| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Editorial |
Department of Medicine M (Endocrinology & Diabetes), University Hospital of Aarhus; and The Institute of Pharmacology, University of Aarhus, DK-8000 Aarhus, Denmark
Address all correspondence and requests for reprints to: Ole Schmitz, Department of Medicine M (Endocrinology & Diabetes), University Hospital of Aarhus, University of Aarhus, DK-8000 Aarhus, Denmark. E-mail: ole.schmitz{at}ki.au.dk.
The pandemic explosion of type 2 diabetes incidence is well established, due to dramatic global changes in lifestyle over the past few decades and in the foreseeable future. To alter the course of this trend may take, at best, up to half a century. Until then, we need, in addition to lifestyle guidance, pharmacological solutions to reduce the markedly increasing incidence of macrovascular complications that, in its wake, confer a huge burden on patients and their families as well as an enormous economic burden to society. In the United States, for example, diabetes afflicts more than 21 million people, with an estimated cost of around $130 billion annually (http://diabetes.niddk.nih.gov/dm/pubs/statistics/index.htm#7). The Steno 2 study has clearly demonstrated the efficacy of aggressive intervention against dyslipidemia, hypertension, and elevated glycemia brought on by diabetes (1). In daily life, however, it is also clear that less than half of the type 2 diabetic patients reach the recommended glycemic level employing so-called standard therapy, i.e. sulfonylureas, metformin, thiazolidinediones, and insulin. Hence, all new efficient blood glucose-lowering compounds without serious side effects are greeted with open arms.
The glucagon like peptide-1 (GLP-1) concept holds promise. GLP-1 is mainly secreted by the intestinal L cells, although neural signals probably also contribute. The active peptides [GLP-1(7-37) and GLP-1(7-36) amide] increase insulin secretion in a glucose-dependent manner, i.e. lowering risk of hypoglycemia, decreasing basal as well as postprandial glucagon secretion, delaying gastric emptying, and increasing satiety by actions in the hypothalamus. Animal and in vitro models have shown that GLP-1 also increases β-cell proliferation and neogenesis and decreases apoptosis (2). In these models, GLP-1 has other extrapancreatic effects beyond those described above, but whether these effects are also reflected in humans remains to be seen.
In the circulation, GLP-1 undergoes enzymatic deactivation to GLP-1(9-37) and GLP-1(9-36) amide within minutes, primarily by dipeptidyl peptidase-4 (DPP-4), an enzyme present on the surface of lymphocytes, macrophages, and endothelial cells, and in tissues, such as the pancreas, liver, intestine, kidneys, and lungs. Localization of DDP-4 in endothelial cells indicates that degradation already takes place upon GLP-1 entry into the capillaries; endothelial DPP-4 in the liver degrades 50% of the protein. From this it can be deduced that only approximately 15% of the secreted GLP-1 reaches the pancreas in its intact form (3). DPP-4 also metabolizes other biologically active peptides of relevance in the regulation of carbohydrate metabolism such as glucose-dependent insulinotropic polypeptide, vasoactive intestinal polypeptide, and gastrin-releasing polypeptide. Active GLP-1 operates through the GLP-1 receptor (GLP-1R) belonging to the class B family of seven-transmembrane-spanning, heterotrimeric G protein-coupled receptors. Apart from the cells of the islet of Langerhans, several tissues express GLP-1R, including the heart, central nervous system, kidneys, lungs, stomach, intestine, and pituitary gland, as well as the nodose ganglion of abdominal vagal afferent nerve fibers, the central branches of which terminate in the brain stem. The presence of GLP-1Rs in skeletal muscle, liver, and adipose tissue is debatable.
Due to the short half-life of native GLP-1, it is not feasible for long-term use, but its parenteral use may be of interest during acute medical and/or surgical conditions (4). For long-term therapy, two classes of pharmaceutical agents are involved and are still in development: 1) long-acting analogs, or so-called GLP-1 mimetics; and 2) DPP-4 inhibitors, or so-called incretin enhancers.
GLP-1 mimetics
Based on evidence that binding the GLP-1 analog to albumin slows its absorption and elimination, such agents have proven successful in prolonging GLP-1 action. Liraglutide is partly a DPP-4 -resistant GLP-1 analog that contains an Arg34Lys substitution and a glutamic acid and 16-C free-fatty acid addition to Lys26. The half-life for this drug in type 2 diabetes patients is 10–12 h (5), making once daily sc administration sufficient. However, the interindividual variation was quite large, approximately 40%. Liraglutide reduces HbA1c by up to 1.7%, induces significant weight loss, suppresses circulating glucagon, and in in vitro and rodent studies, has beneficial effects on the β-cell similar to native GLP-1. Gastrointestinal side effects appear frequently but are usually transient. Liraglutide may be even more effective when combined with oral hypoglycemic agents. Regulatory filing for this drug is scheduled for 2010.
In the search for biologically active peptides, exenatide (synthetic exendin-4) was discovered in the gila monster’s venom (6) and has been approved both in the United States and Europe for patients with type 2 diabetes. Exenatide is 53% identical in its amino acid sequence to human GLP-1 and is relatively degradation-resistant to DPP4. Renal elimination of this drug seems also to be rather sluggish. The half-life of exenatide is 60–90 min, which results in more than a 1,000-fold in vivo potency compared with native GLP-1 (7). Several trials have demonstrated the efficacy of adding exenatide to other therapy in patients not optimally controlled by metformin, thiazolidinediones, or sulfonylureas. Exenatide is normally dosed twice daily. The reduction in HbA1c, and other beneficial effects, in line with gastrointestinal side effects are more or less similar to those of liraglutide. Hypoglycemic events are rare when exenatide is given as monotherapy or in combination with metformin or a thiazolidinedione. Interestingly, a long-acting GLP-1R agonist based upon exenatide has recently been developed: exenatide long-acting release, a polyactide-glycolide microsphere suspension containing 3% exendin-4-peptide. During a 15-wk study, once weekly sc administration of this compound led to a dose-dependent decrease in HbA1c of between 1.4 and 1.7%; 86% of type 2 diabetic patients achieved a HbA1c level of no greater than 7%, and there was a weight reduction of approximately 4 kg in those treated with the highest dose (2 mg exenatide LAR once weekly) (8). Consequently, great expectations have been invested in this slow-release formulation.
DPP4 inhibitors
Several pharmaceutical companies have developed or are developing specific DPP4 inhibitors that prolong the half-life of GLP-1, approximately doubling circulating levels of total and intact GLP-1. Dosing is oral, once or twice daily. Unlike GLP-1 mimetics, DPP4 inhibitors do not alter gastric emptying, and a decrease in weight is rarely seen. These effects are probably due, at least partly, to the relatively modest, only 2- to 3-fold increase in GLP-1 levels. The two DPP4 inhibitors, from which most clinical data have been provided, are vildagliptin and sitagliptin, the former approved in Europe and the latter both in the United States and Europe. Other DPP-4 inhibitors include saxagliptin and alogliptin (currently in phase III), and approximately another 10 have entered a clinical development program. The two approved compounds are well tolerated in combination with metformin and thiazolidinediones, decreasing HbA1c by an additional 1% (9). Finally, DPP-4 inhibition seems to modify β-cell mass in a beneficial manner similar to native GLP-1 and GLP-1 mimetics. From a clinical viewpoint, DDP-4 inhibitors may be candidates for first-line treatment, together with metformin. For more details the reader is referred to recent reviews (9, 10).
Ever since the initial and classic work of the administration of GLP-1 in humans by Gutniak et al. (11) using the "artificial pancreas," it has been elusive whether GLP-1 per se improves insulin sensitivity. In that study a direct effect was asserted, but differences in glucagon levels probably account for the decrease in glucose infusion during GLP-1 exposure. Although some in vitro studies seem to infer direct effects of GLP-1 and GLP-1 mimetics on peripheral tissues and the liver, the majority of human studies fail to demonstrate direct effects (12, 13, 14).
In this issue of JCEM, Azuma et al. (15) investigate the impact of the DPP-4 inhibitor, vildagliptin, in a randomized, double-blinded trial on islet function and glucose metabolism in type 2 diabetic subjects. In essence, 6 wk of vildagliptin treatment improved insulin action as assessed by a two-step hyperinsulinemic clamp. The glucose responsiveness of insulin secretion and postprandial circulating glucagon secretion was improved or reduced by 50 and 16%, respectively. As expected, glycemic control was further enhanced substantially during vildagliptin administration compared with placebo despite rather good glycemic control before the study. However, these beneficial metabolic effects can all be explained by less glucotoxicity and lipotoxicity. Hence, the study gives no evidence for a direct clinical effect on glucose metabolism. One can, of course, always hypothesize, but data controlling for a difference in glycemia, and consequently, in glucotoxicity, are eagerly awaited.
This same mechanism of effect uncertainty applies for the GLP-1 mimetics. One option is treating the placebo group to a similar glycemic level by for example insulin, or evaluating metabolism after a shorter off-drug period. Alternatively, important information on effects of GLP-1, GLP-1 mimetics, and DPP-4 inhibitors beyond the GLP-R may be provided through studies in the dual incretin receptor knockout (DIRKO) mice. Recent data have revealed that long-term administration of vildagliptin to this mouse reduced expression of genes relevant for aspects of hepatic lipid metabolism. In contrast, vildagliptin did not induce changes in glucose homeostasis (16). Furthermore, using the clamp technique, the same group has shown in DIRKO mice that incretin receptor signaling during certain circumstances (e.g. high-fat feeding) can influence insulin action independent of insulin stimulation (17).
Space does not allow discussion on all extrapancreatic effects, and the reader is referred elsewhere (e.g. Ref. 18). However, in brief, GLP-1 infusion has shown to diminish the size of myocardial infarction, not only in animal models but also in humans (19). We do not know whether the effect in humans is directly due to stimulation of the myocardial GLP-1R or secondary to an improved metabolic milieu. GLP-1Rs are widespread in the brain, and GLP-1 has been shown to be involved in learning and memory (20). Using the pancreatic pituitary clamp, Lerche et al. (21) recently demonstrated that GLP-1 administration in healthy humans inhibits blood-brain glucose transfer in almost all brain areas, using 18-flouro-deoxy-glucose PET scanning. However, GLP-1 also diminished glucose metabolism, suggesting that GLP-1Rs are involved in neuroprotection.
Another elusive topic is the preservation of the β-cell function in humans after long-term GLP-1R activation. The literature is rather limited, but in a group at high risk for developing diabetes, i.e. subjects with impaired fasting glucose, 6 wk of vildagliptin treatment increased the acute insulin response and insulin sensitivity evaluated by the frequently sampled iv glucose tolerance test. Unfortunately, the beneficial effect was not sustained after a 2-wk washout (22). This disappointing observation could be due to either the short treatment period or simply that the beneficial β-cell alterations seen in rodents do not occur in humans. Longer term studies are urgently needed, including treatment with GLP-1 mimetics.
When introducing novel drugs, unknown side effects are always a cardinal issue. They will often only be revealed after a huge amount of patient years on drug. The widespread expression of DPP-4 in numerous cells is a matter for concern, especially nonselectivity related to the enzymes DPP-8 and DPP-9 could be deleterious. However, both vildagliptin and sitagliptin show high selectivity for DPP-4 over other enzymes within this family.
The GLP-1 concept is clearly of great interest and will no doubt improve metabolic control in type 2 diabetic patients. However, the pharmaceutical industry is searching for novel compounds so that we can tailor medication to individuals by taking advantage of the enormous progress in genetic research. Last, but not least, it is important not to dismiss the classic oral medications, such as sulfonylureas and metformin, for which almost half a century of surveillance exists. In terms of metformin, we have also recently gained further insight into the pharmacogenetics behind the metformin response (23).
Acknowledgments
The secretarial assistance of Linda Edge is highly appreciated.
Footnotes
Abbreviations: DPP-4, Dipeptidyl peptidase-4; GLP-1, glucagon like peptide-1; GLP-1R, GLP-1 receptor.
Received December 13, 2007.
Accepted December 26, 2007.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
