Articles
`
`Lancet 2009; 374: 39–47
`Published Online
`June 8, 2009
`DOI:10.1016/S0140-
`6736(09)60659-0
`See Comment page 4
`Division of Endocrinology,
`University of North Carolina
`School of Medicine, Chapel Hill,
`NC, USA (Prof J B Buse MD);
`Dallas Diabetes and Endocrine
`Center at Medical City, Dallas,
`TX, USA (J Rosenstock MD);
`Department of Experimental
`and Clinical Medicine, Magna
`Graecia University of
`Catanzaro, Catanzaro, Italy
`(Prof G Sesti MD); Department
`of Medicine I, St Josef-Hospital,
`Ruhr-University Medical
`Faculty, Bochum, Germany
`(Prof W E Schmidt MD); Hospital
`Universitari Bellivtge-IDIBELL,
`University of Barcelona, CIBER
`de Diabetes y Enfermedades
`Metabólicas Asociadas
`(CIBERDEM), Barcelona, Spain
`(Prof E Montanya MD); Novo
`Nordisk, Princeton, NJ, USA
`(J H Brett MD); Novo Nordisk,
`Bagsvaerd, Denmark
`(M Zychma MD); and Ochsner
`Diabetes Clinical Research Unit,
`Department of Endocrinology,
`Ochsner Medical Center, New
`Orleans, LA, USA (L Blonde MD)
`Correspondence to:
`Prof John B Buse, Division of
`Endocrinology, University of
`North Carolina School of
`Medicine, Chapel Hill, NC, USA
`jbuse@med.unc.edu
`
`Liraglutide once a day versus exenatide twice a day for type 2
`diabetes: a 26-week randomised, parallel-group,
`multinational, open-label trial (LEAD-6)
`
`John B Buse, Julio Rosenstock, Giorgio Sesti, Wolfgang E Schmidt, Eduard Montanya, Jason H Brett, Marcin Zychma, Lawrence Blonde, for the
`LEAD-6 Study Group*
`
`Summary
`Background Unlike most antihyperglycaemic drugs, glucagon-like peptide-1 (GLP-1) receptor agonists have a
`glucose-dependent action and promote weight loss. We compared the effi cacy and safety of liraglutide, a human
`GLP-1 analogue, with exenatide, an exendin-based GLP-1 receptor agonist.
`
`Methods Adults with inadequately controlled type 2 diabetes on maximally tolerated doses of metformin, sulphonylurea,
`or both, were stratifi ed by previous oral antidiabetic therapy and randomly assigned to receive additional liraglutide
`1·8 mg once a day (n=233) or exenatide 10 μg twice a day (n=231) in a 26-week open-label, parallel-group, multinational
`(15 countries) study. The primary outcome was change in glycosylated haemoglobin (HbA1c). Effi cacy analyses were by
`intention to treat. The trial is registered with ClinicalTrials.gov, number NCT00518882.
`
`Findings Mean baseline HbA1c for the study population was 8·2%. Liraglutide reduced mean HbA1c signifi cantly more
`than did exenatide (–1·12% [SE 0·08] vs –0·79% [0·08]; estimated treatment diff erence –0·33; 95% CI –0·47 to –0·18;
`p<0·0001) and more patients achieved a HbA1c value of less than 7% (54% vs 43%, respectively; odds ratio 2·02;
`95% CI 1·31 to 3·11; p=0·0015). Liraglutide reduced mean fasting plasma glucose more than did exenatide (–1·61
`mmol/L [SE 0·20] vs –0·60 mmol/L [0·20]; estimated treatment diff erence –1·01 mmol/L; 95% CI –1·37 to –0·65;
`p<0·0001) but postprandial glucose control was less eff ective after breakfast and dinner. Both drugs promoted similar
`weight losses (liraglutide –3·24 kg vs exenatide –2·87 kg). Both drugs were well tolerated, but nausea was less
`persistent (estimated treatment rate ratio 0·448, p<0·0001) and minor hypoglycaemia less frequent with liraglutide
`than with exenatide (1·93 vs 2·60 events per patient per year; rate ratio 0·55; 95% CI 0·34 to 0·88; p=0·0131; 25·5%
`vs 33·6% had minor hypoglycaemia). Two patients taking both exenatide and a sulphonylurea had a major
`hypoglycaemic episode.
`
`Interpretation Liraglutide once a day provided signifi cantly greater improvements in glycaemic control than did
`exenatide twice a day, and was generally better tolerated. The results suggest that liraglutide might be a treatment
`option for type 2 diabetes, especially when weight loss and risk of hypoglycaemia are major considerations.
`
`Funding Novo Nordisk A/S.
`
`Introduction
`Type 2 diabetes is an increasingly common chronic
`disease. Although diagnosed on
`the basis of
`hyperglycaemia, it is associated with broad metabolic
`abnormalities that contribute to microvascular and
`macrovascular
`complications.
`Importantly, unmet
`pharma cological needs remain despite great advances in
`diabetes care and treatment, and availability of ten
`diff erent antihyperglycaemic medication classes.
`To reach glycaemic targets, various antihyperglycaemic
`drugs—alone or in combination—are commonly required
`in addition to lifestyle interventions. Some agents are
`eventually combined with insulin in complex regimens
`that need daily titration based on glucose monitoring.
`Careful selection of therapies and follow-up is crucial to
`achieve glycaemic control while avoiding other substantial
`problems, particularly weight gain and hypoglycaemia.1
`Glucagon-like peptide-1 (GLP-1) is secreted by intestinal
`L-cells, mainly in response to food intake. It has broad
`
`physiological eff ects, including stimulation of insulin
`secretion and reduction of glucagon secretion, both in a
`glucose-dependent manner, and resulting in reduced
`hepatic glucose production. Furthermore, GLP-1 slows
`gastrointestinal motility and increases satiety with
`reduced food intake. In animal models, it promotes β-cell
`proliferation and probably neogenesis, while reducing
`apoptosis.2–4 Because GLP-1 is rapidly degraded by
`dipeptidyl peptidase-4,5 GLP-1 receptor agonists based on
`exendin or human analogues resistant to dipeptidyl
`peptidase-4 have been developed.
`The current consensus statement from the American
`Diabetes Association (ADA) and the European Association
`for the Study of Diabetes (EASD) about the medical
`management of hyperglycaemia in type 2 diabetes
`suggests
`that comprehensive
`lifestyle management
`combined with metformin should be initiated at diagnosis,
`except in cases of severely uncontrolled hyperglycaemia.1
`Subsequently, treatment should be intensifi ed promptly if
`
`www.thelancet.com Vol 374 July 4, 2009
`
`39
`
`MPI EXHIBIT 1114 PAGE 1
`
`

`

`Articles
`
`glycosylated haemoglobin (HbA1c) values exceed the ADA
`target of less than 7%. Recently, the consensus panel
`added GLP-1 receptor agonists as options when weight
`loss or risk of hypoglycaemia are major considerations.
`This decision was based on clinical data for the
`exendin-based GLP-1 receptor agonist exenatide, a
`molecule with 53% aminoacid identity with human GLP-1.
`Exenatide causes a decrease
`in HbA1c values of
`0·5–1·0%, and treatment is associated with weight loss1
`and with frequent gastrointestinal side-eff ects that tend to
`subside over time but can lead to treatment discontinuation.
`With elimination by glomerular fi ltration and a half-life of
`2·4 h, administration of exenatide twice a day 0–60 min
`before meals is recommended.6 The drug’s predominant
`eff ect
`is
`the
`reduction of postprandial glucose
`concentration, especially after breakfast and dinner.7
`Liraglutide is a human GLP-1 analogue with one
`aminoacid substitution (Arg34Lys) and a C-16 palmitic-
`acid side chain attached via a glutamyl spacer. These
`modifi cations result in slower absorption from sub-
`cutaneous tissue, reversible albumin binding, and
`resistance to GLP-1 inactivation by dipeptidyl peptidase-4.
`Unlike exenatide, liraglutide is 99% bound to albumin,
`with free liraglutide degraded by endogenous peptidases,
`and not via renal elimination.8 Liraglutide injection
`produces maximal concentrations within 10–14 h after
`administration, with a half-life of 13 h.9 Liraglutide has
`been developed as a once-a-day treatment for type 2
`diabetes, as an adjunct to lifestyle therapy and in
`combination with oral antidiabetic drugs.7
`Because the molecular structure, aminoacid sequence
`identity shared with human GLP-1, metabolism, and
`pharmacokinetics of exenatide and liraglutide diff er, we
`designed the liraglutide eff ect and action in our diabetes
`(LEAD-6) study to compare their effi cacy and safety. We
`report the results of the 26-week randomised comparator
`trial.
`
`Methods
`Participants
`Participants aged 18–80 years with type 2 diabetes were
`eligible if their HbA1c value was 7–11% and if they had a
`body-mass index (BMI) of 45·0 kg/m² or less on stable
`treatment with maximally tolerated doses of metformin,
`sulphonylurea, or both, for 3 months or more. Exclusion
`criteria included previous insulin treatment (except short-
`term treatment for intercurrent illness), previous exposure
`to exenatide or liraglutide, impaired liver or renal function,
`clinically signifi cant cardiovascular disease, retinopathy
`or maculopathy requiring acute treatment, uncontrolled
`hypertension (≥180/100 mm Hg), or cancer.
`All participants provided written consent before any
`procedure. The trial was done in accordance with the
`Declaration of Helsinki10 and Good Clinical Practice
`guidelines.11 Before trial initiation, the protocol, its
`amendments, consent form, and patient information
`sheets were approved by independent local ethics
`
`committees. The study is registered with ClinicalTrials.
`gov, number NCT00518882.
`
`Trial design and interventions
`This study was a 26-week randomised, open-label, active-
`comparator, parallel-group, multinational (132 offi ce-based
`sites across 15 countries) trial. Participants were screened
`for eligibility and enrolled by investigators. They were
`randomly assigned (1:1) to subcutaneous liraglutide
`1·8 mg once a day (Novo Nordisk A/S, Bagsvaerd,
`Denmark) or subcutaneous exenatide 10 μg twice a day
`(Byetta, Amylin Pharmaceuticals Inc, San Diego, CA,
`USA), and were stratifi ed by previous oral antidiabetic
`drug
`treatment. Randomisation was done with
`telephone-based or web-based systems. Participants were
`randomly assigned by investigators to the lowest available
`number from the range of numbers allocated to the site.
`The study began on Aug 24, 2007, and was completed on
`April 9, 2008.
`After randomisation, participants underwent a 2-week
`liraglutide dose-escalation period (during which the initial
`dose of 0·6 mg was increased by 0·6 mg a week to a
`maximum dose of 1·8 mg once a day) or 4-week exenatide
`dose-escalation period (during which 5 μg twice a day was
`increased to 10 μg twice a day after 4 weeks).6 This was
`followed by a 22–24-week maintenance period when no
`dose reduction of liraglutide or exenatide was allowed.
`Intolerance to these doses required study discontinuation.
`Background oral antidiabetic drugs were maintained at
`prestudy doses unless unacceptable hypoglycaemia
`occurred, in which case sulphonylurea doses could be
`reduced to no less than 50% of the starting dose.
`Both liraglutide and exenatide were injected in the
`upper arm, abdomen, or thigh with a pre-fi lled pen.
`Participants were encouraged to take liraglutide at the
`same time each day. Exenatide was administered
`0–60 min before breakfast and dinner (or before each of
`the two main daily meals, about 6 h or more apart).
`Participants completing this study could enrol in a
`52-week liraglutide 1·8-mg extension phase.
`
`Assessments and endpoints
`The primary effi cacy outcome was change in HbA1c values
`from baseline to week 26. Secondary effi cacy endpoints
`included proportion of patients reaching HbA1c targets
`(<7·0% and ≤6·5%), changes in fasting plasma glucose,
`self-measured
`7-point
`plasma
`glucose
`profi les,
`bodyweight, β-cell function, glucagon, blood pressure,
`and lipid profi les. Assays were done by central laboratories
`(MDS Pharma Services in Canada, France, Germany,
`Switzerland, and USA). Participants used Precision
`Xceed or Precision Xtra glucose meters
`(Abbott
`Diagnostics Inc, Abbott Park, IL, USA) to measure
`plasma glucose, and values were recorded in diaries.
`Overall treatment satisfaction was assessed with the
`Diabetes Treatment Satisfaction Questionnaire in a
`subgroup of participants.12 Overall treatment satisfaction
`
`40
`
`www.thelancet.com Vol 374 July 4, 2009
`
`MPI EXHIBIT 1114 PAGE 2
`
`

`

`Articles
`
`663 screened
`
`467 exposed
` 3 patients were not randomised
` but received treatment
`
`464 randomised
`
`Liraglutide 1·8 mg once a day
` 235 exposed
` 233 randomised
`
`Exenatide 10 μg twice a day
` 232 exposed
` 231 randomised
`
`33 withdrawn
` 23 adverse events
` 1 ineffective therapy
` 4 non-compliance with protocol
` 1 withdrawal criteria
` 4 other
`
`45 withdrawn
` 31 adverse events
` 0 ineffective therapy
` 3 non-compliance with protocol
` 1 withdrawal criteria
` 10 other
`
`202 completed
`
`187 completed
`
`Analysis populations
` 233 intention to treat
` 235 safety
`
`Analysis populations
` 231 intention to treat
` 232 safety
`
`Figure 1: Trial profi le
`Of the adverse events leading to withdrawal, nausea was the most common (14 patients in the liraglutide group
`and 16 in the exenatide group). Participants were exposed to treatment if they had received at least one dose of
`study medication.
`
`with treatment, country, and background oral antidiabetic
`drug as fi xed eff ects, and baseline Diabetes Treatment
`Satisfaction Questionnaire summary score as covariate.
`Missing data were not imputed.
`We analysed hypoglycaemic episodes using a
`generalised linear model with treatment, background
`oral antidiabetic drug, and country as fi xed eff ects. We
`compared other safety data with descriptive statistics.
`Signifi cance level was set at p<0·05, and data are
`expressed as least square means (SE) unless stated
`otherwise.
`
`Role of the funding source
`The sponsor was involved in study design, data collection,
`data review, and data analysis. All authors had full access
`to the data and had fi nal responsibility for the content of
`the manuscript; JBB had fi nal decision to submit for
`publication.
`
`Results
`464 participants were randomly assigned to treatment
`(fi gure 1). Three participants received treatment without
`randomisation (2 in the liraglutide group, 1 in exenatide
`group), and they were included in the safety but not
`intention-to-treat populations. 33 of 235 participants
`withdrew from liraglutide and 45 of 232 from exenatide
`treatment; withdrawal rates were not signifi cantly
`
`was based on six of the eight items in the questionnaire
`(each item was scored on a scale from +3 [better] to
`–3 [worse]).
`Safety variables included adverse events, vital signs,
`electrocardiogram, biochemical and haematological
`measures, and patient-reported hypoglycaemic episodes.
`A serious adverse event was defi ned as an adverse event
`that resulted in death, hospitalisation, disability, a birth
`defect, was life-threatening, or that required medical or
`surgical intervention to prevent one of the other outcomes.
`A severe adverse event was defi ned as an adverse event
`causing unacceptable and considerable interference with
`the patient’s daily activities. Major hypoglycaemic
`episodes were defi ned as requiring third-party assistance
`with food only, glucagon, or intravenous glucose. Minor
`episodes were defi ned as those that the participant could
`self-treat and for which the plasma glucose concentration
`was less than 3·1 mmol/L. At glucose concentrations of
`3·1 mmol/L or more, or in the absence of glucose
`measurements, episodes were regarded as symptoms
`only. Because of the nature of the antibody assay, analysis
`of emergent antibodies against liraglutide cannot be
`completed until participants have been through a
`washing-out period from therapy. Antibody data are not
`reported here and will be analysed once the liraglutide
`extension phase is completed.
`
`Statistical analysis
`The primary endpoint was the diff erence between
`treatment groups in HbA1c values from baseline to
`week 26. 163 individuals in each group were needed for
`85% power to detect a diff erence of 0·4% between groups
`(assuming a SD of 1·2%), a clinically meaningful margin
`for non-inferiority. Assuming a 25% drop-out rate,
`434 participants (217 per group) were needed at
`randomisation.
`Analyses of effi cacy outcomes were based on the
`intention-to-treat population. The primary endpoint was
`also analysed for the per-protocol population. We analysed
`most endpoints with the analysis of covariance (ANCOVA)
`with treatment, country, and current antidiabetic drug as
`explanatory variables, and baseline HbA1c values as
`covariate. We imputed missing values by carrying the last
`observation forward. We did hierarchical tests for
`non-inferiority and superiority of liraglutide and back-
`ground oral antidiabetic drugs versus exenatide and
`background oral antidiabetic drugs. We fi rst established
`non-inferiority and then tested superiority, each at 2·5%
`signifi cance level. We assumed non-inferiority if the
`upper limit of the two-sided 95% CI for treatment
`diff erence was less than 0·4%, and superiority if the
`upper limit was less than 0. We compared the proportions
`of patients achieving HbA1c target values using logistic
`regression with treatment, country, and background oral
`antidiabetic drug as explanatory variables, and baseline
`HbA1c values as covariate. We developed estimates of
`overall treatment satisfaction from an ANCOVA model
`
`www.thelancet.com Vol 374 July 4, 2009
`
`41
`
`MPI EXHIBIT 1114 PAGE 3
`
`

`

`Articles
`
`Men
`Age (years)
`Race
`White
`Asian/Pacifi c Islander
`Black*
`Other
`Hispanic or Latin American ethnic
`origin
`Weight (kg)
`Body-mass index (kg/m²)
`Duration of diabetes (years)
`Fasting C-peptide (nmol/L)
`Prestudy antidiabetic treatment
`Metformin and SU combination
`SU alone
`Metformin alone
`HbA1c
`Fasting plasma glucose (mmol/L)
`Systolic blood pressure (mm Hg)
`Diastolic blood pressure (mm Hg)
`
`Liraglutide
`1·8 mg once a
`day (n=233)
`
`Exenatide
`10 μg twice a
`day (n=231)
`
`114 (49%)
`56·3 (9·8)
`
`127 (55%)
`57·1 (10·8)
`
`216 (93%)
`1 (<1%)
`13 (6%)
`3 (1%)
`32 (14%)
`
`210 (91%)
`5 (2%)
`12 (5%)
`4 (2%)
`25 (11%)
`
`93·1 (20·1)
`32·9 (5·5)
`8·5 (6·2)
`1·25 (0·56)
`
`93·0 (19·5)
`32·9 (5·7)
`7·9 (5·9)
`1·26 (0·58)
`
`145 (62%)
`24 (10%)
`64 (27%)
`8·2% (1·0%)
`9·8 (2·5)
`132 (16·2)
`79·6 (8·4)
`
`147 (64%)
`21 (9%)
`63 (27%)
`8·1% (1·0%)
`9·5 (2·4)
`134 (17·0)
`78·9 (8·9)
`
`Data are mean (SD) or number (%). HbA1c=glycosylated haemoglobin.
`SU=sulphonylurea. *Includes African-American.
`
`Table 1: Baseline demographic and disease characteristics
`
`diff erent between groups. Adverse events were the most
`common reason for withdrawal in both groups. The
`characteristics of the study population were typical for
`participants with
`type 2 diabetes, and baseline
`characteristics were well matched between treatment
`groups (table 1).
`HbA1c values decreased more in the group treated with
`liraglutide 1·8 mg once a day than in that treated with
`exenatide 10 μg twice a day over 26 weeks (fi gure 2A). The
`mean change from baseline to week 26 was signifi cantly
`greater in the group treated with liraglutide than in that
`treated with exenatide (–1·12% [0·08] vs –0·79% [0·08];
`estimated treatment diff erence [ETD] –0·33; 95% CI
`–0·47 to –0·18; fi gure 2B). Reduction of HbA1c values with
`liraglutide was statistically superior to that seen with
`exenatide. Diff erences in HbA1c values between treatment
`groups did not depend on baseline therapy, BMI, country,
`sex, ethnic origin, or age because the interaction eff ects
`were not signifi cant (p>0·05). The signifi cance of
`treatment-by-race interaction (p=0·0256) might be due to
`the small number of non-white participants (table 1). Data
`in the intention-to-treat population were similar to those
`in the per-protocol population (change from baseline to
`week 26 HbA1c: liraglutide –1·16% [0·09] vs exenatide
`–0·87% [0·09]; ETD –0·29%; 95% CI –0·45 to –0·13;
`p<0·0001). We confi rmed robustness of the ETD using
`last-observation
`carried-forward
`data
`with
`repeated-measures analysis and multiple imputation
`
`methods (data not shown). Mean reductions in HbA1c
`values were generally greater for the liraglutide group
`than for the exenatide group across the spectrum of HbA1c
`values. However, the diff erence was greatest for patients
`with baseline HbA1c of 10% or more (liraglutide –2·4%
`[SE 0·21] vs exenatide –1·2% [0·37]).
`The proportion of participants achieving HbA1c targets
`was signifi cantly higher in the liraglutide than in the
`exenatide group (target of <7%: 54% vs 43%; odds ratio
`[OR] 2·02; 95% CI 1·31 to 3·11; target of ≤6·5%: 35% vs
`21%; OR 2·73; 95% CI 1·68 to 4·43; fi gure 2C). Liraglutide
`also reduced fasting plasma glucose from baseline
`signifi cantly more than did exenatide (–1·61 mmol/L
`[0·20] vs –0·60 mmol/L [0·20]; ETD –1·01 mmol/L;
`95% CI –1·37 to –0·65; p<0·0001; fi gure 2D). In contrast,
`exenatide reduced postprandial plasma glucose increment
`more than did liraglutide (self-measured with 7-point
`plasma glucose profi les; fi gure 2E) after breakfast and
`dinner (breakfast: ETD 1·33 mmol/L; 95% CI 0·80 to
`1·86; p<0·0001; dinner: ETD 1·01 mmol/L; 95% CI
`0·44 to 1·57; p=0·0005); treatment diff erences after lunch
`were not signifi cant.
`Liraglutide and exenatide were associated with similar
`weight losses (liraglutide –3·24 kg [0·33] vs exenatide
`–2·87 kg [0·33]; ETD –0·38 kg; 95% CI –0·99 to 0·23;
`p=0·2235; fi gure 2F) and similar proportions of
`participants who lost weight (liraglutide 78% [182 of 233]
`vs exenatide 76% [176 of 231]). Mean reductions in HbA1c
`values were clinically meaningful irrespective of whether
`participants lost weight (weight loss: liraglutide –1·3% vs
`exenatide –0·9%; no weight loss: liraglutide –1·0% vs
`exenatide –0·5%).
`Table 2 shows changes in islet function, blood pressure,
`and lipids. Increases in fasting insulin and the associated
`homoeostasis model assessment index of β-cell function
`(HOMA-B) were signifi cantly greater for the liraglutide
`than for the exenatide group. Treatment diff erences for
`fasting C-peptide or proinsulin-to-insulin ratio were not
`signifi cant. Fasting glucagon and blood pressure
`decreased with both treatments, and diff erences between
`treatments were not signifi cant for fasting glucagon or
`either systolic or diastolic blood pressures. Reductions of
`triglycerides and free fatty acid values were signifi cantly
`greater in the liraglutide group than in the exenatide
`group, and increases in very low-density lipoprotein
`cholesterol were smaller in the liraglutide group than in
`the exenatide group.
`Overall treatment satisfaction was signifi cantly better
`in the liraglutide group (n=161) than in the exenatide
`group (n=143) (15·18 [0·58] vs 13·30 [0·58]; ETD 1·89;
`95% CI 0·85 to 2·92; p=0·0004).
`Despite an overall lower reporting of adverse events in
`the liraglutide group than in the exenatide group (74·9%
`vs 78·9%), the liraglutide group had more serious and
`severe adverse events (serious: 5·1% vs 2·6%; severe:
`7·2% vs 4·7%; table 3). Serious adverse events showed
`no consistent pattern for system organ class and only
`
`42
`
`www.thelancet.com Vol 374 July 4, 2009
`
`MPI EXHIBIT 1114 PAGE 4
`
`

`

`Articles
`
`Liraglutide
`Baseline 8·2%
`
`Exenatide
`Baseline 8·1%
`
`p<0·0001
`
`Liraglutide
`Exenatide
`
`B
`0
`
`–0·5
`
`–1·0
`
`Change in HbA1C (%)
`
`–1·5
`
`D
`10·5
`
`10·0
`
`9·5
`
`9·0
`
`8·5
`
`8·0
`
`7·5
`
`7·0
`
`FPG (mmol/L)
`
`Liraglutide
`Exenatide
`
`0
`
`2
`
`4
`
`6
`
`8
`
`10
`
`14
`12
`Time (weeks)
`
`16
`
`18 20
`
`22
`
`24
`
`26
`
`HbA1C <7·0%
`
`HbA1C ≤6·5%
`
`p=0·0015
`
`p<0·0001
`
`Liraglutide
`
`Exenatide
`
`Liraglutide
`
`Exenatide
`
`A
`8·5
`
`8·0
`
`7·5
`
`7·0
`
`6·5
`
`C
`100
`
`90
`
`80
`70
`60
`50
`
`40
`30
`20
`10
`0
`
`HbA1C (%)
`
`Patients (%)
`
`0
`
`2
`
`4
`
`6
`
`8
`
`10
`
`14
`12
`16
`Time (weeks)
`
`18 20
`
`22
`
`24
`
`26
`
`Liraglutide
`Exenatide
`
`F
`98
`
`96
`
`94
`
`92
`
`90
`
`88
`
`86
`
`Bodyweight (kg)
`
`Liraglutide
`Baseline
`Week 26
`
`Exenatide
`Baseline
`Week 26
`
`Before dinner
`Dinner+90 min
`
`Bedtime
`
`0
`
`2
`
`4
`
`6
`
`8
`
`10
`
`16
`14
`12
`Time (weeks)
`
`18 20
`
`22
`
`24
`
`26
`
`
`
`Before lunchLunch+90 min
`
`Time
`
`E
`14
`
`13
`
`12
`
`11
`
`10
`
`9 8 7 6
`
`Breakfast+90 min
`Before breakfast
`
`Self-measured plasma glucose (mmol/L)
`
`Figure 2: Effi cacy of treatment with liraglutide 1·8 mg once a day or exenatide 10 μg twice a day
`(A) Glycosylated haemoglobin (HbA1c) values from baseline to week 26. (B) Change in HbA1c values from baseline to week 26. (C) Percentage of patients achieving
`HbA1c target values. (D) Fasting plasma glucose (FPG) concentrations from baseline to week 26. (E) 7-point self-measured plasma glucose profi les. (F) Bodyweight
`from baseline to week 26. Data are mean (1·96 SE) unless stated otherwise, with last observation carried forward (except for panel E, observed case).
`
`one event (severe hypoglycaemia requiring medical
`attention in the exenatide group) was judged probably
`related to study medication by the investigator. The
`most frequent severe adverse events were dyspepsia in
`the liraglutide group (n=3) and nausea in the exenatide
`group (n=4). The distribution of most adverse events
`
`was similar between groups (table 3). Although the
`incidence of nausea was similar initially, it was less
`persistent with liraglutide (estimated treatment rate
`ratio 0·448 for liraglutide vs exenatide; proportion of
`participants with nausea at week 26, 5 of 202 [3%] vs
`16 of 186 [9%]; fi gure 3).
`
`www.thelancet.com Vol 374 July 4, 2009
`
`43
`
`MPI EXHIBIT 1114 PAGE 5
`
`

`

`concentrations were reported. Liraglutide was continued
`for another 10 weeks until the patient was withdrawn
`because of lung adenocarcinoma (unlikely to be related to
`treatment). No episodes of acute pancreatitis were reported
`with either agent.
`No major hypoglycaemia occurred with liraglutide but
`there were two episodes in patients receiving exenatide
`and a sulphonylurea. The proportion of patients who had
`minor hypoglycaemia (26% [60 out of 235] with liraglutide
`vs 34% [78 out of 232] with exenatide) and event rate for
`minor hypoglycaemia (1·932 vs 2·600 events per
`participant per year; rate ratio 0·55, 95% CI 0·34 to 0·88;
`p=0·0131) were lower with liraglutide than with exenatide,
`with greater diff erences between treatments during the
`evening (fi gure 4). The proportion of patients who had
`episodes of minor hypoglycaemia was lower in the
`subgroups using metformin as background therapy (6%
`[4 out of 64] and 11% [7 out of 63] for liraglutide and
`exenatide groups, respectively) than in those taking a
`sulphonylurea with or without metformin (33% [56 out
`of 171] and 42% [71 out of 169], respectively). However,
`most patients could continue sulphonylurea treatment at
`the dose used in the period before enrolment (liraglutide
`89% [150 of 169] and exenatide 85% [142 of 168]).
`in all
`Calcitonin concentrations were monitored
`participants to assess any eff ect of GLP-1 receptor
`agonists on C-cell function. At baseline, mean calcitonin
`concentrations were less than 1 ng/L, below the upper
`normal reference range for both women (5·0 ng/L) and
`men (8·4 ng/L). Small decreases in calcitonin occurred
`during the trial in both treatment groups, without
`signifi cant diff erence between groups at any timepoint.
`Heart rates increased slightly in both treatment groups
`(liraglutide 3·28 [0·83] beats per minute; exenatide 0·69
`[0·84] beats per minute) but the increase was signifi cantly
`greater for liraglutide (ETD: 2·58; 95% CI 1·03 to 4·13;
`p=0·0012).
`
`Discussion
`This trial provides a direct comparison of effi cacy and
`safety between liraglutide and exenatide, both of which
`interact with the GLP-1 receptor but diff er in aminoacid
`identity with human GLP-1, frequency and timing of
`administration,
`clearance, and especially pharma-
`cokinetics. The results show that liraglutide provides
`superior overall glycaemic control on the basis of HbA1c
`data.
`We show HbA1c reductions that are consistent with
`those from other studies with liraglutide and exenatide.
`Liraglutide 1·8 mg once a day reduced HbA1c values by
`1·12% in the present study compared with 1·14% in the
`LEAD-3 monotherapy study13 and 1·00–1·48% in LEAD
`studies in which it was administered together with oral
`antidiabetic drugs.14–17 Exenatide 10 μg twice a day reduced
`HbA1c values by 0·79%, which is consistent with
`reductions of 0·78–0·89% reported elsewhere.18–21
`Greater mean reductions were shown in a recent study
`
`Articles
`
`Liraglutide
`
`Exenatide
`
`p value
`
`Estimated treatment
`diff erence
`(liraglutide–exenatide)
`(95% CI)
`
`12·43 (6·93)
`0·05 (0·05)
`0·00 (0·03)
`
`–1·38 (6·92)
`–0·02 (0·05)
`–0·02 (0·03)
`
`13·81 (0·94 to 26·68)
`0·07 (–0·02 to 0·16)
`0·02 (–0·03 to 0·08)
`
`0·0355
`0·1340
`0·4309
`
`32·12% (6·75%)
`
`2·74% (6·75%)
`
`–19·44(5·18)
`
`–12·33 (5·12)
`
`29·37% (16·81% to
`41·93%)
`–7·12 (–16·66 to 2·43)
`
`<0·0001
`
`0·1436
`
`–2·51 (1·15)
`
`–2·00 (1·18)
`
`–0·51 (–2·66 to 1·64)
`
`0·6409
`
`–1·05 (0·71)
`
`–1·98 (0·71)
`
`0·93 (–0·37 to 2·23)
`
`0·1610
`
`–0·20 (0·07)
`–0·44 (0·06)
`0·20 (0·04)
`–0·04 (0·02)
`–0·41 (0·10)
`–0·17 (0·02)
`–0·06 (0·02)
`
`–0·09 (0·07)
`–0·40 (0·06)
`0·27 (0·04)
`–0·05 (0·02)
`–0·23 (0·10)
`–0·10 (0·02)
`–0·03 (0·02)
`
`–0·11 (–0·23 to 0·02)
`–0·04 (–0·15 to 0·06)
`–0·07 (–0·13 to –0·01)
`0·01 (–0·02 to 0·04)
`–0·18 (–0·37 to 0·00)
`–0·07 (–0·11 to –0·03)
`–0·02 (–0·05 to 0·01)
`
`0·0946
`0·4412
`0·0277
`0·5105
`0·0485
`0·0014
`0·1119
`
`Islet function
`Fasting insulin (pmol/L)
`Fasting C-peptide (nmol/L)
`Fasting proinsulin-to-insulin
`ratio
`HOMA-B
`
`Fasting glucagon (ng/L)
`Blood pressure
`Systolic blood pressure
`(mm Hg)
`Diastolic blood pressure
`(mm Hg)
`Lipid profi les
`Total cholesterol (mmol/L)
`LDL cholesterol (mmol/L)
`VLDL cholesterol (mmol/L)
`HDL cholesterol (mmol/L)
`Triglycerides (mmol/L)
`Free fatty acids (mmol/L)
`Apolipoprotein B (g/L)
`
`Data are least square means (SE). HOMA-B=homoeostasis model assessment β-cell function. LDL=low-density
`lipoprotein. HDL=high-density lipoprotein. VLDL=very low-density lipoprotein.
`
`Table 2: Change in indices of islet function and cardiovascular risk from baseline to week 26
`
`p <0·0001
`
`Liraglutide 1·8 mg once a day
`Exenatide 10 μg twice a day
`
`20
`
`18
`
`16
`
`14
`
`12
`
`10
`
`8
`
`6 4 2 0
`
`Patients (%)
`
`0
`
`2
`
`4
`
`6
`
`8
`
`10
`
`12
`14
`Time (weeks)
`
`16
`
`18
`
`20
`
`22
`
`24
`
`26
`
`Figure 3: Proportion of patients with an episode of nausea between baseline and week 26
`
`One episode of mild pancreatitis, which the investigator
`regarded as chronic and unlikely to be related to study
`drug, was diagnosed after 88 days of liraglutide therapy in
`a 69-year-old man with a history of abdominal distension
`and hypercholesterolaemia. No data for pancreatic enzyme
`
`44
`
`www.thelancet.com Vol 374 July 4, 2009
`
`MPI EXHIBIT 1114 PAGE 6
`
`

`

`Articles
`
`Liraglutide
`(n=235)
`
`Exenatide
`(n=232)
`
`Serious adverse events*
`Severe adverse event
`Gastrointestinal disorders
`Nervous system disorders
`Infections and infestations
`Cardiac disorders
`General disorders and administration-site conditions
`Hepatobiliary disorders
`Musculoskeletal and connective tissue disorders
`Neoplasms (benign, malignant, and unspecifi ed, including
`cysts and polyps)
`Reproductive system and breast disorders
`Metabolism and nutrition disorders
`Respiratory, thoracic, and mediastinal disorders
`Vascular disorders
`Adverse events (of any severity) reported by more than 5% of
`patients
`Gastrointestinal disorders
`Constipation
`Diarrhoea
`Dyspepsia
`Nausea
`Vomiting
`Infections and infestations
`Bronchitis
`Nasopharyngitis
`Upper respiratory tract infection
`Nervous system disorders
`Headache
`Musculoskeletal and connective tissue disorders
`Back pain
`Metabolism and nutrition disorders
`Respiratory, thoracic, and mediastinal disorders
`General disorders and administration-site conditions
`Injury, poisoning, and procedural complications
`Skin and subcutaneous tissue disorders
`
`12 (5·1%)
`17 (7·2%)
`8 (3·4%)
`3 (1·3%)
`2 (0·9%)
`1 (0·4%)
`1 (0·4%)
`1 (0·4%)
`1 (0·4%)
`1 (0·4%)
`
`1 (0·4%)
`0
`0
`0
`176 (74·9%)
`
`107 (45·5%)
`12 (5·1%)
`29 (12·3%)
`21 (8·9%)
`60 (25·5%)
`14 (6·0%)
`78 (33·2%)
`12 (5·1%)
`27 (11·5%)
`15 (6·4%)
`43 (18·3%)
`21 (8·9%)
`32 (13·6%)
`14 (6·0%)
`28 (11·9%)
`22 (9·4%)
`21 (8·9%)
`11 (4·7%)
`8 (3·4%)
`
`6 (2·6%)
`11 (4·7%)
`5 (2·2%)
`0
`1 (0·4%)
`2 (0·9%)
`1 (0·4%)
`0
`0
`0
`
`0
`3 (1·3%)
`1 (0·4%)
`1 (0·4%)
`183 (78·9%)
`
`99 (42·7%)
`6 (2·6%)
`28 (12·1%)
`11 (4·7%)
`65 (28·0%)
`23 (9·9%)
`85 (36·6%)
`16 (6·9%)
`31 (13·4%)
`14 (6·0%)
`37 (15·9%)
`24 (10·3%)
`29 (12·5%)
`8 (3·4%)
`32 (13·8%)
`14 (6·0%)
`21 (9·1%)
`14 (6·0%)
`16 (6·9%)
`
`Data are number (%) of participants. *Liraglutide group: adenocarcinoma of the pancreas, adenocarcinoma of the lung,
`thyroid neoplasm, coronary artery stenosis, supraventricular tachycardia, cerebellar infarction, sciatica, diplopia,
`cholelithiasis, pneumonia, postmenopausal vaginal bleeding, and asthma. Exenatide group: acute myocardial
`infarction, coronary artery disease, cerebrovascular accident, cataract, arthralgia, and hypoglycaemia.
`
`Table 3: Treatment-emergent adverse events
`
`fatal
`and
`severe
`Acute pancreatitis—including
`episodes—has been spontaneously reported in patients
`taking exenatide, raising concerns about a potential
`association with GLP-1 receptor agonists. However, obesity,
`hypertriglyceridaemia, and gallstones are all known risk
`factors for acute pancreatitis, and are all associated with
`type 2 diabetes. In fact, acute pancreatitis seems t