Safety of GLP-1 receptor agonists
GLP-1 receptor agonists are associated with dose-dependent gastrointestinal side effects including nausea, vomiting and diarrhoea, and 5-10% of patients may choose to abandon therapy for these reasons. These effects diminish with time, however, and overall patient satisfaction is high. Since stimulation of insulin secretion is glucose-dependent, hypoglycaemia is uncommon. Antibodies may form against the analogues. The main safety concerns relate to the pleiotropic actions of GLP-1 when this is delivered in pharmacological doses, since it has the potential to interact with receptors which subserve functions unrelated to glucose metabolism. Heart rate increases by 2-4 beats per minute, and might potentially have adverse long-term consequences. Subclinical rises of pancreatic enzymes are commonly observed, and acute pancreatitis has been reported in association with both liraglutide and exenatide, although observational studies have to date largely been negative, with one recent exception. Acute renal failure sometimes requiring dialysis/transplantation has been reported. Thyroid C-cell cancers develop in rodents, and the exocrine pancreas increases in weight and displays other morphological changes. A signal for carcinoma of the pancreas has been reported in two regulatory databases.
GLP-1 receptors are widely distributed, being for example present in pancreatic islets, pancreatic exocrine tissue, blood vessels, heart, kidneys, bone and thyroid. They subserve a wide variety of physiological functions at these sites unrelated to their effects on insulin and glucagon secretion, gastric emptying and satiety.
Furthermore, in the physiological situation GLP-1 has a very short-lived interaction with its receptor owing to its short half-life. Long acting agonists, in contrast, are present at pharmacological levels for hours at a time, and the consequences of prolonged high dose stimulation are not well understood. An additional complication is that effects described in animal models have an uncertain relevance to the human situation. Despite this uncertainty, safety reviews to date have been reassuring.
In this overview, the potential adverse consequences of GLP-1 agonist administration will be considered under the following heads: - Metabolic effects - Gastrointestinal effects - Antibodies - Cardiovascular effects - Renal effects - Effects on the exocrine pancreas - Effects on the thyroid - GLP-1 agonists and cancer
One of the advantages of GLP-1 agonist administration is that insulin secretion is glucose-dependent, and is inhibited at low glucose levels. This is an important safety feature since it means that, in contrast to injected insulin or sulfonylureas, insulin is no longer produced at low glucose levels. Use of the GLP-1 agonists in combination therapy may however aggravate the hypoglycaemic potential of the partner therapy.
Delayed gastric emptying can cause discomfort, nausea and vomiting; diarrhoea may also occur. Although these effects tend to diminish with time, and most patients find them tolerable, severe vomiting with dehydration may occur and precipitate pre-existing circulatory or renal disorders.
Antibody formation to therapeutic peptides is common. Recently an analysis characterizing the time-course and cross-reactivity of anti-exenatide antibodies and potential effects on efficacy and safety was published. No obvious correlation between change in HbA1c and titre was observed for exenatide twice daily, although mean reductions were attenuated in the small subset of patients (5%) with higher titres1. A significant correlation have been observed for exenatide once-weekly with no difference between antibody-negative and low-titre patients, but an attenuated mean reduction in the subset of patients (12%) with higher titres. Thus, low-titre anti-exenatide antibodies were common with exenatide treatment (32% exenatide twice-daily, 45% exenatide once-weekly), but had no apparent effect on efficacy1. Among liraglutide-treated patients only 4 to 10% developed antibodies (low titres) and no correlation to impaired efficacy have been observed. The impact of auto-antibodies on efficacy and safety in the longer term remains to be established.
An increase in heart rate (by 2-4 beats per minute) has been reported during treatment with liraglutide and exenatide. Even a small increase in heart rate accompanying a decrease in blood pressure is, however, potentially troubling, as an increased heart rate is an independent risk factor for cardiac mortality. The mechanism behind the change in heart rate is not known, but might involve increased natriuresis and lowered blood pressure. In one study, patients with obesity but without diabetes were treated with liraglutide and an increase was detected in the heart rate for only the first 30 weeks of treatment. The patients’ heart rates subsequently returned to basal levels. Whether the benefit of the decrease in blood pressure outweighs the harm of the increase in heart rate remains to be determined. Several large cardiovascular outcome trials (LEADER (liraglutide), EXSCEL (exenatide once-weekly), ELIXA (lixisenatide), REWIND (dulaglutide)) including up to 9,500 patients with type 2 diabetes are ongoing and are expected to be completed between 2016 and 20195.
There have been a number of reports of renal impairment or renal failure following administration of exenatide or liraglutide. One likely mechanism is that GLP-1 agonists delay gastric emptying, and nausea and vomiting are well-recognised side effects. Nausea and vomiting, if prolonged, result in dehydration. Severe dehydration leads to falling blood pressure, reduced perfusion of the kidneys, and may cause acute renal failure. Those with compromised renal function, whether because of pre-existing vascular disease or because of a reduced number of functioning nephrons, are less able to compensate for acute dehydration, and are therefore at increased risk of acute renal failure.
On 11/02/2009 FDA approved revisions to the drug label for exenatide. It noted that 78 cases of altered kidney function (62 cases of acute renal failure (ARF) and 16 of renal insufficiency) had been reported by the company. The data summary indicates that cases of ARF/renal insufficiency occurred from 3 days - 2 years after initiation of exenatide, patient ages ranged from 23-83 years (mean 60 years), 74/78 had at least one contributory factor (concomitant medical conditions or drugs), 42 reported symptoms associated with volume depletion, e.g. diarrhoea or vomiting, 71/78 required hospitalization, with 4 deaths. Eighteen required renal dialysis and 2 received kidney transplants. Of those requiring dialysis, 6 had no previous history of renal disease, 2 had reduced kidney function, and there was no information on 10. Exenatide was discontinued in 63/78 patients, with reported benefit in 50%.
Effects on the Exocrine Pancreas
Both exenatide and liraglutide have been associated with acute pancreatitis. Acute pancreatitis is an elusive entity to analyze in administrative databases, however, due to differences in diagnostic tests and diagnostic criteria. Another concern is that the incidence of acute pancreatitis is increased in the background type 2 diabetic population, possibly because of the coexistence of occult exocrine pancreatic abnormalities in the diabetic population, or possibly because of a high prevalence of obesity and gallstones.
Most studies in administrative databases have not confirmed an increased incidence of acute pancreatitis with GLP-1-agonist therapy, but these lacked the statistical power to exclude a smaller effect (e.g. up to 2-fold increase). A more recent larger and independent study found an approximate 2-fold increase for exenatide and sitagliptin. Adverse event reporting to the FDA is consistent with an increased rate of acute pancreatitis with all GLP-1 based therapies. Further more definitive studies are awaited.
Structural changes in the human pancreas
Increases in pancreatic weight, presumably mainly due to overgrowth of exocrine tissue, have been reported in some rodent models of diabetes. It should be noted that such examinations are somewhat operator dependent, given the small size of the pancreas, problems with accurate dissection and drying of the specimen that may occur if it is not weighed promptly.
A report of 8 human pancreases obtained post-mortem via the nPOD donor scheme in the USA showed considerable pancreatic enlargement in some instances, with a 40% increase overall. Only one of these had been treated with exenatide; the remainder were on sitagliptin. There was also overgrowth of endocrine cells, with alpha cell hyperplasia and proliferation of presumably immature insulin-containing cells. It has been suggested that alpha cell hyperplasia may be a consequence of glucagon inhibition.
Carcinoma of the Pancreas
GLP-1 receptors are abundantly expressed in the exocrine pancreas, and increased pancreatic weight has been observed, consistent with a trophic effect upon duct cells. A model has been proposed whereby proliferation of duct cells leads to localized duct occlusion and low-grade pancreatic inflammation, more typically manifest by subclinical increases in pancreatic enzymes, and more rarely in severe acute pancreatitis. It should however be noted that, although subclinical increases in pancreatic enzyme levels are regularly seen in those on GLP-1 based therapies, their significance is unknown.
The model further proposes that low grade inflammation and high levels of GLP-1 activity will predispose to the development of pancreatic cancer. Precancerous changes known as pancreatic intraepithelial neoplasia (PanIN) lesions precede the onset of pancreatic cancer, and are frequently present in the pancreas of middle-aged and elderly people (see Pancreatic cancer) and both PanIN lesions and pancreatic adenocarcinoma may carry the GLP-1 receptor. An excess of pancreatic cancer has been reported in regulatory databases, and further developments are expected in this area.
In carcinogenicity studies with liraglutide, C cell tumours were observed in thyroid tissue of mice and rats, and C-cells were observed to proliferate in response to GLP-1 agonist therapy. Other studies suggested that C-cells in humans do not express the GLP-1 receptor, that humans exposed to liraglutide had, in aggregate, little or no rise in calcitonin levels, and that non-human primates exposed to liraglutide do not develop thyroid tumors. In contrast, later and more exhaustive analysis found that the GLP-1 receptor is expressed in a subpopulation of human C-cells. Furthermore, GLP-1 receptor expression is more abundant in C-cell hyperplasia, a potential precursor of medullary thyroid cancer, and GLP-1 receptor expression is also present in 20% of those with papillary thyroid cancer, a much more common tumor for which calcitonin levels would be irrelevant. Medullary thyroid cancer is rare, but a relatively high proportion of the population has micro foci of papillary thyroid cancer.
The ADA and US Endocrine Society have both called for greater transparency in terms of access to trial data accrued in the course of development programmes. Important safety data will result from the Safety Evaluation of Adverse Reactions in Diabetes (SAFEGUARD) program which was established by the European Medicines Agency and funded by the European Union (EU). It will combine data from nine population-based health care databases in six countries in the EU and the United States, capturing drug exposure from 1999-2012 and event data on >35 million patients (1.7 million with type 2 diabetes). This is compatible with >240 million patient-years. It also will include an intervention arm with appropriate biochemical and imaging studies. In addition, individual level data will become available over the next few years as a result of the ongoing large cardiovascular safety trial of DPP-4 inhibitors alogliptin (EXAMINE), linagliptin (CAROLINA), saxagliptin (SAVOR-TIMI53), and sitagliptin (TECOS) and GLP-1 receptor agonists duraglutide (REWIND), exenatide (EXSCEL), liraglutide (LEADER), and lixisenatide (ELIXA).
^ Drucker DJ et al. The safety of incretin based
^ Fineman MS et al. Clinical relevance of anti-exenatide safety, efficacy and cross-reactivity with long-term treatment. Diabetes, Obesity & Metabolism [Internet]. 2012 Jan 11 [cited 2012 Mar 19]; Available //www.ncbi.nlm.nih.gov/pubmed/22236356
^ Holst JJ. The physiology of glucagon-like peptide 1. Physiol. Rev. 2007 Oct;87(4):1409–39.
^ Lund A et al. Emerging GLP-1 receptor agonists. Expert Opin Emerg Drugs. 2011 Dec;16(4):607–18.
^ Sivertsen J et al. The effect of glucagon-like peptide 1 on cardiovascular risk. Nature Reviews. Cardiology [Internet]. 2012 Jan 31 [cited 2012 Mar 6]; Available //www.ncbi.nlm.nih.gov/pubmed/22290234
^ Noel RA et al. Increased risk of acute pancreatitis and biliary disease observed in patients with type 2 a retrospective cohort study. Diabetes Care. 2009 May;32(5):834–8.
^ Singh S et al. Glucagonlike peptide-1 based therapies and risk of hospitalization for acute pancreatitis in type 2 diabetes mellitus: a population-based matched case-control study. JAMA Intern Med 2013 Feb 25:1-6. doi: 10.1001/jamainternmed.2013.2720. [Epub ahead of print]
^ Elashoff M et al. Pancreatitis, pancreatic, and thyroid cancer with glucagon-like peptide-1-based therapies.
^ Butler AE et al. Marked expansion of exocrine and endocrine pancreas with incretin therapy in humans with increased exocrine pancreas dysplasia and the potential for glucagon-producing neuroendocrine tumors. Diabetes 2013 March 22nd [Epub before print]
^ Butler PC et al. GLP-1-based therapy for diabetes: what you do not know can hurt you. Diabetes Care 2010;33:453–455
^ Bjerre Knudsen L et al. Glucagon-like Peptide-1 receptor agonists activate rodent thyroid C-cells causing calcitonin release and C-cell proliferation. Endocrinology 2010;151:1473-1486
^ Gier B et al. Glucagon like peptide-1 receptor expression in the human thyroid gland. J Clin Endocrinol Metab 2012;97:121-131