`
`Cardiovascular Actions and Clinical
`Outcomes With Glucagon-Like Peptide-1
`Receptor Agonists and Dipeptidyl
`Peptidase-4 Inhibitors
`
`ABSTRACT: Potentiation of glucagon-like peptide-1 (GLP-1) action
`through selective GLP-1 receptor (GLP-1R) agonism or by prevention of
`enzymatic degradation by inhibition of dipeptidyl peptidase-4 (DPP-4)
`promotes glycemic reduction for the treatment of type 2 diabetes mellitus
`by glucose-dependent control of insulin and glucagon secretion. GLP-
`1R agonists also decelerate gastric emptying, reduce body weight by
`reduction of food intake and lower circulating lipoproteins, inflammation,
`and systolic blood pressure. Preclinical studies demonstrate that both GLP-
`1R agonists and DPP-4 inhibitors exhibit cardioprotective actions in animal
`models of myocardial ischemia and ventricular dysfunction through
`incompletely characterized mechanisms. The results of cardiovascular
`outcome trials in human subjects with type 2 diabetes mellitus and
`increased cardiovascular risk have demonstrated a cardiovascular benefit
`(significant reduction in time to first major adverse cardiovascular event)
`with the GLP-1R agonists liraglutide (LEADER trial [Liraglutide Effect
`and Action in Diabetes: Evaluation of Cardiovascular Ourcome Results],
`‒13%) and semaglutide (SUSTAIN-6 trial [Trial to Evaluate Cardiovascular
`and Other Long-term Outcomes with Semaglutide], ‒24%). In contrast,
`cardiovascular outcome trials examining the safety of the shorter-acting
`GLP-1R agonist lixisenatide (ELIXA trial [Evaluation of Lixisenatide in
`Acute Coronary Syndrom]) and the DPP-4 inhibitors saxagliptin (SAVOR-
`TIMI 53 trial [Saxagliptin Assessment of Vascular Outcomes Recorded in
`Patients With Diabetes Mellitus-Thrombolysis in Myocardial Infarction
`53]), alogliptin (EXAMINE trial [Examination of Cardiovascular Outcomes
`With Alogliptin Versus Standard of Care in Patients With Type 2 Diabetes
`Mellitus and Acute Coronary Syndrome]), and sitagliptin (TECOS [Trial
`Evaluating Cardiovascular Outcomes With Sitagliptin]) found that these
`agents neither increased nor decreased cardiovascular events. Here we
`review the cardiovascular actions of GLP-1R agonists and DPP-4 inhibitors,
`with a focus on the translation of mechanisms derived from preclinical
`studies to complementary findings in clinical studies. We highlight
`areas of uncertainty requiring more careful scrutiny in ongoing basic
`science and clinical studies. As newer more potent GLP-1R agonists and
`coagonists are being developed for the treatment of type 2 diabetes
`mellitus, obesity, and nonalcoholic steatohepatitis, the delineation of
`the potential mechanisms that underlie the cardiovascular benefit and
`safety of these agents have immediate relevance for the prevention and
`treatment of cardiovascular disease.
`
`Michael A. Nauck, MD
`Juris J. Meier, MD
`Matthew A. Cavender,
`MD, MPH
`Mirna Abd El Aziz, MD
`Daniel J. Drucker, MD
`
`Correspondence to: Michael
`A. Nauck, MD, Department of
`Medicine I, Diabetes Center
`Bochum-Hattingen, St Josef-
`Hospital (Ruhr-Universität
`Bochum), Gudrunstr 56, D-44791
`Bochum, Germany. E-mail michael.
`nauck@rub.de
`
`Key Words: acute myocardial
`infarction ◼ cardiovascular events
`◼ cardiovascular outcomes trials
`◼ congestive heart failure ◼ DPP-
`4Is ◼ GLP-1 ◼ GLP-1R agonists
`◼ incretin ◼ stroke
`
`© 2017 American Heart
`Association, Inc.
`
`Circulation. 2017;136:849–870. DOI: 10.1161/CIRCULATIONAHA.117.028136
`
`August 29, 2017
`
`849
`
`STATE OF THE ART
`
`Downloaded from http://ahajournals.org by on January 17, 2024
`
`PROTECTIVE ORDER MATERIAL
`
`Novo Nordisk Exhibit 2105
`Mylan Pharms. Inc. v. Novo Nordisk A/S
`IPR2023-00724
`Page 00001
`
`
`
`Nauck et al
`
`Glucagon-like peptide-1 (GLP-1) was initially dis-
`
`covered as an insulinotropic hormone produced
`in and secreted from the gut after food intake.1
`It has received attention because of its role in the physi-
`ology of glucose metabolism (ie, its function as an in-
`cretin2) but more so as a parent compound mediating
`the actions of 2 classes of glucose-lowering medica-
`tions used in the treatment of type 2 diabetes mellitus
`(T2D), GLP-1 receptor (GLP-1R) agonists, and dipepti-
`dyl peptidase-4 inhibitors (DPP-4Is).1 GLP-1R agonists,
`either small peptides or much larger peptidomimetics,
`mediate their glucoregulatory actions by a single GLP-
`1R. In contrast, inhibitors of the protease DPP-4 prevent
`the degradation and inactivation of both GLP-1 and the
`incretin hormone glucose-dependent insulinotropic
`polypeptide.1
`GLP-1R agonists and DPP-4Is are approved for the
`treatment of hyperglycemia in patients with T2D.2 Al-
`though glycemic control reduces the microvascular
`complications of diabetes mellitus (neuropathy, ne-
`phropathy, and retinopathy), the relationship between
`glucose control and reduction of macrovascular events
`is more challenging.3 It is notable that incretin-based
`therapies (GLP-1R agonists [GLP-1RAs] and DPP-4Is)
`exert multiple nonglycemic actions in the cardiovascu-
`lar system, heightening the interest in their potential
`for cardiovascular benefit.4–6 The recent findings that
`2 GLP-1RAs, liraglutide7 and semaglutide,8 significantly
`reduced the combined primary outcome of 3 point ma-
`jor adverse cardiovascular events in large cardiovascular
`outcome trials elevates the importance of understand-
`ing how activation of the GLP-1R translates into clinical
`cardiovascular benefit. The purpose of the present re-
`view is to summarize the literature on indirect (through
`lowering glucose and modifying known cardiovascular
`risk factors) and direct (through stimulating GLP-1Rs
`and inhibition of DPP-4) effects of (1) GLP-1, (2) GLP-
`1RAs, and (3) DPP-4Is on the heart and blood vessels.
`Herein we discuss concepts of incretin action in the
`context of results of cardiovascular outcomes trials with
`DPP-4Is and GLP-1RAs and, wherever possible, link un-
`derlying mechanisms to observed clinical benefits.
`
`GLP-1RS IN THE CARDIOVASCULAR
`SYSTEM AND EFFECTS ELICITED BY
`STIMULATING WITH GLP-1, GLP-1RAS,
`OR DPP-4IS (PRECLINICAL STUDIES)
`GLP-1Rs in the Cardiovascular System
`GLP-1R expression has been detected in various cardio-
`vascular tissues and cell types at the mRNA and pro-
`tein levels. Although native GLP-1 improves endothelial
`function, augments ventricular contractility, enhances
`myocardial glucose uptake, and exerts cytoprotective
`
`and metabolic actions on blood vessels and cardiomyo-
`cytes, the endogenous canonical GLP-1R is not highly
`expressed in many of the cell types responsive to GLP-1
`or GLP-1RAs. Hence, some of the well-described actions
`of GLP-1 in preclinical studies may reflect indirect mech-
`anisms or the actions of ≥1 GLP-1 degradation prod-
`ucts acting through GLP-1R-independent mechanisms.
`Details are summarized in Table I in the online-only Data
`Supplement, which highlights well-documented effects
`on the heart (contractile function, substrate supply, cor-
`onary and myocardial blood flow, rate control), blood
`pressure, and platelet aggregation. Because the present
`review is focused on human studies, we refer to Table I
`and accompanying text in the online-only Data Supple-
`ment and several recent reviews for details of preclinical
`studies.4–6,9,10
`
`Potential Mechanisms Explaining
`Biological Effects of N-Terminal GLP-1
`Fragments GLP-1 [9–36] Amide, GLP-1
`[9–37], or GLP-1 [28–36] Amide
`Considerable evidence supports biological activity for
`N-terminally truncated GLP-1 peptides, principally GLP-
`1 [9–36] amide10,11 and GLP-1 [28–36] amide12 in the
`cardiovascular system. Although a distinct receptor
`for these peptides has not been identified, they suc-
`cessfully target cytoplasmic and mitochondria-linked
`pathways, leading to a reduction of reactive oxygen
`species in hepatocytes, endothelial cells, and cardio-
`myocytes.10,11 Moreover, GLP-1 [28–36] directly acti-
`vates prosurvival kinases in the ischemic mouse heart
`or vascular cells through mechanisms linked to soluble
`adenylate cyclase and cAMP generation in isolated car-
`diomyocytes ex vivo.12 Hence, studies using native GLP-
`1 may be associated with activation of dual cardiovas-
`cular pathways mediated through the classical GLP-1R
`and nonclassical cAMP-mediated pathways activated
`by truncated peptides converging on cardiomyocyte
`and vascular protection.10
`
`DPP-4 in the Cardiovascular System
`DPP-4 is widely expressed in most cells and tissues and
`exhibits enzymatic activity against dozens of chemo-
`kines and peptide hormones with roles in inflamma-
`tion, vascular function, stem cell homing, and cell sur-
`vival.13 DPP-4 exhibits exopeptidase activity through its
`2 principal molecular forms, a membrane-tethered 766
`amino acid protein with a small intracellular tail and a
`soluble form that is 39 amino acids smaller, devoid of
`the short membrane spanning domain and intracellular
`tail, and yet otherwise structurally identical.13 Although
`soluble DPP-4 exerts vascular, immune, and proinflam-
`matory actions independent of its catalytic activity, the
`
`850
`
`August 29, 2017
`
`Circulation. 2017;136:849–870. DOI: 10.1161/CIRCULATIONAHA.117.028136
`
`Downloaded from http://ahajournals.org by on January 17, 2024
`
`PROTECTIVE ORDER MATERIAL
`
`Novo Nordisk Exhibit 2105
`Mylan Pharms. Inc. v. Novo Nordisk A/S
`IPR2023-00724
`Page 00002
`
`
`
`majority of the experimental literature has studied the
`importance of DPP-4-mediated peptide cleavage in the
`pathophysiology and treatment of cardiovascular dis-
`ease.
`Attribution of mechanism(s) linking reduction of
`DPP-4 activity to attenuation of cardiovascular injury
`or preservation of cardiovascular function is difficult
`for several reasons. First, DPP-4 cleaves dozens of sub-
`strates simultaneously, initiating complex changes in
`multiple signaling pathways.4,5 Second, the majority of
`DPP-4 substrates circulate at low levels and are difficult
`to quantitate. Third, highly sensitive and specific as-
`says distinguishing full length from DPP-4-cleaved pep-
`tides are generally not commercially available. Hence,
`measurements of total immunoreactive peptide detect
`a mixture of cleaved versus intact substrates. Fourth,
`many of the DPP-4-cleaved peptide metabolites retain
`biological activity in the cardiovascular system, albeit
`through different receptors and signaling pathways.
`Hence, DPP-4 simultaneously inactivates and potenti-
`ates the activity of numerous cardioactive substrates.5,13
`Last, only a few highly selective antagonists for DPP-
`4 peptide substrates are available, and these reagents
`have not been widely used in cardiovascular studies.
`
`Preclinical Effects in Myocardial Infarction
`Models and Cardiovascular Function
`When myocardial infarction is experimentally induced
`by occluding (ligating) a coronary artery, the myocar-
`dial area receiving blood supply through the vessel to
`be occluded can be defined as an area at risk, and the
`resulting area of necrosis can be identified by specific
`staining methods.14 Administration of GLP-1, GLP-
`1RAs (eg, exenatide, liraglutide), and DPP-4Is (eg, si-
`tagliptin, vildagliptin, alogliptin) reduces the resulting
`necrosis (relative to the area at risk), as summarized
`in Figure I in the online-only Data Supplement. Exam-
`ples encompass in vivo and ex vivo (isolated perfused
`heart) studies, studies in rodents and larger mammals,
`and with various pharmacological agents (GLP-1 [7–36
`amide], DPP-4Is, and GLP-1RAs) (Figure I in the online-
`only Data Supplement). Additional studies examining
`effects of the GLP-1RAs exenatide,15 lixisenatide,16
`and albiglutide17 and the DPP-4Is sitagliptin18 and lina-
`gliptin19 have been published. Although occasional re-
`ports do not replicate these findings (eg, with liraglu-
`tide in a porcine model20), the majority of studies found
`a significant reduction in the necrotic area in hearts of
`animals treated with GLP-1 or GLP-1RAs (Figure I in
`the online-only Data Supplement). The cardioprotec-
`tive effects of GLP-1 can be inhibited by the specific
`GLP-1RA exendin [9–39]. Thus, these effects seem to
`be mediated by an interaction with the canonical GLP-
`1R.14 More details are described in the online-only Data
`Supplement.
`
`Cardiovascular Effects of Incretin-Based Drugs
`
`CARDIOVASCULAR ACTIONS IN
`HUMANS
`Table 1 summarizes human studies examining cardio-
`vascular function or changes in renal function, lipopro-
`tein metabolism, and hepatic fat accumulation.
`
`GLP-1R in Human Cardiovascular Tissues
`The atrial expression of the GLP-1R protein was identi-
`fied in nonhuman primate and human hearts using a
`highly specific monoclonal antibody, localizing an im-
`munoreactive GLP-1R protein to cells within the sino-
`atrial node.21 Nevertheless, some studies have detected
`partial GLP-1R mRNA transcripts by reverse transcrip-
`tion polymerase chain reaction techniques using RNA
`isolated from human ventricles, although GLP-1RAs
`such as exenatide failed to augment contractility in
`the majority of isolated strips from human ventricles
`in the same experiments.27 RNASeq analyses have de-
`tected the presence of GLP-1R mRNA transcripts in RNA
`from human left ventricles (http://www.gtexportal.org/
`home/gene/GLP1R). Hence, these findings imply that
`under some circumstances, transcriptional or transla-
`tional control may dictate whether a ventricular GLP-1R
`mRNA transcript is expressed and gives rise to func-
`tional GLP-1R protein in the human heart (including
`the working myocardium in atria and ventricles). The
`presence or absence of a functional GLP-1R in human
`coronary arteries is not clearly established.6
`
`Cardiac Output
`Intravenous GLP-1 at a pharmacological dose improved
`left ventricular function, maximum oxygen uptake, and
`physical performance in subjects with congestive heart
`failure.25 Likewise, intravenous exenatide (GLP-1RA) im-
`proved cardiac index and pulmonary capillary wedge
`pressure and reduced atrial natriuretic peptide.28 How-
`ever, in vitro, exenatide increased contractility in human
`atrial but not ventricular myocardium.27 Larger random-
`ized controlled clinical trials with albiglutide or liraglutide
`failed to demonstrate any beneficial effect of sustained
`GLP-1RA treatment in human subjects with moderate
`to severe heart failure and reduced ejection fraction,29,30
`independent of the presence or absence of diabetes
`mellitus. In patients with advanced heart failure, lira-
`glutide did not improve a composite end point of car-
`diovascular events that included changes in N-terminal
`pro-brain natriuretic peptide. A numeric but statistically
`nonsignificant increase in mortality and hospitalization
`for heart failure was detected (hazard ratio [HR], 1.30;
`95% confidence interval [CI], 0.92–1.83; P=0.14), in-
`dicating a potential for harm in patients with reduced
`ejection fraction and a prior history of hospitalization
`for heart failure. It is possible that this may be related to
`
`Circulation. 2017;136:849–870. DOI: 10.1161/CIRCULATIONAHA.117.028136
`
`August 29, 2017
`
`851
`
`STATE OF THE ART
`
`Downloaded from http://ahajournals.org by on January 17, 2024
`
`PROTECTIVE ORDER MATERIAL
`
`Novo Nordisk Exhibit 2105
`Mylan Pharms. Inc. v. Novo Nordisk A/S
`IPR2023-00724
`Page 00003
`
`
`
`Nauck et al
`
`Table 1. Effects of Stimulating GLP-1 Receptors With GLP-1, GLP-1 Receptor Agonists, or DPP-4 Inhibitors in
`Human Studies, Which Lead to a Modified Cardiovascular Function (Directly or Indirectly)
`
`GLP-1 Receptor Agonists
`
`DPP-4 Inhibitors
`
`Organ
`
`Heart
`
`Effect(s) on
`
`GLP-1 [7–36 Amide] or
`[7–37]
`
`Myocardial glucose
`uptake
`
`• Intravenous GLP-1
`(pharmacological dose): ≈22
`
`Left ventricular function
`
`• Intravenous GLP-1
`(pharmacological dose, 5
`wk): LVEF ↑, VO2 max. ↑,
`6-min walk, distance ↑25
`• Improved LVEF not
`confirmed at lower dose of
`GLP-126
`
`• Exenatide (intravenous,
`pharmacological dose, type 2
`diabetes mellitus, no CAD): ≈23
`
`• Exenatide: In vitro contractility of
`atrial, but not ventricular human
`myocardium ↑27; intravenous: cardiac
`index ↑, PCWP ↑, and ANP ↓28
`• Albiglutide: no significant effects29
`• Liraglutide: trend for reduced rate of
`hospitalization for congestive heart
`failure (LEADER)7; however, trends
`for worse outcomes (not significant)
`in dedicated heart failure trials30,31
`
`• Sitagliptin (subjects without diabetes
`mellitus, subjects with nonischemic
`cardiomyopathy): ↑24
`
`• Sitagliptin (chronic congestive heart
`failure): left ventricular diastolic
`function ↑32
`• Rate of hospitalization for congestive
`heart failure ≈ (TECOS)33
`• Saxagliptin: rate of hospitalization for
`congestive heart failure ↑ (significant)
`SAVOR-TIMI 5334,35
`• Alogliptin: rate of hospitalization
`for congestive heart failure ↑
`(nonsignificant) EXAMINE36,37
`• Vildagliptin: trend to reduced left
`venticular function (VIVIDD trial,
`unpublished)
`
`• Sitagliptin (dobutamine-induced
`stress): LVEF ↑, regional contractility
`↑. Preferential effect in ischemic
`segments46,47
`
`Cardioprotection
`against ischemia/
`myocardial stunning
`
`• Intravenous GLP-1
`(pharmacological dose,
`dobutamine-induced
`stress) LVEF ↑, regional
`contractility ↑38,39
`• Coronary balloon
`occlusion: preserved left
`ventricular function38,40
`• 72 h after acute myocardial
`infarction: LVEF ↑, regional
`wall motility ↑41
`
`• ST-segment elevation myocardial
`infarction: intravenous exenatide:
`salvage index (non-necrosed
`proportion of area at risk) ↑42
`• Subcutaneous exenatide: infarct
`size ↓43
`• Liraglutide preserved LVEF after PCI44
`• Non–ST-segment elevation myocardial
`infarction: liraglutide-preserved LVEF
`after PCI45
`
`Heart rate
`
`• Intravenous GLP-1: ↑
`(small), no decrease in
`vagal control48
`
`• ↑ by 2–3 beats per min49,50
`• Sympathetic activation with
`exenatide?51
`
`• Not reported in a study demonstrating
`lowering in systolic blood pressure by
`≈ 2 mm Hg52
`
`Peripheral
`arteries
`
`Angiogenesis,
`endothelial cell
`proliferation
`
`Endothelium-derived
`vasodilation (NO
`production)
`
`Endothelium-
`independent
`vasodilation
`
`Anti-atherosclerotic
`effects
`
`Blood
`pressure
`
`Systolic
`
`• New vessel formation from
`human endothelial cells
`improved by high doses of
`GLP-153
`
`• Endothelial nitric oxide
`synthase ↑ in HUVECs55
`• Intravenous GLP-1
`(pharmacological dose):
`acetyl choline–induced
`vasodilation ↑ in healthy
`subjects56 and in type 2
`diabetes mellitus with
`stable CAD57
`
`• Nitroprusside-induced
`forearm vasodilation not
`augmented by intravenous
`GLP-1 (pharmacological
`dose)56
`
`• No immediate effects
`
`• Transient increase with
`GLP-1 (intravenous;
`pharmacological dose:
`transient ↑71; physiological
`dose: ≈72)
`
`• Exenatide-stimulated proliferation of
`human coronary artery endothelial
`cells54
`
`• Not reported
`
`• Exenatide: endothelial nitric oxide
`synthase in HUVECs ↑58; postprandial
`endothelial function ↑59
`• Liraglutide: endoplasmic reticulum
`stress (induced by hyperglycemia) ↓60
`and TNFα-induced oxidative stress ↓ and
`inflammation ↓ in HUVECs61; eNOS ↑,
`endothelin-1 expression ↓62
`• Liraglutide: Acetyl choline–mediated
`forearm blood flow (↑) (n.s.)63
`
`• Not reported
`
`• Sitagliptin: reactive hyperemia
`peripheral artery tonometry index ↑64,
`flow- mediated vasodilation (type 2
`diabetes mellitus) ↑65
`• Effect of DPP-4 inhibition on
`endothelial function not confirmed by
`other studies66,67
`
`• Nitroglycerin-mediated dilatation not
`changed by sitagliptin65
`
`• Liraglutide: intimamedia thickness ↓
`over 8 months68
`
`• Sitagliptin,69 linagliptin70: intimamedia
`thickness progression ↓
`
`• ↓ by 2–3 mmHg49,50
`
`• Occasional reports of lowering
`systolic blood pressure in hypertensive
`subjects52
`
`Natriuretic peptides
`
`• ANP ≈ (n.s.)71,73
`
`• Liaglutide: pro-ANP ↓,74 but ANP and
`pro-BNP ≈75,76; ANP ↑ and BNP ↑ also
`reported77
`
`• Not reported
`
`(Continued )
`
`852
`
`August 29, 2017
`
`Circulation. 2017;136:849–870. DOI: 10.1161/CIRCULATIONAHA.117.028136
`
`Downloaded from http://ahajournals.org by on January 17, 2024
`
`PROTECTIVE ORDER MATERIAL
`
`Novo Nordisk Exhibit 2105
`Mylan Pharms. Inc. v. Novo Nordisk A/S
`IPR2023-00724
`Page 00004
`
`
`
`Cardiovascular Effects of Incretin-Based Drugs
`
`Table 1. Continued
`
`Effect(s) on
`
`GLP-1 [7–36 Amide]
`or [7–37]
`
`Glomerular filtration
`
`• Acutely ↑78
`
`Organ
`
`Renal
`function
`
`GLP-1 Receptor Agonists
`
`DPP-4 Inhibitors
`
`• Exenatide. ≈79
`• Lixisenatide: ≈80
`• Liraglutide: ≈7
`
`• Generally no significant effect36
`• Sitagliptin: minor ↓ in the TECOS
`trial33
`
`Albumin excretion
`
`• No immediate effects
`known
`
`• Liraglutide: ↓7,74
`• Lixisenatide: ↓ (P=0.004)80
`
`• Saxagliptin ↓,34 linagliptin ↓81
`
`Metabolic
`milieu
`
`Hyperglycemia
`
`• Plasma glucose ↓82
`
`Fasting lipoproteins/lipid
`concentrations
`
`• No immediate effect,
`nonesterified fatty acids ↓
`(transient)82
`
`• See Figure 2
`
`• See Figure 2
`
`• See Figure 2
`
`• See Figure 2
`
`Postprandial lipid
`concentrations
`
`• Postprandial triglycerides
`↓ (deceleration of gastric
`emptying)83
`
`• Exenatide,84 liraglutide85: triglycerides
`↓, apolipoprotein B-48 ↓,85 and in
`chylomicron remnant lipids ↓84
`
`• Sitagliptin,86 vildagliptin,87 and
`alogliptin88: triglycerides ↓,
`apolipoprotein B-48 ↓
`
`Liver
`
`Hepatic fat deposition
`(hepatic steatosis,
`NAFLD)
`
`• No effects reported
`
`Inflammatory
`responses
`
`Reactive oxygen species/
`oxidative stress
`
`HUVECs: ROS ↓94
`
`• Mechanistic study describes the
`role of exenatide and liraglutide
`in stimulating lipoautophagy
`(macroautophagy and chaperone-
`mediated autophagy) in preventing
`apoptiosis, fat-induced hepatocyte
`death, and progression to hepatic
`fibrosis and cirrhosis89
`• Exenatide: better reversal of fatty
`liver (ultrasonography) than with
`insulin90
`• Liraglutide: resolution of definite
`nonalcoholic steatohepatitis
`(histology) vs placebo ↑91
`
`• Exenatide: ROS generation ↓95,
`anti-oxidative potential in human
`monocytes/ macrophages ↑96
`
`• Vildagliptin: hepatic triglyceride
`content ↓ vs placebo92
`• Sitagliptin: ≈ vs placebo93
`
`• No effects reported
`
`NF-κB binding/activation
`
`No immediate effects
`reported
`
`• Exenatide: nuclear factor-κB binding
`(mononuclear blood cells) ↓95
`
`• Sitagliptin: nuclear factor-κB binding
`(mononuclear blood cells) ↓97
`
`Expression of
`inflammatory cytokines
`in mononuclear cells
`
`IL-6 ↓98
`
`C-reactive protein
`
`Adiponectin
`
`Platelet aggregation
`
`SDF-1 stabilization
`
`No immediate effects
`reported
`
`• No immediate effects
`reported
`
`• No immediate effects
`reported
`
`• No immediate effects
`reported
`
`Platelet
`function
`
`Stem cell
`homing
`
`• Exenatide: TNFα ↓, IL-1ß ↓, etc.95
`• Liraglutide: TNFα ↓, IL-1ß ↓, IL-6 ↓,
`etc.99
`
`• Exenatide: ↓ by 61%101
`• Liraglutide ↓ by 23%102
`
`• Exenatide: ↑ by 12%101
`• Liraglutide: ↑ by 40%99
`
`• Exenatide: platelet aggregation ↓104
`
`• No immediate effects reported
`
`• Sitagliptin: significant reduction in
`IL-6, IL-18, sICAM-1, E-selectin100;
`significant reduction in TNFα, TLR-4,
`TLR-2, CCR-297
`
`• Sitagliptin: ↓ by 44%100
`
`• Increase more substantial with
`vildaglptin than with sitagliptin103
`
`• Potential for reduced platelet
`aggregation(?)105
`
`• Circulating endothelial progenitor
`cells (reduced in subjects with type 2
`diabetes) enhanced after 4 weeks of
`treatment with sitagliptin106
`• Benefits of improved stem cell
`homing not supported by results
`of the SITAGRAMI study (sitagliptin
`for 28 days and granulocyte-colony
`stimulating factor for 5 days after
`acute myocardial infarction)107
`
`Ach indicates acetyl choline; ANP, atrial natriuretuc peptide; BNP, brain-type natriuretic peptide; CAD, coronary artery disease; CCR-2, chemokine receptor type 2;
`DPP-4, dipeptidyl peptidase-4; eNOS, endothelial nitric oxide synthase; EXAMINE, Examination of Cardiovascular Outcomes with Alogliptin versus Standard of Care;
`GLP-1, glucagon-like peptide-1; HUVECs, human umbilical vein endothelial cells; IL-1ß, interleukin 1ß; IL-6, interleukin 6; IL-18, interleukin-18; LEADER, Liraglutide
`Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results; LVEF, left ventricular ejection fraction; NAFLD, nonalcoholic fatty liver disease; NF-κB,
`nuclear factor κB; n.s., not significant; PCI, percutaneous coronary intervention; PCWP, pulmonary capillary wedge pressure; ROS, reactive oxygen species; SAVOR-
`TIMI 53, Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus-Thrombolysis in Myocardial Infarction; SDF-1, stromal-derived
`factor-1; sICAM, soluble intercellular adhesion molecule; SITAGRAMI, Sitagliptin Plus Granulocyte-colony Stimulating Factor in Acute Myocardial Infarction; TECOS,
`Trial to Evaluate Cardiovascular Outcomes after Treatment with Sitagliptin; TLR-2, toll-like receptor-2; TLR-4, toll-like receptor-4; TNFα, tumor necrosis factor α;
`VIVIDD, Vildagliptin in Ventricular Dysfunction Diabetes; VO2, velocity of opxygen uptake; ↑, improved, enhanced; ↓, reduced, worsened; and ≈, no significant change.
`
`Circulation. 2017;136:849–870. DOI: 10.1161/CIRCULATIONAHA.117.028136
`
`August 29, 2017
`
`853
`
`STATE OF THE ART
`
`Downloaded from http://ahajournals.org by on January 17, 2024
`
`PROTECTIVE ORDER MATERIAL
`
`Novo Nordisk Exhibit 2105
`Mylan Pharms. Inc. v. Novo Nordisk A/S
`IPR2023-00724
`Page 00005
`
`
`
`Nauck et al
`
`increases in heart rate or development of arrhythmias.30
`Coronary blood flow was not changed by liraglutide
`treatment (Table 1). In contrast, in a small pilot study
`of 18 subjects with T2D treated with sitagliptin for 24
`weeks, an improvement in diastolic but not systolic ven-
`tricular function was shown.32 Intravenous GLP-1 and
`exenatide (GLP-1RA) did not affect myocardial glucose
`uptake22,23 (Table 1), whereas treatment with the DPP-4I
`sitagliptin increased myocardial glucose uptake (in sub-
`jects without diabetes mellitus with nonischemic cardio-
`myopathy24; Table 1). Overall, GLP-1RA use in human
`subjects with or without diabetes mellitus does not ap-
`pear detrimental to cardiac function, except for patients
`with advanced heart failure. The effects of DPP-4Is in
`short-term mechanistic studies have to be reconciled
`with the observation that the DPP-4 I saxagliptin signifi-
`cantly increased the risk for hospitalization for conges-
`tive heart failure in the SAVOR TIMI 53 trial35 (vide infra
`in the section on clinical trial results).
`
`Ischemic Conditioning
`Intravenous GLP-1 improved left ventricular ejection frac-
`tion and regional contractility during dobutamine-induced
`stress and preserved left ventricular function during coro-
`nary balloon occlusion.38,39 Perioperative intravenous in-
`fusions of GLP-1 during and after aortocoronary bypass
`grafting did not result in changes in left ventricular ejection
`fraction or cardiac index but reduced the need for inotro-
`pic medications.108 In a pilot study with a small number of
`patients with and without diabetes mellitus, 72 hours of
`intravenous GLP-1 infusion to patients with acute myocar-
`dial ischemia undergoing percutaneous revascularization
`improved left ventricular ejection fraction and regional
`wall motility.41 Similar evidence with GLP-1RAs suggests
`an improved salvage of myocardium at risk for necrosis
`with intravenous exenatide42 and a reduced infarct size
`with subcutaneous exenatide.43 Liraglutide treatment re-
`duced the resulting necrotic area109 and improved left ven-
`tricular ejection fraction after percutaneous intervention
`for ST-segment elevation44 and non-ST-segment elevation
`myocardial infarction.45 The DPP-4I sitagliptin improved
`left ventricular ejection fraction and regional contractility
`during dobutamine-induced stress,46 with a preferential
`effect in ischemic segments of the heart.
`
`Heart Rate
`Changes in heart rate have not been consistently de-
`scribed with intravenous infusions of GLP-1, at both
`physiological and pharmacological concentrations. How-
`ever, a small rise in heart rate (usually by 2–3 beats per
`minute) has been described in short-term controlled
`studies with a GLP-1RA (Figure 1). Studies using ambu-
`latory 24-hour monitoring have found larger average
`changes in heart rate of up to 9 beats per minute.110
`
`Also, 24-hour monitoring shows that variation in dura-
`tion of heart rate changes may exist depending on the
`exposure to the GLP-1RA. Thus, long-acting GLP-1RAs
`elevate heart rate for 24 hours, whereas short-acting
`compounds only lead to a transient rise in heart rate for
`the period characterized by effective drug levels.110 The
`relative contributions of the autonomic nervous system
`(as suggested by a study of exenatide51), versus the direct
`actions of GLP-1RAs on the GLP-1R located in the sino-
`atrial node,21 on heart rate (Figure 1) is difficult to ascer-
`tain in human subjects (Table 1). The functional conse-
`quences, if any, of longer term increases in heart rate in
`subjects treated with GLP-1RAs are not entirely known.
`Although such an increment in heart rate does not seem
`to prevent overall beneficial results in terms of clinical
`end points in the LEADER trial (Liraglutide Effect and Ac-
`tion in Diabetes: Evaluation of Cardiovascular Outcome
`Results)7 and SUSTAIN-6 trial (Trial to Evaluate Cardio-
`vascular and Other Long-term Outcomes with Semaglu-
`tide in Subjects with Type 2 Diabetes),8 we do not know
`whether patients with more marked increments in heart
`rate exhibit different outcomes compared with those
`with no or less marked increments. Additional analyses
`will be required to clarify the clinical consequences of
`changes in heart rate with GLP-1RA treatment.
`
`Effects on the Endothelium
`GLP-1 and the GLP-1RA exenatide stimulate prolifera-
`tion of human endothelial cells in ex vivo studies (Ta-
`ble 1). This finding suggests a possible effect of GLP-1
`on new vessel formation, but similar results have not
`been reported for DPP-4Is. GLP-1 and the GLP-1RAs
`exenatide and liraglutide increased nitric oxide syn-
`thase activity in human endothelial cells58,62 (Table 1).
`Furthermore, intravenous GLP-1 improves endothelial
`function,56,57 and the GLP-1RA liraglutide augmented
`acetyl choline-induced vasodilation.63 The effect of lira-
`glutide alone on endothelial function, which is relatively
`resistant to degradation by DPP-4 and thus does not
`form significant amounts of the DPP-4 metabolite, was
`not significant65 (Table 1). This finding could mean that
`peptides structurally related to GLP-1 [9–36] amide or
`GLP-1 [9–37] are needed for this effect (vide supra). In-
`travenous exenatide increased myocardial blood flow.23
`Exenatide improved endothelial function after a lipid-
`rich meal84, and a single subcutaneous dose of exena-
`tide reduced peripheral vascular resistance. The DPP-4I
`sitagliptin improved flow-mediated vasodilation in sub-
`jects with T2D.65 However, other studies did not confirm
`beneficial effects of DPP-4Is on endothelial function66,67
`(Table 1). In contrast to endothelium-dependent vasodi-
`lation, endothelium-independent vasodilation (induced
`by nitroglycerin or nitroprusside) has not been changed
`with either GLP-156 or DPP-4Is.65 Liraglutide reduced en-
`dothelin-1, a peptide able to induce vasoconstriction, in
`
`854
`
`August 29, 2017
`
`Circulation. 2017;136:849–870. DOI: 10.1161/CIRCULATIONAHA.117.028136
`
`Downloaded from http://ahajournals.org by on January 17, 2024
`
`PROTECTIVE ORDER MATERIAL
`
`Novo Nordisk Exhibit 2105
`Mylan Pharms. Inc. v. Novo Nordisk A/S
`IPR2023-00724
`Page 00006
`
`
`
`Cardiovascular Effects of Incretin-Based Drugs
`
`B
`
`D
`
`F
`
`A
`
`C
`
`E
`
`G
`
`Figure 1. Effects of treatment with glucagon-like peptide-1 (GLP-1) receptor agonists and dipeptidyl peptidase-4
`(DPP-4) inhibitors on cardiovascular risk factors as described in placebo-controlled clinical trials using incretin-
`based medications as monotherapy.
`The placebo-subtracted differences to baseline (± standard error of the mean) are shown for glycohemoglobin (A), body
`weight (B), systolic blood pressure (C), pulse rate (D), serum triglycerides (E), low-density lipoprotein (LDL) cholesterol (F), and
`high-density lipoprotein (HDL) cholesterol (G). BID indicates twice daily; q.w., once weekly. *Significant difference (P<0.05).
`†Median instead of mean value reported. ‡Based on 24-hour monitoring. §Reported as no relevant change in vital param-
`eters or lipid levels. Data displayed in Figure 2 are from Moretto et al122 and Simó et al123 (exenatide BID); Rosenstock et al124,125
`and Meier et al110 (lixisenatide); Nauck et al,126 Buse et al,127 and Dungan et al128 (liraglutide); Drucker et al129 and Diamant et
`al130 (exenatide q.w.); Nauck et al131 and Dungan et al128 (dulaglutide); Pratley et al132 and Nauck et al133 (albiglutide); Nauck et
`al134 (semaglutide, 0.8 mg/wk, dose initially slowly escalated; semaglutide has not been approved for t