicated for use as adjunctive therapy to basal insulin (eMC 2014c,i,j).
With saxagliptin, sitagliptin, and vildagliptin, dosing adjustment is required in moderate and severe renal impairment, as these compounds are largely renally excreted and drug accumulation has been reported (see SPC for specific products). Linagliptin is mostly excreted non-renally and can be used at all stages of renal disease (eMC, 2014f). Therefore, DPP-4 inhibitors may be a more appropriate choice for patients with moderate or severe renal impairment. However, safety of liraglutide in this population has been demonstrated (Idorn et al, 2014; Umpierrez et al, 2014).
Safety and tolerability
Both classes of incretin therapy are generally well tolerated. Furthermore, due to their glucose-dependent mechanism of action, hypoglycaemia rates are low when these agents are not combined with insulin or an insulin secretagogue.
The most common side effects of GLP-1RAs are gastrointestinal. Nausea is more common with GLP-1RAs than DPP-4Is, but tends to be mild and transient (Bergenstal et al 2010; Pratley et al, 2010; Russell-Jones et al, 2012). Among GLP-1RAs, head-to-head studies found a lower incidence of nausea with liraglutide and lixisenatide (compared with exenatide BD; Buse et al 2009; Rosenstock et al, 2013) and exenatide OW (compared with exenatide BD or liraglutide; Drucker et al, 2008, Buse et al, 2013).
Concerns have been raised regarding the risk of pancreatitis with incretin therapies; however, an assessment by the European Medicines Agency (EMA) has concluded that assertions concerning a causal association between incretin-based drugs and pancreatitis or pancreatic cancer are inconsistent with the current clinical data (EMA, 2013). All SPCs for DDP-4Is and GLP-1RAs advise informing patients of the characteristic symptom of acute pancreatitis: persistent, severe abdominal pain and these agents are contraindicated in people with a history of pancreatitis. Both DPP-4Is and GLP-1RAs should be discontinued if pancreatitis is suspected.
Recently, the first few of several cardiovascular outcome trials required by regulatory agencies for DPP-4Is have reported results. In the EXAMINE trial, carried out in patients with type 2 diabetes who had a recent acute coronary syndrome, the rates of major adverse cardiovascular events were not increased with alogliptin compared with placebo (White et al, 2013). In the SAVOR-TIMI trial, people with type 2 diabetes who had a history of, or were at risk for, cardiovascular events found no change in the rate of ischemic events, but an increase in hospitalisation for heart failure with saxagliptin (Scirica et al, 2013). There are no data available as yet for ongoing trials with GLP-1RAs.
Practical considerations
Potential advantages of DPP-4Is are less frequent nausea and oral administration. Nausea with GLP-1RAs typically occurs early and can be lessened with an incremental dosing approach, where possible. Injecting at meal times and eating smaller meals may decrease nausea; returning to a lower GLP-1RA dose for a week and repeating incremental dosing is another option, where possible (Unger and Parkin, 2011; Novo Nordisk, 2013). Practical demonstration and patient education is important to allay injection fears. Although GLP-1RAs are more expensive than DPP-4Is in pure cost, cost-utility analyses make the situation more complex due to differences in efficacy, and GLP-1RAs have been found to be comparable or better in some health-economic analyses (Lage et al, 2009; Davies et al, 2012). GLP-1RAs may be a better choice when adding therapy to people close to target (King et al, 2013) due possibly to superphysiological levels of GLP-1 activation.
Conclusion
Differences between DPP-4Is and GLP-1RAs include efficacy, tolerability, mechanism of action, and administration route. GLP-1RAs yield higher levels of GLP-1 than DPP-4Is, which may account for increased anti-hyperglycaemic and weight benefits and increased GI side effects. Reductions in HbA1c and weight are generally greater with GLP-1RAs.
PRO data suggest that patient acceptance of injectables is not a major problem, and GLP-1RAs are actually associated with greater TS compared to other anti-diabetes drugs. Both types of incretin therapies provide a low risk of hypoglycaemia when not used in combination with insulin or insulin secretagogues. Nausea is common with GLP-1RAs, but tends to be transient and should be considered in the context of HbA1c improvements and weight loss. Treatments should be tailored to individual patients, based on their specific needs, comorbidities, and adverse effects with different therapies, and cost considerations should be taken into account.
Acknowledgements
The author is grateful to Jenna Steere of Watermeadow Medical for writing assistance in the development of this manuscript. This assistance was funded by Novo Nordisk, which also had a role in the review of the manuscript for scientific accuracy.
Author disclosure
The author has attended an advisory board in 2011 on behalf of Novo Nordisk A/S.
New handbook for healthcare professionals aims to improve outcomes
21 Nov 2024