Since our previous review of pharmacotherapy to treat diabesity (Day and Bailey, 2012), the epidemic of coexistent type 2 diabetes with obesity has continued to escalate (International Diabetes Federation, 2013). An individualised management strategy is advocated with lifestyle (mainly diet and exercise) advice initially and throughout. This is supplemented with pharmacotherapy to assist blood glucose control, ideally without weight gain and preferably with weight loss, as well as treatments for comorbidities and to reduce cardiovascular risk (Rajeswaran et al, 2012). Metformin is usually the preferred initial blood glucose-lowering agent as it counters insulin resistance without weight gain and without causing hypoglycaemia (Bailey et al, 2007). Further agents with complementary actions are then added as required (Figure 1), aiming to treat both the hyperglycaemia and adiposity (NICE, 2009; Bailey, 2011; Inzucchi et al, 2012; Day, 2013).
In the last 3 years, several new pharmacotherapies have become available to address diabesity. This update considers the role of these therapies to intervene against the diversity of diabesity presentations.
Glucagon-like peptide-1 receptor agonists
Glucagon-like peptide-1 (GLP-1) receptor agonists comprise a class of peptides that bind and activate receptors for the natural incretin hormone GLP-1. They potentiate insulin secretion and reduce glucagon secretion in a glucose-dependent manner, lowering blood glucose particularly after meals and carrying negligible risk of interprandial or fasting hypoglycaemia. Since GLP-1 receptor agonists delay gastric emptying they may cause temporary nausea in some individuals. This class of agents also exerts satiety effects that facilitate weight loss (Drucker and Nauck, 2006; Meier, 2012; Campbell and Drucker, 2013).
Table 1 lists the available and pre-launch GLP-1 receptor agonists (at the time of writing); they are variously administered twice daily, once daily or once weekly by subcutaneous injection, depending upon their pharmacokinetic properties and formulations. Various published meta-analyses have suggested subtle differences in the glucose-lowering and weight-lowering efficacy of members of the class, although it is difficult to fairly account for variability in patient populations, baseline characteristics, trial designs and data analyses (Amori et al, 2007; Monami et al, 2009; Aroda et al, 2012). In randomised trials, the average reduction of HbA1c is typically in the range of 9–16 mmol/mol (0.8–1.5%), occasionally greater, and the weight-lowering effect is mostly about 2–3 kg (e.g. Aroda et al, 2012).
A recent audit of “real life” use of exenatide twice daily and liraglutide in the UK has confirmed the results seen in clinical trials and noted that while a majority of people respond to these agents, they may have little or no effect in some individuals (Thong et al, 2014). Indeed, NICE (2009) recommends that treatment with these agents should only be continued beyond 6 months if HbA1c is reduced by ≥11 mmol/mol (≥1%) and there is ≥3% weight loss. Since there is no clear marker to predict who will respond, as with most glucose-lowering treatments, it is appropriate to “try and see” on an individual basis. A finger-stick measure of glucose control after 2–4 weeks can be a useful indicator of longer-term responsiveness. We suggest that, even if treatment with a GLP-1 receptor agonist is not lowering both glucose and weight to the extent required by NICE, the agent may still be preventing increases in these factors (or improving one of them). Furthermore, discontinuation of the agent may leave insulin as the only option, because such individuals have usually already exhausted suitable oral agents. Advice from NICE (2009; 2010; 2012; 2013) to generally restrict use of a GLP-1 receptor agonist to a BMI ≥35 kg/m2 in individuals of European descent (with adjustment for other ethnic groups) can, we should remember, be over-ridden by clinical need.
To date, responders have shown good durability of glucose control and weight reduction during exenatide once-weekly treatment for up to 6 years (Diamant et al, 2014; Henry et al, 2014). Although pre-clinical data suggest that this and other GLP-1 receptor agonists can improve beta-cell mass (Tahrani et al, 2011), there is, as yet, no obvious evidence in human type 2 diabetes that the progressive decline in beta-cell function can be reversed.
Nausea remains an initial tolerability factor for some individuals receiving a GLP-1 receptor agonist, but this is mostly temporary (electronic Medicines Compendium [eMC], 2014b; 2015a; 2015b; 2015c; 2015d). Questions concerning a possible association between GLP-1 receptor agonists and acute pancreatitis have been asked in recent years. Present data from registration trials, insurance databases, audits and post-marketing studies have not indicated any substantive increase in risk, but careful monitoring is ongoing (Egan et al, 2014). Accordingly, regulators have advised caution in patients with a history of pancreatitis and discontinuation if pancreatitis is suspected or confirmed (European Medicines Agency [EMA], 2013; Medicines and Healthcare products Regulatory Agency, 2013; US Food and Drug Administration, 2013).
Fixed-ratio combinations
Emerging clinical experience and controlled studies have noted the benefits of combination therapy using a GLP-1 receptor agonist together with a basal insulin (Holst and Vilsbøll, 2013). To simplify the administration of two such agents into a single injection, a fixed-ratio combination, IDegLira (Xultophy®), has recently been launched in the UK (eMC, 2014c). IDegLira combines 50 units of the long-acting insulin analogue degludec with 1.8 mg of the GLP-1 receptor agonist liraglutide as the maximum recommended daily dose. This is titrated up in the same way as the titration of insulin alone; thus, for every 1 unit of insulin injected, the individual also receives 0.036 mg liraglutide at the same time. In a 1-year randomised clinical trial with obese and overweight people with type 2 diabetes inadequately controlled on metformin with or without pioglitazone (the DUAL-I trial), IDegLira reduced HbA1c from a baseline of 68 mmol/mol (8.3%) by 21 mmol/mol (1.8%) without incurring the weight gain seen with insulin alone (Gough et al, 2014a; 2014b). The glucose-lowering effect was greater than achieved with a higher dose of insulin degludec (IDeg) alone, indicating an insulin-sparing effect of IDegLira (Table 2), and there were about one-third fewer recorded episodes of hypoglycaemia with IDegLira than IDeg alone (Gough et al, 2014a). Other fixed-ratio combinations of a basal insulin with a GLP-1 receptor agonist (such as insulin glargine with lixisenatide – LixiLan) are in development (Rosenstock et al, 2014).
Dipeptidyl peptidase-4 inhibitors
The enzyme dipeptidyl peptidase-4 (DPP-4) degrades endogenous incretin hormones such as GLP-1; thus, DPP-4 inhibitors increase the circulating levels of endogenous incretin hormones by reducing their rate of loss from the circulation.
Five DPP-4 inhibitors are available in the UK at the time of writing (Table 3). Meta-analyses of controlled trial (Esposito et al, 2014) and “real life” databases indicate similar glucose-lowering efficacy across the class, with decreases in HbA1c by about 7–11 mmol/mol (0.6–1.0%). DPP-4 inhibitors are weight neutral and carry a very low risk of frank hypoglycaemia, due to the glucose-dependent nature of the incretin effect (Deacon, 2011). Although there are differences in pharmacokinetic properties, the only practical differences noted by most prescribers are that they are all once daily except vildagliptin (twice daily), and all require dose reduction with impaired renal function except linagliptin.
Although DPP-4 inhibition is known to affect the half-lives of various hormones and signalling peptides beyond the incretin system (Flatt et al, 2008), extensive use of DPP-4 inhibitors has raised minimal safety issues (Goossen and Graber, 2012). Post-marketing cardiovascular safety studies have recently been reported for saxagliptin and alogliptin, noting a small increase in the number of individuals with heart failure who were hospitalised, and this is receiving further investigation (Scirica et al, 2013; White et al, 2013). There were also more reports of acute pancreatitis in the active arms than the placebo arms, but the absolute numbers were small and large database analyses have not confirmed these issues (Patil et al, 2012).
Several fixed-dose combinations of DPP-4 inhibitors with metformin are available (see Table 4), and these provide a convenient opportunity to intensify treatment without increasing the pill burden (Bailey and Day, 2009). For the future, once-weekly DPP-4 inhibitors are in development (Biftu et al, 2014).
Sodium–glucose cotransporter 2 inhibitors
There are now three sodium–glucose cotransporter 2 (SGLT2) inhibitors available in the UK (Table 5). These agents competitively inhibit the reabsorption of filtered glucose from the proximal tubules, eliminating up to about 100 g glucose daily in the urine (Bailey and Day, 2010). Because the mode of action is independent of insulin secretion or action, the glucose-lowering efficacy of these agents is seen at any stage during the natural history of type 2 diabetes provided that the individual has adequate renal function.
The effectiveness of SGLT2 inhibitors to generate glycosuria is dependent on adequate renal filtration indicated by an estimated glomerular filtration rate (eGFR) >60 mL/min/1.73 m2 for maximal dosing. Dosage reduction is required for canagliflozin and empagliflozin down to eGFR 45 mL/min/1.73 m2. There are no data to indicate renal damage with these agents (Bailey and Day, 2014).
Clinical trials, some up to 4 years in duration, have consistently shown a reduction of HbA1c in the range of 6–11 mmol/mol (0.5–1%) as monotherapy or add-on to other glucose-lowering treatments including insulin (Tahrani et al, 2013). The renal elimination of glucose also assists weight loss (typically 2–3 kg) and may contribute to a small reduction in blood pressure (similar to a low-dose thiazide diuretic). By reducing glucotoxicity and lowering body weight, SGLT2 inhibitors can reduce the dose requirement for other drugs including insulin (Tahrani et al, 2013). The glucosuria increases the risk of genital and urinary tract infections, but evidence to date indicates that risk can be limited with appropriate patient education and managed by standard procedures. While analyses of cardiovascular events during pre-registration clinical trials have been well within the requirements of regulatory agencies, long-term post-marketing cardiovascular safety studies are in progress.
SGLT2 inhibitors are metabolised by a uridine system. This avoids interactions with most commonly used drugs including other classes of glucose-lowering agents (Bailey and Day, 2014), and fixed-dose combinations of SGLT2 inhibitors with metformin are licensed (see Table 4).
Anti-obesity agents
Although weight reduction is a key part of the treatment strategy for diabesity, there is limited access in Europe to pharmacotherapies that are primarily directed against adiposity. Orlistat is the only agent specifically indicated for weight loss in the UK, and this has been shown to modestly assist glycaemic control in people with type 2 diabetes (Yanovski and Yanovski, 2013). Several additional anti-obesity agents have been approved in the US, and others are in advanced development (Scheen and Van Gaal, 2014). Of note, a single daily subcutaneous injection of 3 mg liraglutide is, at the time of writing, under review by European regulators as a treatment for obesity (EMA, 2015). A 1-year study (the SCALE™ trial) in obese individuals (baseline weight 106 kg, BMI 37 kg/m2) with type 2 diabetes showed a 5.9% weight loss with liraglutide compared with 2% for placebo. Half of the liraglutide group achieved >5% weight loss, and 22% achieved >10% weight loss. The weight loss in the liraglutide group was accompanied by a reduction of HbA1c by 14 mmol/mol (1.3%) versus 3 mmol/mol (0.3%) with placebo from a baseline HbA1c of 63 mmol/mol (7.9% [Davies et al, 2014]).
Conclusions
Patient-centred “personalised” care for diabesity remains founded on appropriate lifestyle advice to address adiposity and hyperglycaemia, with attention to cardiovascular risk and treatment of comorbidities. Pharmacotherapy to improve glycaemic control without weight gain typically begins with metformin followed by add-on of a DPP-4 inhibitor, SGLT2 inhibitor or GLP-1 receptor agonist as appropriate. Alternative approaches not discussed in this review may be considered for morbidly obese people, notably bariatric surgery and intensive dietary interventions.
Among agents primary directed to reduce blood glucose, the SGLT2 inhibitors and GLP-1 receptor agonists facilitate weight loss, while the DPP-4 inhibitors are generally weight neutral or produce very modest weight reduction. It is emphasised that early intervention is important to gain the long-term benefits of glycaemic control and to address the plethora of debilitating complications that accompany protracted obesity.
