Evidence of the Month

Commentaries on both new and classic studies of importance for the treatment of diabetes are posted here monthly.

Mathijs C. Bunck and Michaela Diamant Department of Endocrinology / Diabetes Center VU University Medical Center Amsterdam, the Netherlands	Click here to download PowerPoint slides of this commentary.
	Click here for the programme evaluation and post-test.

Background
Type 2 diabetes is characterized by progressive pancreatic β-cell failure against a background of obesity-related insulin resistance. Data from the United Kingdom Prospective Diabetes Study (UKPDS) indicate that, irrespective of the therapy used, glycaemic control worsens over the years due to progressive loss of β-cell function. Interventions that could improve or even those that could halt or slow down the decline of β-cell function are expectantly awaited. Based on preclinical data, incretin-based treatments, such as glucagon-like peptide-1 (GLP-1) receptor agonists and inhibitors of the GLP-1–degrading enzyme dipeptidyl peptidase-4 (DPP-4), have now emerged as a promising therapeutic option that could indeed improve β-cell function in humans. In the study by Heine et al, the efficacy of a 26-week treatment with the GLP-1 receptor agonist exenatide was compared with that of insulin glargine in people with Type 2 diabetes failing on oral blood glucose–lowering agents.

Methods and Key Results
A total of 844 people with Type 2 diabetes were screened in 13 countries comprising 82 study sites for this 26-week, open-label, randomized, controlled trial (Clinical Trials.gov number: NCT00082381). Individuals, aged 30 to 75 years, with HbA_1c levels ranging from 7 % to 10 %, stable body weight for 3 months, and body mass index between 25 and 45 kg/m² who were in suboptimal glycaemic control while using stable doses of metformin and sulfonylurea for at least 3 months prior to the screening were eligible. Finally, 551 individuals were randomized, of whom 535 (97%; exenatide 10 µg twice daily, n=275, and insulin glargine, n=260) had at least one post-baseline HbA_1c sample and therefore were included in the intention-to-treat population. The study’s primary objective was to test the hypothesis that exenatide was ‘noninferior’ to insulin glargine in achieving glycaemic control after 26 weeks of treatment. A priori, the non-inferiority margin for the difference between treatments was defined as 0.4%.

At 26 weeks, both treatments similarly reduced HbA_1c from baseline (exenatide baseline, 8.2 % ±1.0; insulin glargine baseline, 8.3 %±1.0) by 1.11%, resulting in a between-group difference of 0.017% (95% CI:
–0.123 to 0.157%), thus achieving the ‘noninferiority’ primary endpoint. In both treatment arms, a similar proportion of people with HbA_1c of 7 % at baseline achieved the target HbA_1c goal of ≤7.0 % (46% for exenatide and 48% for insulin glargine; P=ns). During the study, individuals randomized to insulin glargine were instructed to adjust their insulin dose according to their self-monitored fasting blood glucose levels. In short, people had to increase their insulin dose by 2 IU if their blood glucose level was ≥5.6 mmol/l (≥100 mg/dl) on 3 consecutive days, starting with a baseline insulin dose of 10 IU daily. Using this algorithm, insulin glargine reduced fasting plasma glucose (FPG) more than exenatide
(–2.9 mmol/l [–51.5 mg/dl] and –1.4 mmol/l [–25.7 mg/dl], respectively; P<0.001). However, still only 21.6% of those treated with glargine reached the treat-to-target FPG goal of <5.6 mmol/l (<100 mg/dl), as compared with 8.6% in the exenatide group (P<0.001). At study termination, the mean insulin dose used was 25 IU a day. Baseline mean body weight was 87.5±16.9 kg and 88.3±17.9 kg for the exenatide- and insulin glargine–treated arms, respectively. At 26 weeks, exenatide reduced body weight by 2.3 kg whereas insulin glargine increased body weight by 1.8 kg, resulting in a between-group difference of –4.1 kg (95% CI: –4.6 to –3.5 kg; P<0.0001).

The most frequently reported adverse events in the exenatide group were nausea (57.1%) and vomiting (17.4%), mostly of mild to moderate intensity, but only 13% reported nausea during the last 8 weeks of the study; in the insulin glargine group, 8.6% and 3.7% reported nausea and vomiting, respectively (P<0.001 vs exenatide). The reported reduction in body weight in the exenatide group appeared to be independent of nausea or vomiting. The overall rate of hypoglycaemia was similar in the treatment groups, but exenatide-treated individuals reported fewer episodes of nocturnal hypoglycaemia than individuals treated with insulin glargine. Conversely, daytime episodes of hypoglycaemia were more frequent in the exenatide-treated group. None of the randomized individuals withdrew due to hypoglycaemia. Finally, 43% of those treated with exenatide tested positive for anti-exenatide antibodies. Nevertheless, mean reduction in HbA_1c was not affected by the presence of antibodies.

Clinical Implications
Type 2 diabetes is characterized by a progressive decline in pancreatic β-cell functioning, in the presence of reduced β-cell mass. Pharmacological interventions that slow down or even reverse this decline are of great interest in the treatment of Type 2 diabetes. Exenatide has been shown to improve β-cell function and reduce β-cell apoptosis in preclinical studies (Baggio and Drucker, 2006).

In the present 26-week study, exenatide showed non-inferiority relative to the current standard of insulin therapy in Type 2 diabetes, that is, insulin glargine, with respect to overall glycaemic control. Both treatments resulted in an equal reduction in HbA_1c; however, exenatide reduced postprandial glucose excursions whereas insulin glargine primarily decreased FPG. Unfortunately, only 21.6% of glargine-treated individuals reached the treat-to-target FPG goal of <5.6 mmol/l (<100 mg/dl), and this may be explained by a relative low mean insulin dose at study termination due to the concomitant use of sulfonylurea and the associated risk for hypoglycaemia that could have prevented further dose titration. Thus, if the insulin glargine dose could have been increased more aggressively, a greater HbA_1c reduction could have been achieved in the insulin group, as shown in a recent study by Yki-Jarvinen et al. However, since in patients with baseline HbA_1c in a similar range as in this study, fasting and postprandial glucose contribute equally to HbA_1c level (Monnier et al, 2003), one may speculate that even if a higher insulin dose were used, exenatide may have been equally effective.

A subset of people was given a standardized meal tolerance test at baseline and after 26 weeks of treatment. The 4-hour post-meal insulin response did not differ between the exenatide (n=41) and insulin glargine (n=37) groups. However, the 4-hour post-meal plasma glucose response was significantly lower in the exenatide-treated group compared with the insulin glargine group. As insulin secretion must be considered in the context of ambient glucose concentrations (Kahn, 2003), one could argue that although it was not an endpoint of the study, β-cell function seems to have improved in this short-term study.

If long-term studies could indeed demonstrate durable beneficial effects of exenatide on β-cell function preservation, then the timing of initiating incretin-based therapies in the early phase of the disease, that is, when β-cell mass and function can still be salvaged, needs to be seriously considered. Studies that could dramatically improve our outlook on Type 2 diabetes as a progressive disease are at present eagerly awaited.

This Website Feature is funded by an unrestricted educational grant from Pfizer Inc.