New Drug Developments
`
`Incretin Mimetics as Emerging Treatments for Type 2 Diabetes
`
`Scott V Joy, Philip T Rodgers, and Ann C Scates
`
`OBJECTIVE: To review the physiology, pharmacology, and clinical efficacy of glucagon-like peptide (GLP-1) and the incretin mimetics
`exenatide and liraglutide in clinical studies.
`DATA SOURCES: Primary literature obtained via MEDLINE (1966–April 2004) and International Pharmaceutical Abstracts (1970–April
`2004) searches; abstracts obtained from meeting sources and manufacturers.
`STUDY SELECTION AND DATA EXTRACTION: All English-language studies and abstracts evaluating GLP-1, exenatide, and liraglutide in
`the treatment of patients with type 2 diabetes were reviewed. Data from animal studies were also included if human data were not
`available. Primary and review articles related to the physiology, development, and evaluation of GLP-1s were reviewed.
`DATA SYNTHESIS: GLP-1, exenatide (exendin-4, AC2993), and liraglutide (NN2211) are incretin mimetics that have been shown in
`human studies to be an effective treatment to improve glycemic control in patients with type 2 diabetes. Mechanisms by which these
`compounds improve glycemic control include enhancing glucose-dependent pancreatic secretion of insulin in response to nutrient
`intake, inhibiting glucagon secretion, delaying gastric emptying, and promoting early satiety. GLP-1 has been shown to promote
`pancreatic progenitor cell differentiation and improve β-cell function and lifespan. Reported adverse effects of exenatide and liraglutide
`include nausea, vomiting, and transient headache, as well as increased risk of hypoglycemia when used with sulfonylureas.
`CONCLUSIONS: Clinical studies show that GLP-1, exenatide, and liraglutide improve glycemic control for patients with type 2 diabetes
`through unique mechanisms not available with current pharmaceutical products. Ongoing Phase III studies will help to further
`position these compounds as treatment options for patients with type 2 diabetes.
`KEY WORDS: exenatide, glucagon-like peptides, incretin mimetics, liraglutide, type 2 diabetes.
`Ann Pharmacother 2005;39:110-8.
`
`Published Online, 23 Nov 2004, www.theannals.com, DOI 10.1345/aph.1E245
`
`Currently, 18.2 million Americans have type 2 diabetes,
`
`and treatment consists of lifestyle modifications, oral
`hypoglycemic agents (sulfonylureas, biguanides, acarbose,
`thiazolidinediones), and/or insulin therapy.1 Unresponsive-
`ness to oral antihyperglycemic agents is high: after 5–7
`years of therapy with sulfonylureas, 50% of these patients
`will require insulin therapy.2 β-Cell dysfunction leads to
`decreased insulin secretion and resulting hyperglycemia.3
`An emerging area of study involves the enteroinsular axis,
`a process by which peptides (gut hormones, incretin hor-
`mones) are secreted by intestinal cells in response to food
`
`Author information provided at the end of the text.
`Dr. Joy serves as a consultant for Eli Lilly and Amylin Pharmaceu-
`ticals and is a member of the Speaker’s Bureau for Pfizer and
`Wyeth Pharmaceuticals.
`
`intake to affect pancreatic insulin secretion. This article re-
`views the available data on the naturally occurring incretin
`hormone glucagon-like peptide (GLP-1) and similar in-
`cretin mimetics in Phase III studies (exenatide, liraglutide),
`with emphasis on the physiology, pharmacokinetics, effi-
`cacy, and safety of these compounds and their emerging
`role in the treatment of patients with type 2 diabetes.
`
`Data Sources
`
`MEDLINE (1966–April 2004) and International Phar-
`maceutical Abstracts searches (1970–April 2004) were con-
`ducted for English-language articles using the terms gluca-
`gon-like peptides, incretin mimetics, exendin- 4, AC2993,
`exenatide, NN2211, and liraglutide. Articles relevant to the
`physiology, pharmacology, and efficacy in published clini-
`cal trials were reviewed.
`
`110 I The Annals of Pharmacotherapy I 2005 January, Volume 39
`
`www.theannals.com
`
`MPI EXHIBIT 1073 PAGE 1
`
`

`

`Physiology of Incretin Hormones
`
`Incretin hormones are peptides released from cells in the
`gastrointestinal tract in response to a nutrient stimulus,
`leading to glucose-dependent insulin release from the pan-
`creas.4 GLP-1 is a naturally occurring incretin hormone in
`humans derived from the proglucagon gene. In mammals,
`the proglucagon RNA transcript (Figure 1) is translated
`and processed differently in pancreatic islet cells and in-
`testinal cells.5,6 The α cells of the pancreas have prohor-
`mone convertase 2 that converts proglucagon into func-
`tional glucagon, and intestinal L cells of the jejunum and
`ileum process proglucagon into GLP-1.7-9 Glucagon is the
`counterregulatory hormone of insulin, raising plasma lev-
`els of glucose in response to insulin-induced hypoglycemia
`by enhancing glycogenolysis and gluconeogenesis.10 The 2
`forms of GLP-1 secreted after meal ingestion are GLP-1
`(7-37)amide and GLP-1 (7-36)amide, which differ by a
`single amino acid.11 Approximately 80% of circulating ac-
`tive GLP-1 is in the form of GLP-1 (7-36)amide.12 These 2
`hormones have identical half-lives, equal potency, and
`similar biological activities.13 GLP-1 acts by binding to a
`G-protein–linked receptor expressed on islet β cells.14
`No reports exist of genetic mutations in patients with
`type 2 diabetes or evidence that patients with maturity-on-
`set diabetes of the young have a genetic linkage to defects
`of the GLP-1 receptor gene.15 In patients with type 2 dia-
`betes or impaired glucose tolerance, there are modest but
`significant reductions in meal-stimulated circulating levels
`of GLP-1.16-18 GLP-1 is rapidly secreted by the L cells of
`the intestine in response to food ingestion in humans, by
`both neural and hormonal signaling initiated by exposing
`the proximal gastrointestinal tract to ingested nutrients, as
`well as by subsequent direct contact of those nutrients as they
`are exposed to the L cells in the distal jejunum and ileum,
`particularly in response to a mixed meal or a meal high in fat
`and complex carbohydrates.11,19,20 Release of GLP-1 has been
`shown to potentiate glucose-dependent insulin secretion by
`stimulating β-cell growth and differentiation and insulin gene
`expression.21-24 GLP-1 has been shown to inhibit β-cell death
`in human islets cultured in vitro.23,25 The mechanism of this
`
`Incretin Mimetics for Type 2 Diabetes
`
`action has been shown in rodent models to involve GLP-1 in-
`ducing expression of the homeodomain transcription factor
`islet duodenum homeobox-1 (IDX-1), a master regulator of
`pancreas development and β-cell function, inducing progeni-
`tor cells found in the pancreatic ducts to develop into β-
`cells.24,26,27 These pancreatic progenitor cells have been isolat-
`ed from humans, and proof of concept that these cells reside
`within human pancreatic ducts has been completed.28 These
`data suggest that GLP-1 can improve β-cell differentiation,
`increase β-cell mass, and increase β-cell lifespan. Other ef-
`fects of GLP-1 have been shown to inhibit glucagon secre-
`tion, delay gastric emptying, and act through the central
`nervous system to decrease appetite, increase sensation of
`satiety, and promote weight loss.29-33
`
`Pharmacology of GLP-1
`
`GLP-1 is inactivated rapidly by dipeptidyl peptidase IV
`(DPP-IV) in plasma by cleaving the penultimate alanine
`residue to generate GLP-1 (9-36)amide, with the half-life
`of intact GLP-1 in vivo <2 minutes.34,35 A study of healthy
`volunteers receiving subcutaneous GLP-1 (7-36) amide
`showed a dose-related increase of GLP-1 plasma concen-
`trations, and after an intravenous glucose bolus, more than
`in the basal state, GLP-1 augmented the insulin secretory
`response and suppressed plasma glucagon.36 A study of
`healthy patients given an intravenous infusion of GLP-1
`during ingestion of a meal showed that GLP-1 increased
`and augmented the dose–response relationship between
`glucose concentration and insulin secretion.37 The effect of
`GLP-1 on glucose control appears to be a function of en-
`hanced, glucose-dependent insulin secretion and not im-
`proved peripheral insulin resistance.38,39 Comparing acute
`administration of GLP-1 with a 3-hour infusion of GLP-1
`in diabetic and nondiabetic patients showed that both
`groups had improved insulin secretion, but only the 3-hour
`infusion led to augmented improvement in first-phase in-
`sulin response.40 Comparison of 16- and 24-hour infusions
`of GLP-1 showed that the most efficacious dosing based
`on nocturnal and fasting plasma glucose levels was in the
`24-hour continuous infusion group.41 Evalua-
`tion of GLP-1 (7-36) given by a single buccal
`tablet showed that therapeutic plasma levels of
`GLP-1 in healthy humans were achieved, con-
`sistent with a relative bioavailability of 7%
`versus intravenous injection and 47% versus
`subcutaneous injection.42
`The clinical applicability of GLP-1 there-
`fore is limited in humans, as the insulin-secret-
`ing effects of GLP-1 are best observed with
`continuous infusion and that GLP-1 is rapidly
`degraded by DPP-IV.
`
`Figure 1. Proglucagon mRNA transcript. Copyright 2003, The Endocrine Society.5 GLP =
`glucagon-like peptide; GRPP = glicentin-related pancreatic polypeptide; IP = intervening
`peptide; MPGF = major proglucagon fragment.
`
`Exenatide
`
`Exenatide is a naturally occurring 39-amino
`acid GLP-1 agonist isolated from the salivary
`gland venom of the lizard Heloderma suspec-
`
`www.theannals.com
`
`The Annals of Pharmacotherapy I 2005 January, Volume 39 I 111
`
`MPI EXHIBIT 1073 PAGE 2
`
`

`

`SV Joy et al.
`
`tum (Gila monster).43 This peptide has 53% amino acid
`similarity to mammalian GLP-1, effectively binds to the
`GLP-1 receptor, and is highly resistant to DPP-IV.44 Com-
`pared with GLP-1, exenatide has up to 3000-fold greater
`potency for glucose-lowering in vivo and is a more potent
`stimulator of glucose-dependent insulin release when giv-
`en intravenously.45-47 Infusion of exenatide and GLP-1 in
`animal models has replicated a bell- shaped dose response
`for both compounds, with exenatide demonstrating a
`greater maximal effect than GLP-1.48 Phase I clinical trials
`have shown that subcutaneous exenatide was generally
`well tolerated at doses of ≤0.1 µg/kg, with nausea and
`vomiting being dose-limited at 0.3 µg/kg, and all exenatide
`doses increased plasma insulin.49 Exenatide suppresses
`glucagon secretion, slows gastric emptying, reduces food
`intake, and promotes β-cell proliferation and neogenesis
`from precursor cells. Exenatide does not acutely enhance in-
`sulin activity in nondiabetic humans and, unlike GLP-1,
`does not suppress gastric acid secretion.39,50-54
`Phase II trials with exenatide in 323 patients with type 2
`diabetes identified an optimal glucose-lowering subcuta-
`neous dose range of 0.05– 0.2 µg/kg.55 Transient nausea
`and vomiting were dose-limiting adverse events, with this
`nausea mitigated by a dose initiation period of one month
`at 5 µg twice daily followed by a maintenance dose of 10
`µg twice daily.56
`
`PHARMACOKINETICS
`
`Exenatide was administered to rats for the purpose of
`investigating the pharmacokinetics after intravenous, sub-
`cutaneous, and intraperitoneal administration, and showed
`dose-dependent increases in plasma concentration, regard-
`less of administration route.57 Exenatide plasma levels ex-
`hibited a gradual single-phase decay, occurring 30 minutes
`after injection of all doses. Clearance of exenatide from
`plasma was 4–8 mL/min and was primarily cleared via a
`renal route.
`Table 1 describes the pharmacokinetics of exenatide in
`humans.58,59 Exenatide demonstrated first-order pharma-
`cokinetics, with the concentration decreasing in a logarith-
`mic fashion over time and plasma concentrations detectable
`6 hours after the injection. The volume of dis-
`tribution following intravenous infusions to
`healthy volunteers was 64 ± 7 mL/kg, suggest-
`ing that exenatide is not likely to be highly tis-
`sue bound. Exenatide is renally cleared, with a
`documented clearance rate of 1.8 ± 0.2
`mL/kg/min.
`
`CLINICAL EFFICACY
`
`Several clinically relevant trials have docu-
`mented the beneficial effects of exenatide
`(Table 251,58-61). A randomized, double-blind,
`single-dose crossover study in healthy individ-
`uals was performed.58 Subjects received either
`exenatide or a placebo infusion of NaCl 0.9%
`
`Cmax = peak plasma concentration; Cl = clearance; NR = not reported; t1/2 = half-
`life; tmax = time to maximum plasma concentration; Vd = volume of distribution.
`aMean ± SEM.
`
`112 I The Annals of Pharmacotherapy I 2005 January, Volume 39
`
`www.theannals.com
`
`for a total of 6 hours on day 1, returning in one week to re-
`ceive another infusion in a double-blind fashion after an 8-
`hour fast. A standard breakfast was given one hour after
`start of infusion, and 3 hours later, the subjects were given
`a free-choice lunch buffet. Food quantities were weighed
`before and after participants ate lunch, and feelings of full-
`ness were rated.
`No patients experienced adverse events from the infu-
`sion. In the fasting state in the first hour, fasting plasma
`glucose levels fell significantly from 81 to 72 mg/dL (p <
`0.005), with no change in the placebo group. There was no
`significant change in the 3-hour postprandial glucose level,
`although the peak postprandial glucose level (glucose ex-
`cursion) was lower in the exenatide group and gastric emp-
`tying time was decreased. Food intake was 19% lower in
`the exenatide group (208 calories) at the buffet lunch com-
`pared with the placebo group, and the effect appeared to
`carry over to the evening meal. Noting this small decrease
`in absolute caloric intake, subjects reported no difference
`in feelings of fullness or nausea.58
`Protocols to study the effects of exenatide in both a hy-
`perglycemic and fluctuating state of blood glucose concen-
`trations were performed in patients with type 2 diabetes
`and healthy volunteers.60 To reproduce a hyperglycemic
`state in patients with diabetes, diabetic medications were
`withheld for 3 days before the study period. A hyper-
`glycemic clamp (continuous glucose infusion) that raised
`subjects’ blood glucose levels by approximately 100 mg/dL
`from the fasting state for 5 continuous hours was used to
`measure changes in insulin release. An exenatide infusion
`was then given for one hour, and a standard meal was giv-
`en 5.5 hours later. A second protocol allowed the hyper-
`glycemic clamp to be stopped in the third hour to allow
`plasma glucose to fall to fasting levels, then increased
`again by 100 mg/dL during the fifth and final hours to de-
`termine the level of glucose dependency of exenatide.
`The results from the first protocol revealed that insulin
`release was potentiated during and in the hours following
`the exenatide infusion. Results from the second protocol
`showed that, when the fasting state was resumed, insulin
`levels fell, but when plasma glucose was raised again,
`plasma insulin levels were potentiated to an even greater de-
`
`Table 1. Pharmacokinetic Parameters of Exenatide
`
`Parameter
`Dose/route
`a
`Cmax
`
`Edwards et al. (2001)58
`0.05 pmol/kg/min intravenous
`16.4 ± 0.9 pmol/L
`
`t1/2 (min)a
`
`Cl (mL/min•kg)a
`Vd (mL/kg)a
`tmax (h)
`
`26 ± 3
`
`1.8 ± 0.2
`64 ± 7
`2–3
`
`Fineman et al. (2003)59
`0.08 µg/kg sc
`163 ± 86 pg/mL (day 1)
`159 ± 81 pg/mL (day 28)
`202 ± 182 (day 1)
`226 ± 170 (day 28)
`NR
`NR
`NR
`
`MPI EXHIBIT 1073 PAGE 3
`
`

`

`gree than during the initial phase. The results demonstrated a
`glucose-dependent insulinotropic effect from exenatide, sug-
`gesting an amplification of β-cell insulin release when glu-
`cose concentrations were above the normal range, but not
`when glucose concentrations were below or within the
`normal range.60
`To evaluate the effect of multiple doses of exenatide
`over a one-month period, a study of 9 patients with type 2
`diabetes (mean age 57 y, mean body mass index [BMI] 33
`kg/m2, mean glycosylated hemoglobin [HbA1c] 9.1%) not
`being treated with insulin was performed.61 Oral diabetic
`agents were stopped a week prior to the study, and patients
`were placed on hyperglycemic and hyperinsulinemic eug-
`lycemic clamps. Subcutaneous injections of exenatide at
`an initial dose of 12 pmol/kg were titrated weekly to a
`maximum of 96 pmol/kg (0.4 µg/kg). Exenatide was con-
`tinued for one month, and patients were instructed to mon-
`itor their blood glucose levels at least 8 times a day (before
`and 1 hour after every meal as well as before and after the
`exenatide injection). Hyperglycemic and hyperinsulinemic
`euglycemic clamps were repeated at the end of the study.
`Results showed that exenatide reduced HbA1c (9.1% vs
`8.3%; p = 0.009) and lowered glucose one hour after
`breakfast (p < 0.0001) after twice-daily dosing. No differ-
`ence was noted in glucose response, plasma insulin levels,
`or glucagon after a month of exenatide compared with be-
`fore treatment. However, C-peptide levels were signifi-
`cantly higher during the clamp. A favorable response to
`
`Incretin Mimetics for Type 2 Diabetes
`
`adipose tissue sensitivity to insulin was noted, with nones-
`terified free fatty acid levels higher after exenatide treat-
`ment, and no desensitization to the effects of exenatide
`over the course of the month were observed.61
`Kolterman et al.51 published a trial that included 2 sub-
`studies of exenatide use in patients with type 2 diabetes.
`The aim of study A was to investigate postprandial glucose
`response, and study B evaluated the effect on fasting plas-
`ma glucose levels. In study A, patients with type 2 diabetes
`(mean age 56 y, mean BMI 29 kg/m2, mean HbA1c ≤12%,
`fasting plasma glucose <260 mg/dL) were divided into
`groups based on their baseline diabetes therapy: (1) diet
`management alone, (2) oral therapy with HbA1c <8%, (3)
`oral therapy with HbA1c 8–12%, or (4) insulin treated.
`Subjects were randomly given a dose of exenatide 0.1
`µg/kg or placebo subcutaneously twice a day for 5 days,
`then switched to the other arm after a 2- to 3-day washout
`period. Patients were given a standardized liquid breakfast
`10 minutes after their morning injection, and postprandial
`information was recorded. A dose of acetaminophen was
`given to measure gastric emptying.
`The authors observed that postprandial glucose levels
`were significantly lower over the 5 hours after the standard
`meal for the exenatide group versus placebo, with no dif-
`ference between day 1 or day 5, nor between any of the
`predefined patient groups. Plasma glucose levels in the ex-
`enatide group were lower in the hours after the meal (from
`160 to 126 mg/dL at 3 h), compared with placebo where it
`
`Table 2. Summary of Clinical Trials with Exenatide and Liraglutide
`
`Pts.
`
`Treatment
`
`Dose
`
`Significant Clinical Response
`
`healthy non-diabetics
`(n = 8)
`
`single 6-h infusion
`
`0.05 pmol/kg/min
`
`lower FBG, lower glucose excursion, reduced
`food intake
`
`healthy non-diabetics
`(n = 7)
`non–insulin-treated
`type 2 diabetics
`(n = 7)
`non–insulin-treated
`type 2 diabetics
`(n = 9)
`
`single 1-h infusion
`
`0.15–0.59 pmol/kg/min
`
`overall insulin release potentiated after infusion
`
`single 1-h infusion
`
`0.15–0.59 pmol/kg/min
`
`insulin release amplified when glucose levels
`above normal
`
`1 mo, once or twice
`daily sc injections
`
`12–96 pmol/kg (0.05–
`0.4 µg/kg)
`
`lower HbA1c, FPG, and 1-h PPPG; elevated
`C-peptide, NEFA levels
`
`Reference
`Exenatide
`Edwards et al.
`(2001)58
`Egan et al.
`(2002)60
`protocol 1
`
`protocol 2
`
`Egan et al.
`(2003)61
`
`Kolterman et al.
`(2003)51
`study A
`
`study B
`
`Fineman et al.
`(2003)59
`Liraglutide
`Juhl et al.
`(2002)66
`
`type 2 diabetics
`(n = 24)
`type 2 diabetics
`(n = 13)
`
`5 days, twice-daily sc
`injections
`single-dose sc in-
`jections
`
`0.1 µg/kg
`
`0.05, 0.1, 0.2 µg/kg
`
`type 2 diabetics
`(n = 109)
`
`28 days, 2 or 3 times
`a day sc injections
`
`0.08 µg/kg
`
`type 2 diabetics
`(n = 11)
`
`single sc dose
`
`10 µg/kg
`
`lower PPPG, postprandial insulin, delayed
`gastric emptying
`lower FPG, increased insulin levels at 3 h, no
`clinical differences among the 3 doses in
`glycemic control
`lower fructosamine, HbA1c, PPPG; increased
`β-cell function
`
`lower FPG and premeal glucose; lower glucose
`exposure and excursion; increased insulin
`levels and delayed gastric emptying
`increased insulin levels in response to elevated
`glucose levels
`
`Chang et al.
`(2003)65
`
`type 2 diabetics
`(n = 10)
`
`single sc dose
`
`7.5 µg/kg
`
`FPG = fasting plasma glucose; HbA1c = glycosylated hemoglobin; NEFA = nonesterified free fatty acid; PPPG = postprandial plasma glucose.
`
`www.theannals.com
`
`The Annals of Pharmacotherapy I 2005 January, Volume 39 I 113
`
`MPI EXHIBIT 1073 PAGE 4
`
`

`

`SV Joy et al.
`
`increased (from 170 to 289 mg/dL). Use of postprandial
`insulin was significantly lower in the exenatide group, with
`similar results found on days 1 and 5. Glucagon levels did
`not change from baseline in the exenatide group, but in-
`creased in the placebo group after the meal. Gastric empty-
`ing was delayed, as measured by postprandial plasma ac-
`etaminophen concentrations, which showed much reduced
`levels for the first 3 hours; peak concentrations at 5 hours
`were not different.51
`Study B was a randomized, double-blind, placebo-con-
`trolled, 4-period crossover trial of patients with type 2 dia-
`betes receiving only oral therapy for diabetes (mean age 49
`y, mean BMI 33 kg/m2).51 In 4 random sequences, subjects
`were given subcutaneous placebo or exenatide 0.05, 0.1, or
`0.2 µg/kg as single doses each evening before a study day.
`All doses of exenatide reduced fasting plasma glucose signif-
`icantly compared with placebo; no dose was found to be sig-
`nificantly better. The nadir of the plasma glucose occurred 3
`hours after the dose given the night before, although it was
`still significantly lower at 8 hours. A dose-dependent rise in
`plasma insulin levels for the first 3 hours after the dose was
`observed, and by 8 hours, insulin levels were similar to
`those of placebo. No differences in glucagon levels were
`noted compared with placebo, although the authors noted a
`trend toward suppression in the exenatide groups. Mild ad-
`verse effects included headache, nausea, and vomiting; no
`subjects withdrew because of adverse effects.
`A multicenter, triple-blind, randomized, parallel, place-
`bo-controlled study examining the effects of exenatide in
`combination with existing oral drug therapy for diabetes
`was performed.59 Patients (mean age 52 y, mean HbA1c
`9.2%, mean BMI 33 kg/m2) were taking either a sulfony-
`lurea, metformin, or both at stable doses for at least 6
`months prior to enrolling in the study and continued these
`medications throughout the study. Patients were assigned
`to 1 of 4 groups for a treatment period of 28 days: placebo
`or exenatide 0.08 µg/kg at breakfast and supper, or at
`breakfast and bedtime, or 3 times a day (all pts. received
`subcutaneous injections 3 times a day, with placebo used
`as needed to maintain assigned group). Measurements of
`fructosamine, HbA1c, body weight, lipids, and other pa-
`rameters were performed at baseline and day 28; fasting
`and postprandial glucose (after a standardized meal) were
`obtained at these times and also on day 14.
`The results showed that fructosamine was significantly
`reduced in the exenatide groups compared with placebo (p <
`0.004). HbA1c was also significantly reduced compared with
`placebo, with an overall mean reduction of 0.9% across the
`treatment groups (p < 0.006). β-Cell function, as calculated
`by the homeostasis model analysis, was 50–100% greater in
`the exenatide groups compared with placebo (p < 0.05).
`Postprandial glucose levels were significantly lower in the
`exenatide groups versus placebo, observed on the first day
`of therapy and maintained until day 28 (p < 0.004). Fasting
`plasma glucose was not significantly different compared
`with placebo at day 28. There were no differences between
`the various exenatide regimens for any of these parameters,
`and no changes in body weight, lipids, vital signs, hemato-
`
`logic parameters, or cortisol were observed. About 20% of
`patients treated with exenatide developed low-titer antibod-
`ies with no effect on therapeutic results. Mild to moderate
`nausea developed in 31% of exenatide patients, most occur-
`ring in the initial days of therapy; only 4 patients withdrew
`from the study due to nausea, and only 13% had persistent
`nausea for the entire treatment period. Fifteen percent of pa-
`tients experienced hypoglycemia, occurring only in those
`taking concurrent sulfonylurea therapy with exenatide.59
`
`Liraglutide
`
`Liraglutide (NN2211) is a synthetic acylated derivative
`of GLP-1 that has agonist activity at GLP-1 receptors be-
`ing developed for once-daily subcutaneous injection. The
`prolonged effects of liraglutide are obtained by attaching a
`fatty acid molecule at one position of the GLP-1 molecule
`that allows for binding to albumin and a slow release from
`the reservoir. Subcutaneous dosing of liraglutide has been
`shown in animal models to reduce food intake and body
`weight, increase insulin secretion, inhibit glucagon secre-
`tion, decrease gastric emptying, reduce blood glucose in a
`dose-dependent manner, and increase the proportion of
`pancreatic β-cells in mice.62-64
`Single doses of liraglutide administered to human sub-
`jects with type 2 diabetes have shown significant increases
`in insulin and C-peptide levels in a glucose-dependent
`manner compared with placebo, and liraglutide adminis-
`tered at bedtime significantly lowered fasting and post-
`prandial glucose, suppressed glucagon secretion, and de-
`layed gastric emptying.65,66
`
`PHARMACOKINETICS
`
`Studies with healthy men to investigate the pharmacoki-
`netics, safety, tolerability, and pharmacodynamics of li-
`raglutide when administered in single doses by the subcu-
`taneous and intravenous routes were performed.66,67 Table
`3 documents the pharmacokinetic results from subjects ad-
`ministered 5 µg/kg intravenously and 10 µg/kg subcuta-
`neously. A subcutaneous dose of 10 µg/kg administered at
`bedtime had a half-life of 10 hours and a time to maximum
`concentration of 12 hours, suggesting that bedtime dosing
`should provide for maximum plasma concentrations to oc-
`cur during the day.66 The absolute bioavailability of liraglu-
`tide was determined to be 55%.67
`A single-center, randomized, double-blind, placebo-con-
`trolled, parallel-group dose-escalation study in healthy men
`(n = 30) was conducted, with data analyzed on day 1 to de-
`termine single-dose kinetics and after days 5–11 to deter-
`mine multiple-dose kinetics. Pharmacokinetic data for
`subjects administered liraglutide 10 µg/kg are summarized
`in Table 3.68
`
`CLINICAL EFFICACY
`
`Published clinical studies of liraglutide are summarized
`in Table 2. A randomized, double-blind, crossover trial
`
`114 I The Annals of Pharmacotherapy I 2005 January, Volume 39
`
`www.theannals.com
`
`MPI EXHIBIT 1073 PAGE 5
`
`

`

`comparing placebo with a single dose of liraglutide for pa-
`tients with type 2 diabetes (mean age 59 y, BMI 29 kg/m2,
`HbA1c 6.5%) previously treated with either diet, sulfony-
`lureas, or metformin was performed.66 Diabetic drug thera-
`py was withheld for 2 days before the start of the trial, and
`then placebo or liraglutide was given subcutaneously at a
`dose of 10 µg/kg at 2200, with fasting blood samples ob-
`tained the following morning. At 900, an accepted stan-
`dard process to measure responsiveness of β-cells to re-
`lease insulin was performed, known as a pulsitile glucose
`infusion, which is a one-minute glucose infusion given ev-
`ery 10 minutes for 90 minutes. Subjects ate a standard
`meal at 1130, and blood samples were obtained for 4 hours
`after the meal. The ability of the patients to finish the meal
`and adverse effects that developed were noted.
`The results showed that fasting plasma glucose was
`lower (124 vs 145 mg/dL; p < 0.05) with liraglutide versus
`placebo. After the meal, overall glucose exposure (as mea-
`sured by AUC) was reduced 23% in the liraglutide group,
`and glucose excursion was 27% lower; no hypoglycemia
`was reported. Fasting insulin secretion and peak insulin
`values after the meal were increased in the liraglutide
`group, gastric emptying was significantly delayed, and no
`effect was observed on insulin sensitivity. No changes in
`fasting glucagon were noted, but postprandial glucagon
`was suppressed. Two patients experienced nausea after
`treatment with liraglutide, with one episode severe enough
`to limit meal consumption.66
`A randomized, double-blind, crossover study to mea-
`sure the effects of a single dose of liraglutide on β-cell sen-
`sitivity to an increasing dose of glucose via a glucose infu-
`sion in patients with type 2 diabetes (mean age 62 y, mean
`BMI 30 kg/m2, mean HbA1c 6.5%) was performed.65 For
`comparison, 10 non-diabetic patients were placed only on
`the glucose infusion described below. Diabetic drug thera-
`py was stopped one day to one week prior to study initia-
`tion. Diabetic patients were given a subcutaneous injection
`of liraglutide 7.5 µg/kg or placebo at 2300 the night before
`the study day. The subjects also received a continuous in-
`sulin infusion of 0.5–5 units/h to ensure baseline fasting
`glucose levels were controlled and similar between sub-
`jects the next morning. The next day, a graded glucose in-
`
`Incretin Mimetics for Type 2 Diabetes
`
`fusion was started. For the first 3 hours, glucose was given
`to raise plasma glucose levels from 90 to about 215 mg/dL,
`with intravenous insulin started to lower the plasma glucose
`to 90 mg/dL. A dextrose infusion was then started to in-
`crease the plasma glucose level again to 215 mg/dL over
`the next 3 hours.
`The results showed that glucose levels throughout the
`study were very similar between all groups (treatment,
`placebo, healthy subjects) as would be expected since the
`goal was to keep glucose levels the same throughout the
`study. However, plasma insulin C-peptide levels, which
`were similar in the fasting state, were significantly higher in
`the treatment group compared with placebo (p < 0.0001).
`The insulin secretion rate was significantly greater with li-
`raglutide than placebo (p < 0.0001) and was similar to that
`of the healthy subjects. The effect was observed as early as
`40–60 minutes into the glucose infusion, and β-cell sensi-
`tivity was estimated to have increased 70% with the li-
`raglutide treatment. All subjects completed the trial, no hy-
`poglycemia was noted, and adverse effects were mild diar-
`rhea and headache.65
`
`Drug Interactions
`
`Hypoglycemia was more common when exenatide was
`used in combination with a sulfonylurea. Care must be tak-
`en to closely monitor blood glucose levels should these new-
`er agents be used as add-on therapy with sulfonylureas. Pa-
`tients should be educated and instructed to monitor for signs
`of hypoglycemia and to self-monitor blood glucose levels
`with the use of these incretin mimetics as monotherapy or if
`used in combination with other hyperglycemic agents.
`
`Cost
`
`Cost information is not available for exenatide or li-
`raglutide since these agents are both investigational. Phar-
`macoeconomic issues that will impact their utilization in-
`clude the individual cost of each therapy, whether the use of
`other antidiabetic agents can be reduced, whether reductions
`in hospital admissions are possible, and whether diabetic
`complication can be prevented, minimized, or eliminated.
`
`Table 3. Pharmacokinetic Parameters of Liraglutide
`
`Parameter
`Dose/route
`Cmax (nmol/L)
`t1/2 (h)
`AUC (h•nmol/L)
`Cl (mL/min•kg)
`Vd (L/kg)
`tmax (h)
`
`Elbrond et al. (2002)67,a
`10 µg/kg sc
`5 µg/kg iv
`7.9 ± 2.7
`NR
`14
`8
`235 ± 33
`216 ± 43
`NR
`0.10 ± 0.02
`NR
`0.07 ± 0.01
`12 ± 6
`NR
`
`Agerso et al. (2002)68,b
`(day 1)
`(day 11)
`10 µg/kg sc
`10 µg/kg sc
`8 ± 1
`11 ± 2
`12.8
`12.8
`246 ± 43
`351 ± 60
`0.16 ± 0.03
`NR
`0.17 ± 0.03
`NR
`11 ± 1
`10 ± 2
`
`Cmax = peak plasma concentration; Cl = clearance; NR = not reported; t1/2 = half-
`life; tmax = time to maximum plasma concentration; Vd = volume of distribution.
`aMean ± SD.
`bMean ± SEM.
`
`Summary
`
`Emerging evidence of the role of the en-
`teroinsular axis in glucose homeostasis has led
`to the development of 2 clinically appealing in-
`cretin mimetics, exenatide and liraglutide. These
`compounds have been shown to improve glu-
`cose control in patients with type 2 diabetes,
`with favorable effects on insulin secretion, β-cell
`function, and gastric emptying. Longer-term
`studies currently ongoing will further position
`these peptides as clinical options to be used as
`monotherapy or in combination with biguanides
`or sulfonylureas to improve glycemic control for
`patients with type 2 diabetes.
`
`www.theannals.com
`
`The Annals of Pharmacotherapy I 2005 January, Volume 39 I 115
`
`MPI EXHIBIT 1073 PAGE 6
`
`

`

`SV Joy et al.
`
`Scott V Joy MD CDE FACP, Assistant Professor of Clinical
`Medicine, Duke University Medical Center, Durham, NC
`Philip T Rodgers PharmD BCPS CDE CPP, Clinical Assistant Pro-
`fessor, University of North Carolina; Clinical Pharmacist, Duke Uni-
`versity Health System
`Ann C Scates PharmD, Clinical Pharmacist and Drug Information
`Spec