GLP-1 receptor agonist and basal insulin co-therapy in type 2 diabetes: Clinical evidence and practicalities of use

Owing to the progressive nature of the condition, many people with type 2 diabetes will eventually require insulin therapy; however, insulin initiation is often substantially delayed in UK clinical practice (Blak et al, 2012a; Khunti et al, 2013). In those people who have started basal insulin therapy, many are not achieving adequate glycaemic control (Dale et al, 2010; Blak et al, 2012a; Blak et al, 2012b). Despite elevated HbA_1c levels, intensification of basal insulin therapy with meal-time or pre-mixed insulins appears infrequent (Blak et al, 2012b). Combining basal insulin and glucagon-like peptide-1 (GLP-1) receptor agonist (RA) therapies is a treatment approach that may help overcome the inertia with basal insulin initiation alone or insulin intensification. This article reviews the potential benefits of this treatment combination in people with type 2 diabetes and addresses the practicalities of implementing such an approach in primary care.

Brief overview of GLP-1 RAs
GLP-1 RAs improve glycaemic control by stimulating insulin secretion and inhibiting glucagon secretion in response to elevated glucose levels, such as post-meal hyperglycaemia (Baggio and Drucker, 2007). As a result of this glucose-dependent mechanism of action, rates of hypoglycaemia are very low with GLP-1 RA treatment, except when combined with sulphonylurea or insulin treatment (Nauck et al, 2009; Marre et al, 2009). Treatment with GLP-1 RA therapy is also generally associated with clinically significant weight loss, the extent of which varies with each agent (Vilsbøll et al, 2012). The most common side effect with GLP-1 RA therapy is transient gastrointestinal (GI) adverse events, although these usually subside within 12 weeks of treatment (MacConell et al, 2012) and with practical meal adjustments (Hicks et al, 2014).

There are currently five GLP-1 RAs available for use in the UK: exenatide once weekly (Bydureon®), exenatide twice daily (Byetta®), liraglutide (Victoza®), lixisenatide (Lyxumia®) and – the most recent addition – dulaglutide once weekly (Trulicity®). An overview of these agents is presented in Table 1.

Rationale for combining GLP-1 RAs and basal insulin
While the licences of some of the GLP-1 RAs may have been updated only relatively recently to include combination with basal insulin therapy, the concept is not new to the UK. In audits of both exenatide (2007) and liraglutide (2009) by the Association of British Clinical Diabetologists, the combination was extensively used off-licence in secondary care, with this being reported in almost 40% of people treated with a GLP-1 RA (Ryder and Thong, 2012). The most recent clinical guidelines for management of type 2 diabetes from NICE (2009) and SIGN (2010) do not include recommendations for this combination; the NICE guideline, which, at the time of writing, is undergoing an update, recommended further studies. However, in 2012, the American Diabetes Association and the European Association for the Study of Diabetes released a joint position statement recommending GLP-1 RA and basal insulin combination therapy for people not meeting glycaemic targets on dual therapy with either metformin and a GLP-1 RA or metformin and basal insulin (Inzucchi et al, 2012). This was reinforced in the recent update of this statement (Inzucchi et al, 2015).

Combining a basal insulin with a GLP-1 RA provides people with the complementary actions of the two products, as shown in a recently published systematic review and meta-analysis, which found that use of the combination in type 2 diabetes can help achieve adequate glycaemic control with no increased risk of hypoglycaemia or weight gain (Eng et al, 2014). While both classes of drug are effective at lowering blood glucose levels, basal insulin primarily acts on fasting plasma glucose (FPG), while short-acting GLP-1 RAs target postprandial glucose (PPG) and long-acting GLP-1 RAs have effects on both FPG and PPG (Drucker et al, 2008). It has been shown that fasting hyperglycaemia has the most impact on HbA_1c when diabetes is poorly controlled while postprandial hyperglycaemia is the major contributor in people nearing HbA_1c goals (Monnier et al, 2003). Additionally, insulin therapy often results in weight gain and an increased risk of hypoglycaemic events, while treatment with GLP-1 RAs can be associated with weight loss without increased risk of hypoglycaemia (Inzucchi et al, 2015).

Evidence to support the combination of basal insulin and GLP-1 RA therapy
There are two approaches to using a GLP-1 RA in combination with a basal insulin; adding a basal insulin to existing GLP-1 RA therapy or adding a GLP-1 RA to basal insulin therapy, both of which have been studied in phase III trials.

Adding a basal insulin to GLP-1 RA therapy
Two studies have demonstrated that adding basal insulin to GLP-1 RA therapy significantly reduces HbA_1c and with a low rate of hypoglycaemia (0.29 and 0.57 events per patient-year for liraglutide in combination with insulin detemir and insulin degludec, respectively; DeVries et al, 2012; Aroda et al, 2014). In this sequence, there was either no change in mean body weight or a small increase. Box 1 presents a case illustrating the rationale for the addition of a basal insulin to the treatment regimen of a person with sub-optimally controlled type 2 diabetes who is already taking a GLP-1 RA.

Adding a GLP-1 RA to basal insulin therapy
The addition of a GLP-1 RA to basal insulin therapy can significantly improve HbA_1c levels without increasing the risk of hypoglycaemia, and the addition of a GLP-1 RA to basal insulin is also commonly associated with clinically significant weight loss (Buse et al, 2011; Riddle et al, 2013a; Riddle et al, 2013b; Ahmann et al, 2014; Mathieu et al, 2014; Rosenstock et al, 2014a). See Box 2 for an example of how a GLP-1 RA may be added to basal insulin therapy.

Why is this combination particularly relevant in primary care?
Owing to the improvements in glycaemic control with a low risk of hypoglycaemia when adding a basal insulin to existing GLP-1 RA therapy (DeVries et al, 2012; Aroda et al, 2014), this approach presents primary care clinicians with a potentially attractive option for appropriate people with diabetes. Additionally, while intensifying basal insulin treatment by adding in a prandial or pre-mixed insulin often requires referral to secondary care, which may lead to increased costs and delays due to longer waiting times (Cuddihy et al, 2011), we feel that the addition of a GLP-1 RA to basal insulin is more suitable for a primary care setting.

Furthermore, intensification of basal insulin with a GLP-1 RA may be more acceptable to people with diabetes than the addition of prandial insulin injections owing to the more favourable weight profile and reduced likelihood of hypoglycaemic events with the GLP-1 RA (Mathieu et al, 2014; Rosenstock et al, 2014a). Basal insulin intensification with a GLP-1 RA has also been shown to be associated with reduced hospitalisation rates and lower all-cause costs compared with intensification using prandial insulin, in a US setting (Dalal et al, 2015).

Practical tips for using the combination
Adjusting prior medication doses
To reduce the risk of hypoglycaemia when adding a GLP-1 RA to basal insulin, a reduction in basal insulin dose should be considered (Table 1). In clinical studies, protocols have generally specified a reduction of around 20–25% of the prior dose in people with HbA_1c <69 mmol/mol (<8.5%) at initiation (Viswanathan et al, 2007; Buse et al, 2011). In addition – again, owing to the risk of hypoglycaemia – for people treated with a GLP-1 RA or basal insulin in combination with a sulphonylurea, withdrawal, or at least a dose reduction, of the sulphonylurea should be considered prior to intensifying therapy with a basal insulin or GLP-1 RA (see Table 1).

Timing of injections
Injections should be administered as recommended in each product’s prescribing information, with regards to timing and frequency. Basal insulin and GLP-1 RA injections can be administered at the same time and in the same general area of the body; however, we strongly recommend that they are not administered at the same injection site. Depending on which product is used, dosing may or may not be independent of meal-times (see Table 1 for GLP-1 dose timing), which should be factored into decisions about the chosen regimen.

Reducing GI side effects
With exenatide twice daily, liraglutide and lixisenatide, in order to reduce the risk of GI side effects following addition of the GLP-1 RA to basal insulin, treatment should be initiated at a reduced dose of the GLP-1 RA (see Table 1) and then gradually increased to the recommended dose (electronic Medicines Compendium [eMC], 2014a; 2014b; 2015c; 2015e). Additionally, in our experience, avoiding large or high-fat meals after injection can also help reduce nausea, especially early in treatment.

The future of combination therapy
Two once-daily, single-injection products that combine a GLP-1 RA and basal insulin at a fixed ratio have been developed. IDegLira (Xultophy®), which has a ratio of 1 unit of insulin degludec to 0.036 mg of liraglutide, is now available in the UK (eMC, 2014c). LixiLan, which combines insulin glargine and lixisenatide), is in phase III trials at the time of writing (ClinicalTrials.gov, 2015a; 2015b; Rosenstock et al, 2014b).

IDegLira is administered once daily at any time (preferably at the same time of the day) based on dose steps, with each step consisting of 1 unit of insulin and 0.036 mg of liraglutide. The recommended starting dose is 10 dose steps (10 units of insulin degludec and 0.36 mg of liraglutide) if added to oral antidiabetes therapy, or 16 dose steps if people were previously receiving a basal insulin (eMC, 2014c).

The dose is adjusted using a dial on the injection pen and is based on fasting self-measured blood glucose readings. We recommend increasing or decreasing the dose by two steps twice a week, with the aim of achieving individualised glycaemic targets, to a maximum of 50 dose steps (50 units insulin degludec and 1.8 mg of liraglutide).

Four phase III studies of IDegLira have been completed, two of which have been published in full (Buse et al, 2014; Gough et al, 2014). When added to oral antidiabetes therapy, these two studies have found reductions in HbA_1c (mean reduction of 21 mmol/mol [1.9%] in both trials) with IDegLira that were non-inferior to (Gough et al, 2014), or significantly greater than (Buse et al, 2014), that with insulin degludec, and that were significantly greater than that with liraglutide (Gough et al, 2014). Gough et al (2014) also reported that fewer participants in the IDegLira group reported gastrointestinal adverse events than in the liraglutide group (nausea, 8.8% versus 19.7%), which was probably attributable to the slower titration of the liraglutide component in IDegLira. The studies also found that IDegLira was weight neutral (reduction of 0.5 kg) when used in insulin-naïve people (Gough et al, 2014) and resulted in weight loss (reduction of 2.7 kg) in those previously treated with a basal insulin (Buse et al, 2014).

Further developments in basal insulins and GLP-1 RAs may also impact the way the combination is used in future.

Conclusions
The combination of a GLP-1 RA and a basal insulin is an efficacious treatment for people with type 2 diabetes. Co-use of these offers complementary effects on glycaemic control with potential advantages in terms of hypoglycaemia risk. The combination is also beneficial in terms of body weight, resulting in weight loss with the addition of a GLP-1 RA to basal insulin.

When initiating the combination, minor adjustments in prior treatment regimen can help reduce the risk of hypoglycaemia. While GI side effects are possible in the early stages of GLP-1 RA treatment, there are a number of ways to help minimise these effects, and these should be discussed with individuals.

Furthermore, once-daily, single-injection combinations of a basal insulin and a GLP-1 RA have been developed, with one of these already having been launched in the UK.

Acknowledgements
The authors identified the need for this topic and had editorial control over both the design and the content of this manuscript. They wish to thank Leah Pears and Gabrielle Parker at Watermeadow Medical for editorial assistance (funded by Novo Nordisk). The authors did not receive payment in relation to the article’s preparation or publication.