A therapeutic approach to hyperglycaemia in the setting of acute myocardial infarction: spotlight on glucagon-like peptide 1

Abstract

Patients with acute myocardial infarction (AMI) frequently have abnormalities of glucose metabolism and insulin resistance, both of which are associated with a poor outcome. Glucagon-like peptide 1 (GLP-1) is a naturally occurring incretin with both insulinotropic and insulinomimetic properties which not only controls glucose levels but also has potential beneficial actions on the ischaemic and failing heart. In this review we highlight the underlying pathophysiological mechanisms for the development of hyperglycaemia in AMI, speculate on the potential relationship between GLP-1 and sphingosine-1-phosphate, and review the literature on the role of GLP-1 as an important approach to treating hyperglycaemia in the setting of AMI.

Keywords

acute myocardial infarction (AMI)glucagon-like peptide 1 (GLP-1)

Introduction

A rise in plasma glucose concentration is often observed in the early hours after the onset of acute myocardial infarction (AMI) both in patients with pre-existing diabetes mellitus and those without [Oswald et al. 1984]. Acute hyperglycaemia and diabetes mellitus are independently associated with adverse outcomes after AMI regardless of the method of revascularization (thrombolysis, percutaneous coronary intervention, or coronary artery bypass graft) [Goode et al. 2009; Lakhdar et al. 1984]. Furthermore, the presence of diabetes is associated with a worse prognosis in patients with heart failure, but only in patients with underlying ischaemic heart disease [Domanski et al. 2003; Goode et al. 2009].

Glucagon-like peptide 1 (GLP-1) is an incretin hormone secreted mainly by the enteroendocrine cells of the intestine in response to the presence of nutrients. It facilitates glucose-induced insulin release, and now, analogues of GLP-1 are a new class of drugs used in the management of type 2 diabetes. Emerging research studies suggest that GLP-1 not only control glucose levels but may also exert beneficial cardiovascular effects independent of its glucose-lowering actions [Inzucchi and McGuire, 2008]. In this review we highlight the underlying pathophysiological mechanisms for the development of hyperglycaemia in AMI, speculate on the potential relationship between GLP-1 and sphingosine-1-phosphate (S1P), and review the literature on the role of GLP-1 as an important approach to treating hyperglycaemia in the setting of AMI.

Hyperglycaemia after acute myocardial infarction: frequency and mechanisms

Raised serum glucose is a common feature early after AMI and abnormal glucose tolerance is common among patients with AMI who have no previous diagnosis of diabetes [Norhammar et al. 2002]. Norhammar and colleagues performed oral glucose tolerance tests (OGTTs) in patients without diabetes with AMI at hospital discharge and found that 31% had frank diabetes while a further 35% had impaired glucose tolerance (IGT) [Norhammar et al. 2002]. The Euro Heart Survey on diabetes and the heart reported that OGTTs identified diabetes in 22% and IGT in 36% of 923 patients without diabetes and with acute coronary syndrome [Bartnik et al. 2004].

Several mechanisms have been proposed to account for hyperglycaemia in patients with AMI (Figure 1). First, there may be pre-existing abnormalities in blood glucose metabolism that had previously been undetected. Second, serious illnesses including AMI are accompanied by a generalized stress reaction which includes activation of the hypothalamic–pituitary–adrenal (HPA) axis [Gauna et al. 2005]. As a result, blood glucocorticoids increase, and activation of the sympathetic nervous system results in catecholamine release [Van den Berghe, 2000]. The acute phase of AMI is accompanied by high levels of cortisol, growth hormone and catecholamines [Leor et al. 1993; Opie, 1975]. Increased stress hormones enhance glycogenolysis, gluconeogenesis, proteolysis and lipolysis, all resulting in increased glucose production [Barth et al. 2007; Seematter et al. 2004]. In addition, adrenaline inhibits glucose transport into cells by inhibiting the binding of insulin to its receptor: increased levels of circulating adrenaline can thus result in insulin resistance with hyperinsulinaemia [Gearhart and Parbhoo, 2006; Hunt and Ivy, 2002].

Figure 1.

Mechanisms leading to hyperglycaemia in patients with acute myocardial infarction (AMI). Three main mechanisms seem to underlie the occurrence of hyperglycaemia in patients with AMI: pre-existing abnormalities in blood glucose metabolism, an increased immune response and stimulation of the hypothalamic–pituitary–adrenal (HPA) axis. IL, interleukin; TNF, tumour necrosis factor.

Third, AMI is associated with an inflammatory response and the release of a wide range of cytokines [Frangogiannis et al. 2002]. Some cytokines, such as tumour necrosis factor, activate the HPA axis [Chrousos, 1995] and the activity of the cytokines is also associated with the development of insulin resistance [del Aguila et al. 1999].

Effect of glucagon-like peptide 1

Overview

GLP-1 analogues are a new class of drugs used in the management of type 2 diabetes. GLP-1 is a hormone secreted after oral glucose administration. It stimulates (pro)insulin biosynthesis, which might explain (at least in part) why there is a greater insulin response to an oral glucose load than to an intravenous glucose load. GLP-1 also lowers glucagon concentrations and slows gastric emptying. In addition, it reduces food intake after intra-cerebroventricular administration in animals [Nauck et al. 1997].

These GLP-1 effects are mediated by binding to the GLP-1 receptor, which is a G-coupled receptor linked to adenylate cyclase via a stimulatory G protein and is expressed in pancreatic islet cells and in the kidney, lung, brain, gastrointestinal tract, and heart [Holst, 2007]. The half life of GLP-1 (7-36) in circulation is very brief (1–2 min), as in addition to undergoing renal excretion, it is rapidly degraded by the enzyme dipeptidyl peptidase IV (DPP-IV) to generate an NH₂ terminally truncated metabolite GLP-1 (9-36), which does not act at the GLP-1 receptor [Holst, 2007]. Importantly, as an antidiabetic agent, GLP-1 does not cause hypoglycaemia, as both its stimulatory effect on insulin secretion and its inhibitory action on glucagon release are inhibited in the presence of plasma glucose levels at or below fasting levels [Holst, 2007].

The use of GLP-1 as a clinical therapy, however, is hampered by the need to administer it as an infusion because it is rapidly broken down in plasma by DPP-IV. Therefore, as a therapeutic antidiabetic strategy, synthetic analogues of GLP-1 (such as exenatide and liraglutide) that are resistant to the degradative actions of DPP-IV have been developed. In addition, pharmacological inhibitors of DPP-IV that are capable of augmenting endogenous levels of GLP-1, such as sitagliptin and vildagliptin, are in clinical use.

The side effects of GLP-1 are predominantly gastrointestinal. Nausea is a common adverse effect of exenatide but is generally mild to moderate in intensity and wanes with the duration of treatment [Buse et al. 2004; DeFronzo et al. 2005; Kendall et al. 2005; Zinman et al. 2007]. Exenatide may also cause dose-dependent and progressive weight loss [Amori et al. 2007]. There have been 36 postmarketing reports of acute pancreatitis and 78 reported cases of acute renal failure or renal insufficiency in patients taking exenatide [US Food and Drug Administration, 2012; Elashoff et al. 2011]. Therefore, exenatide should not be used in patients with a creatinine clearance below 30 ml/min. In patients with moderate renal impairment (creatinine clearance 30–50 ml/min), monitoring of serum creatinine is warranted when initiating therapy and after the usual dose increase from 5 to 10 µg [US Food and Drug Administration, 2012].

Emerging research studies suggest that GLP-1 not only controls glucose levels but may also exert beneficial cardiovascular effects independent of its glucose-lowering actions [Inzucchi and McGuire, 2008].

Effect on acute myocardial infarction

The GLP-1 receptor is widely expressed in the heart [Wei and Mojsov, 1995]. The GLP-1 receptor is a G-protein–coupled receptor and is a distinct member of the glucagon-secretin receptor super family which has been shown to function by causing intracellular calcium influx in addition to upregulating cyclic adenosine monophosphate (cAMP) [Holz et al. 1995]. Interestingly, cAMP has been demonstrated to protect against apoptosis in several cell types other than myocardial [Hui et al. 2003; Kwon et al. 2004]. However, of specific interest is the fact that GLP-1 has also been shown to promote the activity of phosphoinositide 3 kinase (PI3K) in β cells which has been clearly associated with myocardial protection in the setting of ischaemia–reperfusion injury as well as myocardial preconditioning [Buteau et al. 1999; Hausenloy and Yellon, 2004; Mocanu et al. 2002; Tong et al. 2000]. GLP-1 can directly act on cardiac myocytes and protect them from hypoxia-reoxygenation injury mainly by inhibiting their apoptosis [Xie et al. 2008]. When added to standard therapy, GLP-1 infusion improved regional and global left ventricular function in patients with AMI and severe systolic dysfunction after successful primary angioplasty [Nikolaidis et al. 2004].

GLP-1 (7-36) therapy is capable of limiting AMI reperfusion injury, as evidenced by a reduction in myocardial infarct size in both the isolated and in situ murine heart when GLP-1 (7-36), coadministered with a DPP-IV inhibitor, is given to the heart before a period of sustained ischaemia followed by reperfusion [Bose et al. 2005a]. Dokken and colleagues have recently shown that GLP-1 may decrease myocardial infarct size in rodents, probably through attenuation of neutrophil-mediated reperfusion injury [Dokken et al. 2011]. Consistently, albiglutide, a novel GLP-1 analogue with a long half life, may reduce myocardial infarct size and improved postischaemic cardiac function and energetics following myocardial ischaemia–reperfusion injury [Bao et al. 2011]. In addition, GLP-1 (7-36) is protective against myocardial ischaemia–reperfusion injury when given either as a preconditioning mimetic or at reperfusion in an isolated rat heart model [Bose et al. 2005b]. Furthermore, exenatide administered prior to reperfusion increased myocardial salvage in patients with ST-segment elevation myocardial infarction undergoing primary percutaneous coronary intervention [Lønborg et al. 2011].

The mechanism underlying this cardioprotective effect is independent of any effect on glucose and has been linked to GLP-1 receptor activation and recruitment of intracellular prosurvival signalling pathways as well as the downstream phosphorylation and inhibition of the proapoptotic protein BAD [Bose et al. 2005a, 2005b].

Interestingly, recent studies suggest that GLP-1 (9-36), the metabolite that is generated by DPP-IV, is also capable of increasing myocardial glucose uptake, improving left ventricular contractile function in a canine model of pacing-induced dilated cardiomyopathy, improving postischaemic myocardial injury and left ventricular contractile function, and inducing vaso relaxation through a mechanism that involves nitric oxide [Ban et al. 2008; Nikolaidis et al. 2005].

Ongoing clinical trials with different GLP-1 analogues will provide further important information as to whether GLP-1 may constitute a novel therapeutic option to reduce infarct size and preserve cardiac function in adjunction to reperfusion therapy in patients with AMI and type 2 diabetes [Scholte et al. 2011; Scirica et al. 2011].

Effect on sphingosine-1-phosphate signalling

S1P is a circulating bioactive lipid metabolite which has been shown to promote several cellular responses, including cellular proliferation and migration, smooth muscle contraction, and intracellular calcium mobilization [Egom et al. 2010b]. The concentration of S1P in plasma ranges from 0.2 to 0.9 μM. It is tightly associated with albumin and lipoproteins, particularly high-density lipoprotein. S1P may be involved in ischaemic preconditioning-mediated intracellular signalling processes [Egom et al. 2010b]. The protective effect of ischaemic preconditioning can be mimicked by exogenous administration of S1P [Egom et al. 2010b]. We found that the S1P receptor agonist, FTY720, a new generation of S1P receptor modulator in phase III clinical trials as an immunosuppressant agent for the treatment of autoimmune diseases and in organ transplantation [Budde et al. 2006], can prevent ischaemia–reperfusion damage in isolated heart and sino-atrial (SA) nodes in the rat [Egom et al. 2010a]. We also showed that FTY720 reduces ischaemia-induced ventricular arrhythmias and SA nodal dysfunction via activation of p21-activated kinase, a Ser/Thr kinase downstream of small G proteins, and Akt [Egom et al. 2010a, 2010b].

Stimulation of endothelial differentiation gene (EDG) receptors in islets and insulinoma cell lines with S1P inhibits GLP-1-stimulated cAMP production and thus reduces insulin secretion in a concentration-dependent manner [Laychock et al. 2003]. Both EDG receptors and GLP-1 receptors are subclasses of G-protein-coupled receptors, and there may be some cross talk or nonspecific signalling between these receptors. Ban and colleagues have recently shown that GLP-1 increases functional recovery and cardiomyocyte viability after ischaemia–reperfusion injury of isolated hearts and dilated preconstricted arteries from GLP-1 receptor knockout mice [Ban et al. 2008]. Furthermore, GLP-1 (9-36), a by product of GLP-1 metabolism, gives protection against ischaemia–reperfusion injury and induces vasodilation via a mechanism that does not require a functional GLP-1 receptor [Ban et al. 2008]. It is therefore possible that GLP-1 analogues compete for the same receptor as S1P and that their cardiovascular protective effects may be mediated via increased S1P signalling.

Conclusions

The ability of GLP-1 to confer cardioprotection, probably via a mechanism that involves S1P signalling, makes it a suitable candidate for the treatment of hyperglycaemia in patients with AMI as not only does it combat hyperglycaemia but it also appears to improve the cardiovascular outcomes in this clinical setting.

Clearly, further experimental and clinical studies are needed to determine whether synthetic GLP-1 analogues and DPP-IV inhibitors are able to reproduce their direct cardioprotective effect in humans.

Footnotes

Funding

The author acknowledges support from the HYMS Research Strategy Committee Funding (Hull York Medical School), Hull and East Yorkshire Hospitals, National Institute for Health Research (NIHR), Yorkshire and the Humber Deanery.

Conflict of interest statement

The authors declare no conflicts of interest in preparing this article.

References

Amori

Lau

Pittas

A.G.

(2007) Efficacy and safety of incretin therapy in type 2 diabetes: systematic review and meta-analysis. JAMA 298: 194–206.

Ban

Noyan-Ashraf

Hoefer

Bolz

Drucker

Husain

(2008) Cardioprotective and vasodilatory actions of glucagon-like peptide 1 receptor are mediated through both glucagon-like peptide 1 receptor-dependent and -independent pathways. Circulation 117: 2340–2350.

Bao

Aravindhan

Alsaid

Chendrimada

Szapacs

Citerone

. (2011) Albiglutide, a long lasting glucagon-like peptide-1 analog, protects the rat heart against ischemia/reperfusion injury: evidence for improving cardiac metabolic efficiency. PloS one 6: e23570.

Barth

Albuszies

Baumgart

Matejovic

Wachter

Vogt

. (2007) Glucose metabolism and catecholamines. Crit Care Med 35(9 Suppl.): S508–S518.

Bartnik

Ryden

Ferrari

Malmberg

Pyorala

Simoons

. (2004) The prevalence of abnormal glucose regulation in patients with coronary artery disease across Europe. The Euro Heart Survey on Diabetes and the Heart. Eur Heart J 25: 1880–1890.

Bose

Mocanu

Carr

Brand

Yellon

(2005a) Glucagon-like peptide 1 can directly protect the heart against ischemia/reperfusion injury. Diabetes 54: 146–151.

Bose

Mocanu

Carr

Yellon

(2005b) Glucagon like peptide-1 is protective against myocardial ischemia/reperfusion injury when given either as a preconditioning mimetic or at reperfusion in an isolated rat heart model. Cardiovasc Drugs Ther 19: 9–11.

Budde

Schutz

Glander

Peters

Waiser

Liefeldt

. (2006) FTY720 (fingolimod) in renal transplantation. Clin Transplant 20(Suppl. 17): 17–24.

Buse

Henry

Han

Kim

Fineman

Baron

(2004) Effects of exenatide (exendin-4) on glycemic control over 30 weeks in sulfonylurea-treated patients with type 2 diabetes. Diabetes Care 27: 2628–2635.

10.

Buteau

Roduit

Susini

Prentki

(1999) Glucagon-like peptide-1 promotes DNA synthesis, activates phosphatidylinositol 3-kinase and increases transcription factor pancreatic and duodenal homeobox gene 1 (PDX-1) DNA binding activity in beta (INS-1)-cells. Diabetologia 42: 856–864.

11.

Chrousos

(1995) The hypothalamic–pituitary–adrenal axis and immune-mediated inflammation. N Engl J Med 332: 1351–1362.

12.

DeFronzo

Ratner

Han

Kim

Fineman

Baron

(2005) Effects of exenatide (exendin-4) on glycemic control and weight over 30 weeks in metformin-treated patients with type 2 diabetes. Diabetes Care 28: 1092–1100.

13.

del Aguila

Claffey

Kirwan

(1999) TNF-alpha impairs insulin signaling and insulin stimulation of glucose uptake in C2C12 muscle cells. Am J Physiol 276: E849–E855.

14.

Dokken

La Bonte

Davis-Gorman

Teachey

Seaver

McDonagh

(2011) Glucagon-like peptide-1 (GLP-1), immediately prior to reperfusion, decreases neutrophil activation and reduces myocardial infarct size in rodents. Horm Metab Res 43: 300–305.

15.

Domanski

Krause-Steinrauf

Deedwania

Follmann

Ghali

Gilbert

. (2003) The effect of diabetes on outcomes of patients with advanced heart failure in the BEST trial. J Am Coll Cardiol 42: 914–922.

16.

Egom

E.E.

Musa

Mohamed

Wang

Cartwright

. (2010a) FTY720 prevents ischemia/reperfusion injury-associated arrhythmias in an ex vivo rat heart model via activation of Pak1/Akt signaling. J Mol Cell Cardiol 48: 406–414.

17.

Egom

E.E.

Solaro

Lei

(2010b) Cardioprotection in ischemia/reperfusion injury: spotlight on sphingosine-1-phosphate and bradykinin signalling. Prog Biophys Mol Biol 103: 142–147.

18.

Elashoff

Matveyenko

Gier

Elashoff

Butler

P.C.

(2011) Pancreatitis, pancreatic, and thyroid cancer with glucagon-like peptide-1-based therapies. Gastroenterology 141: 150–156.

19.

Frangogiannis

Smith

Entman

(2002) The inflammatory response in myocardial infarction. Cardiovasc Res 53: 31–47.

20.

Gauna

van den Berghe

van der Lely

(2005) Pituitary function during severe and life-threatening illnesses. Pituitary 8: 213–217.

21.

Gearhart

Parbhoo

(2006) Hyperglycemia in the critically ill patient. AACN Clin Issues 17: 50–55.

22.

Goode

John

Rigby

Kilpatrick

Atkin

Bragadeesh

. (2009) Elevated glycated haemoglobin is a strong predictor of mortality in patients with left ventricular systolic dysfunction who are not receiving treatment for diabetes mellitus. Heart 95: 917–923.

23.

Hausenloy

Yellon

(2004) New directions for protecting the heart against ischaemia-reperfusion injury: targeting the Reperfusion Injury Salvage Kinase (RISK)-pathway. Cardiovasc Res 61: 448–460.

24.

Holst

(2007) The physiology of glucagon-like peptide 1. Physiol Rev 87: 1409–1439.

25.

Holz

Leech

Habener

(1995) Activation of a cAMP-regulated Ca(2+)-signaling pathway in pancreatic beta-cells by the insulinotropic hormone glucagon-like peptide-1. J Biol Chem 270: 17749–17757.

26.

Hui

Nourparvar

Zhao

Perfetti

(2003) Glucagon-like peptide-1 inhibits apoptosis of insulin-secreting cells via a cyclic 5′-adenosine monophosphate-dependent protein kinase A- and a phosphatidylinositol 3-kinase-dependent pathway. Endocrinol 144: 1444–1455.

27.

Hunt

Ivy

(2002) Epinephrine inhibits insulin-stimulated muscle glucose transport. J Appl Physiol 93: 1638–1643.

28.

Inzucchi

McGuire

(2008) New drugs for the treatment of diabetes: part II: incretin-based therapy and beyond. Circulation 117: 574–584.

29.

Kendall

Riddle

Rosenstock

Zhuang

Kim

Fineman

. (2005) Effects of exenatide (exendin-4) on glycemic control over 30 weeks in patients with type 2 diabetes treated with metformin and a sulfonylurea. Diabetes Care 28: 1083–1091.

30.

Kwon

Pappan

Marshall

Schaffer

McDaniel

(2004) cAMP Dose-dependently prevents palmitate-induced apoptosis by both protein kinase A- and cAMP-guanine nucleotide exchange factor-dependent pathways in beta-cells. J Biol Chem 279: 8938–8945.

31.

Lakhdar

Stromberg

McAlpine

(1984) Prognostic importance of hyperglycaemia induced by stress after acute myocardial infarction. Br Med J (Clin Res Ed) 288: 288.

32.

Laychock

Tian

Sessanna

(2003) Endothelial differentiation gene receptors in pancreatic islets and INS-1 cells. Diabetes 52: 1986–1993.

33.

Leor

Nesto

Kloner

(1993) Acute myocardial infarction and diabetes mellitus. Cardiovasc Res 27: 1913–1915.

34.

Lønborg

Vejlstrup

Kelbæk

Bøtker

Kim

Mathiasen

. (2011) Exenatide reduces reperfusion injury in patients with ST-segment elevation myocardial infarction. Eur Heart J 14 September (epub ahead of print). Available at: http://eurheartj.oxfordjournals.org/content/early/2011/09/13/eurheartj.ehr309.abstract (accessed 23 July 2012).

35.

Mocanu

Bell

Yellon

(2002) PI3 kinase and not p42/p44 appears to be implicated in the protection conferred by ischemic preconditioning. J Mol Cell Cardiol 34: 661–668.

36.

Nauck

Holst

Willms

Schmiegel

(1997) Glucagon-like peptide 1 (GLP-1) as a new therapeutic approach for type 2-diabetes. Exp Clin Endocrinol Diabetes 105: 187–195.

37.

Nikolaidis

Elahi

Shen

Shannon

(2005) Active metabolite of GLP-1 mediates myocardial glucose uptake and improves left ventricular performance in conscious dogs with dilated cardiomyopathy. Am J Physiol Heart Circ Physiol 289: H2401–2408.

38.

Nikolaidis

Mankad

Sokos

Miske

Shah

Elahi

. (2004) Effects of glucagon-like peptide-1 in patients with acute myocardial infarction and left ventricular dysfunction after successful reperfusion. Circulation 109: 962–965.

39.

Norhammar

Tenerz

Nilsson

Hamsten

Efendic

Ryden

. (2002) Glucose metabolism in patients with acute myocardial infarction and no previous diagnosis of diabetes mellitus: a prospective study. Lancet 359: 2140–2144.

40.

Opie

(1975) Metabolism of free fatty acids, glucose and catecholamines in acute myocardial infarction. Relation to myocardial ischemia and infarct size. Am J Cardiol 36(7): 938–953, http://www.ncbi.nlm.nih.gov/pubmed/1106170.

41.

Oswald

G.A.

Corcoran

Yudkin

J.S.

(1984) Prevalence and risks of hyperglycaemia and undiagnosed diabetes in patients with acute myocardial infarction. Lancet 1: 1264–1267.

42.

Scholte

Timmers

Bernink

Denham

Beek

Kamp

. (2011) Effect of additional treatment with Exenatide in patients with an Acute Myocardial Infarction (EXAMI): study protocol for a randomized controlled trial. Trials 12: 240.

43.

Scirica

Bhatt

Braunwald

Steg

Davidson

Hirshberg

. (2011) The design and rationale of the saxagliptin assessment of vascular outcomes recorded in patients with diabetes mellitus-thrombolysis in myocardial infarction (SAVOR-TIMI) 53 study. Am Heart J 162: 818–825.

44.

Seematter

Binnert

Martin

Tappy

(2004) Relationship between stress, inflammation and metabolism. Curr Opin Clin Nutr Metab Care 7: 169–173.

45.

Tong

Chen

Steenbergen

Murphy

(2000) Ischemic preconditioning activates phosphatidylinositol-3-kinase upstream of protein kinase C. Circ Res 87: 309–315.

46.

US Food and Drug Administration (2012) MedWatch The FDA Safety Information and Adverse Event Reporting Program: Safety Information – Byetta (exenatide) – renal failure. Available at: http://www.fda.gov/safety/MedWatch/SafetyInformation/SafetyAlertsforHumanMedicalProducts/ucm188703.htm (accessed 2 December 2009).

47.

Van den Berghe

(2000) Novel insights into the neuroendocrinology of critical illness. Eur J Endocrinol 143: 1–13.

48.

Wei

Mojsov

(1995) Tissue-specific expression of the human receptor for glucagon-like peptide-I: brain, heart and pancreatic forms have the same deduced amino acid sequences. FEBS Lett 358: 219–224.

49.

Xie

Wang

Sha

Zhou

Wang

Han

. (2008) Effects and mechanism of glucagon-like peptide-1 on injury of rats cardiomyocytes induced by hypoxia-reoxygenation. Chin Med J 121: 2134–2138.

50.

Zinman

Hoogwerf

Duran Garcia

Milton

Giaconia

Kim

. (2007) The effect of adding exenatide to a thiazolidinedione in suboptimally controlled type 2 diabetes: a randomized trial. Ann Intern Med 146: 477–485.