Glucagon-like peptide 1 deficiency in type 1 diabetes?

The introduction of glucagon-like peptide (GLP-1) analogues for the management of type 2 diabetes has, arguably, been one of the major pharmaceutical success stories of the last decade. These treatments jointly improve insulin secretion and aid weight loss – a desirable combination of effects not seen with other diabetes drugs, the majority of which cause weight gain. Three GLP-1 analogues are currently licensed in the UK (exenatide, liraglutide and lixisenatide), with approval expected for a number of others with extended pharmacokinetic profiles allowing weekly dosing. Pharmaceutical companies are hoping to expand the indications for GLP-1 analogues to include weight loss in obese people without diabetes, with positive results in a recent clinical trial lending a strong argument to this proposition. ¹

With the industry focus on the lucrative type 2 diabetes and obesity markets, a steady trickle of papers have, paradoxically, hinted at a possible role for GLP-1 in type 1 diabetes. At first sight this seems counterintuitive, as these patients are usually not overweight, and historically have been assumed not to retain any beta cell secretory function required for GLP-1’s insulinotropic effect. On the latter point, it is now accepted that most people with type 1 diabetes retain some functional beta cell mass for at least a year after diagnosis – although this eventually declines. ² Such patients, usually identified by the presence of detectable circulating C-peptide, have repeatedly been shown in short-term studies to benefit from incretin treatment, in terms of reduction in exogenous insulin dose and risk of hypoglycaemia.^3–5 Given that GLP-1 inhibits beta cell apoptosis, ⁶ it is conceivable that in addition to its direct insulinotropic effect, it might also extend the lifespan of remaining beta cells, although this remains to be tested. Perhaps more surprisingly, glycaemic benefits are also seen in ‘C-peptide negative’ patients, who should not be able to upregulate endogenous insulin secretion in response to GLP-1. Here, the observed effects are likely to reflect other actions of GLP-1, including more gradual glucose absorption due to delayed gastric emptying, and suppression of glucagon release. Longer term trials are underway which should confirm if short-term effects of GLP-1 analogues translate to longer term benefits.

The study by Zibar et al. ⁷ in this issue of the Annals adds an additional dimension to the potential importance of GLP-1 in type 1 diabetes. Building on the existing body of evidence for pharmacological GLP-1 in this patient group, the authors shift the focus to physiology, and find that patients with type 1 diabetes have significantly lower fasting GLP-1 concentrations than healthy controls. A ‘bihormonal hypothesis’ of diabetes pathophysiology has long been mooted, ⁸ in which lack of glucagon suppression contributes to the phenotype – might it be that lack of GLP-1 plays a role as well? This is of course speculation, but it does provide an additional rationale for trialling GLP-1 analogues in these patients. Interestingly, in this study, type 1 diabetes did not influence postprandial rises in GLP-1, unlike in the fasting state, and also unlike in type 2 diabetes. ⁹ As GLP-1 is generally considered to exert its main effects after a meal, this might argue against an important role for lack of GLP-1 in pathophysiology; on the other hand, non-acute effects of GLP-1 such as on beta cell survival could relate mainly to basal concentrations.

Explaining why people with type 1 diabetes have lower circulating GLP-1 concentrations is difficult. In this study, efforts were made to exclude age, gender and body mass index differences as confounders. It is suggested that exogenous insulin might acutely suppress GLP-1 release, as it does for another incretin peptide, glucose-dependent insulinotropic polypeptide, ¹⁰ but it is important to note that an inhibitory effect of insulin on GLP-1 release has never been conclusively demonstrated. Given that the patients in this study had an average diabetes duration of several years, is it possible that the development of diabetic gastroparesis and ensuing alterations in nutrient delivery to GLP-1-producing enteroendocrine cells could have led to a reduction in GLP-1 production capacity?

The authors also highlight a significant issue which continues to hinder comparison of different GLP-1 studies, namely the lack of standardised assays, and the historical reliance on ‘total GLP-1’ measurement, the majority of which is inactive breakdown products such as GLP-1 (9-36) amide. Here, both total and ‘active’ GLP-1 were measured, the latter referring to the physiologically important (7-36) amide form of the peptide.

In summary then, Zibar et al. ⁷ have identified an intriguing hormonal ‘deficiency state’ in type 1 diabetes. It remains to be seen why this is the case, and whether it plays a role in pathophysiology. What seems likely though is that over-replacing the deficient hormone with pharmacological doses of long-lasting analogues will be a fruitful therapeutic strategy in the future.