REVIEW
`
`review article
`
`Diabetes, Obesity and Metabolism 18: 317–332, 2016.
`© 2015 The Authors. Diabetes, Obesity and Metabolism published by John Wiley & Sons Ltd.
`
`Review of head-to-head comparisons of glucagon-like peptide-1
`receptor agonists
`Sten Madsbad
`
`Department of Endocrinology, Hvidovre Hospital, University of Copenhagen, Hvidovre, Denmark
`
`Currently, six glucagon-like peptide-1 receptor agonists (GLP-1RAs) are approved for treating type 2 diabetes. These fall into two classes based on
`their receptor activation: short-acting exenatide twice daily and lixisenatide once daily; and longer-acting liraglutide once daily, exenatide once weekly,
`albiglutide once weekly and dulaglutide once weekly. The phase III trial of a seventh GLP-1RA, taspoglutide once weekly, was stopped because of
`unacceptable adverse events (AEs). Nine phase III head-to-head trials and one large phase II study have compared the efficacy and safety of these
`seven GLP-1RAs. All trials were associated with notable reductions in glycated haemoglobin (HbA1c) levels, although liraglutide led to greater decreases
`than exenatide formulations and albiglutide, and HbA1c reductions did not differ between liraglutide and dulaglutide. As the short-acting GLP-1RAs delay
`gastric emptying, they have greater effects on postprandial glucose levels than the longer-acting agents, whereas the longer-acting compounds reduced
`plasma glucose throughout the 24-h period studied. Liraglutide was associated with weight reductions similar to those with exenatide twice daily but
`greater than those with exenatide once weekly, albiglutide and dulaglutide. The most frequently observed AEs with GLP-1RAs were gastrointestinal
`disorders, particularly nausea, vomiting and diarrhoea. Nauseaoccurred less frequently, however, with exenatide once weekly and albiglutide than
`exenatide twice daily and liraglutide. Both exenatide formulations and albiglutide may be associated with higher incidences of injection-site reactions
`than liraglutide and dulaglutide. GLP-1RA use in clinical practice should be customized for individual patients, based on clinical profile and patient
`preference. Ongoing assessments of novel GLP-1RAs and delivery methods may further expand future treatment options.
`Keywords: albiglutide, dulaglutide, exenatide, GLP-1 receptor agonist, liraglutide, lixisenatide, taspoglutide, type 2 diabetes
`
`Date submitted 25 June 2015; date of first decision 16 August 2015; date of final acceptance 22 October 2015
`
`Introduction
`The early and intensive treatment of people with type 2 dia-
`betes (T2D) is of key importance for reducing the risk of late
`diabetic complications, such as microvascular disease [1]. T2D
`is linked to obesity [2], and the cornerstone of treatment is
`lifestyle changes to promote weight loss and increase exercise
`[3]; however, because of the progressive nature of T2D, phar-
`macological therapy to address hyperglycaemia becomes nec-
`essary in almost all patients. Pharmacological treatment is,
`unfortunately, often associated with side effects such as weight
`gain (e.g. sulphonylureas, insulin and thiazolidinediones) [4,5],
`hypoglycaemia (e.g. sulphonylureas and insulin) [6,7], gas-
`trointestinal (GI) discomfort [e.g. metformin and glucagon-like
`peptide-1 receptor agonists (GLP-1RAs)] and genital infec-
`tions [sodium-glucose co-transporter 2 (SGLT2) inhibitors]
`[8–10]. Notwithstanding the GI discomfort with GLP-1RAs,
`their introduction over the last decade has greatly improved
`treatment of T2D [11–14].
`Human GLP-1 is a member of the incretin family of
`glucoregulatory hormones, and is secreted in response to
`
`Correspondence to: Sten Madsbad, Department of Endocrinology, Hvidovre Hospital, University
`of Copenhagen, Kettegaards Alle, 2650 Hvidovre, Denmark.
`E-mail: sten.madsbad@regionh.dk
`
`This is an open access article under the terms of the Creative Commons Attribution-
`NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided
`the original work is properly cited, the use is non-commercial and no modifications or adaptations
`are made.
`
`food ingestion [15,16]. Glucagon-like peptide-1 has multiple
`effects that are desirable in the treatment of T2D, includ-
`ing: glucose-dependent increased insulin secretion; glucose-
`dependent decreased glucagon secretion; delayed gastric
`emptying; increased satiety; and, as shown in some animal
`studies, protection of 𝛽-cell mass [17,18].
`Unfortunately, although intravenously infused GLP-1 can
`normalize plasma glucose concentrations in people with T2D
`[19,20], it has an extremely short half-life (1–2 min) [16] that
`limits its therapeutic value [21]. Multiple GLP-1RAs have been
`developed to recapitulate the physiological effects of GLP-1
`but with an extended duration of action (achieved by vari-
`ous changes to the molecular structure) compared with the
`native peptide [22].The present review examines the avail-
`able evidence from published head-to-head clinical trials with
`GLP-1RAs, and contrasts the relative clinical benefits of the
`short- and longer-acting agents.
`
`Characteristics of GLP-1RAs
`Seven GLP-1RAs are included in the present review, all of which
`have been studied in phase III clinical trials. The GLP-1RAs
`are: exenatide twice daily (Byetta®, AstraZeneca; approved in
`Europe in November 2006 and 28 May 2005 in USA [23,24]);
`liraglutide (Victoza®, Novo Nordisk; approved in Europe in
`June 2009 and 25 January 2010 in USA [25,26]); exenatide
`once weekly (Bydureon®, AstraZeneca; approved in Europe in
`
`MPI EXHIBIT 1111 PAGE 1
`
`

`

`review article
`
`June 2011 and 26 January 2012 in USA [27,28]); lixisenatide
`(Lyxumia®, Sanofi; approved in Europe in February in 2013
`[29] but not in the USA); albiglutide (Eperzan® and Tanzeum®,
`GlaxoSmithKline; approved in March 2014 in Europe and April
`2014 in USA [30,31]); dulaglutide (Trulicity™, Lilly; approved
`in Europe in November 2014 and September 2014 in USA
`[32,33]); and taspoglutide (Ipsen/Roche). These have all now
`been approved for use in T2D, with the exception of taspoglu-
`tide, the development of which was halted because of serious
`hypersensitivity reactions and GI adverse events (AEs) during
`clinical trials; however, the available data for this compound
`are included in the present review to give a full picture of the
`GLP-1RA family.
`As a drug class, the GLP-1RAs have proven efficacy for low-
`ering glycated haemoglobin (HbA1c) and decreasing weight
`in T2D, with a reduced risk of hypoglycaemia compared with
`insulin or sulphonylureas [34]. These characteristics under-
`lie the inclusion of GLP-1RAs in various clinical practice
`guidelines. Their use as dual therapy with metformin after
`first-line metformin and as triple therapy (in combination with
`metformin and a sulphonylurea/thiazolidinedione/insulin)
`is part of the European Association for the Study of Dia-
`betes/American Diabetes Association recommendations [34].
`GLP-1RAs are recommended as monotherapy, dual therapy
`and triple therapy by the American Association of Clini-
`cal Endocrinologists/American College of Endocrinology
`guidelines [35]. Nonetheless, they differ substantially in their
`molecular structure and degree of homology to endogenous
`GLP-1, both in their chemical and physiological properties
`and in their durations of action (Table 1).
`Several GLP-1RAs (exenatide twice daily, exenatide once
`weekly and lixisenatide) are based on the exendin-4 molecule,
`a peptide with 53% identity to native GLP-1 [36,42,43], while
`others, such as liraglutide, albiglutide, dulaglutide and tas-
`poglutide are classified as GLP-1RA analogues with 97, 95,
`90 and 93% identity, respectively, to native GLP-1 [38–40].
`The GLP-1RAs are, in addition, commonly considered to fall
`into two different classes based on their duration of receptor
`activation. The short-acting compounds, delivering short-lived
`receptor activation, comprise exenatide twice daily and lixise-
`natide once daily. The long-acting compounds, which activate
`the GLP-1 receptor continuously at their recommended dose,
`comprise liraglutide once daily, and the once-weekly formu-
`lations of exenatide, albiglutide, dulaglutide and taspoglutide
`(Table 1). These different durations of action largely explain
`variations among GLP-1RAs in their impact on fasting plasma
`glucose (FPG), 24-h glucose profile and postprandial plasma
`glucose (PPG) levels [60,61]. Delayed gastric emptying, for
`example, is more strongly associated with short-acting than
`longer-acting GLP-1RAs (Figure 1), and this may underlie the
`greater effects on PPG observed with short-acting GLP-1RAs.
`Meanwhile, the greater half-lives of the longer-acting com-
`pounds allow enhanced effects on the whole 24-h glucose
`level, including FPG. Longer-acting GLP-1RAs do not sig-
`nificantly affect gastric motility. Instead, they exert more of
`their effect via the pancreas, increasing insulin secretion and
`inhibiting glucagon secretion via paracrine release of somato-
`statin (Figure 1) [22].
`
`DIABETES, OBESITY AND METABOLISM
`
`Not only do GLP-1RAs differ from each other in terms of
`their duration of action [39,46–51], they also show varying
`levels of affinity for the GLP-1 receptor [62]. This difference
`between GLP-1RAs is also evident in their varying efficacy with
`regard to HbA1c reduction and weight loss, and differing tol-
`erability profiles and potential for immunogenicity [22,63–65].
`It is important to understand these specific characteristics to
`appropriately tailor the choice of GLP-1RA to the individual
`patient. Head-to-head clinical trials are the best way to eluci-
`date variations in efficacy and tolerability, and a number of such
`studies have been conducted with GLP-1RAs in T2D.
`
`Head-to-head Comparison Trials
`To date, the results from nine phase III randomized trials
`that directly compare different pairs of GLP-1RAs have been
`published [12–14,40,54–56,66,67]. An overview of the designs
`of these studies is provided in Table 2. One large phase II study,
`comparing liraglutide and lixisenatide pharmacodynamics, is
`also included [61].
`Of the GLP1-RAs in the head-to-head trials, exenatide twice
`daily and liraglutide were the most common comparators
`(Table 2). The majority of the phase III studies included in
`the present review were of ∼6 months’ duration, although sev-
`eral also had extension phases to give trial durations up to
`12 months (Table 2). All of the phase III trials examined changes
`in HbA1c as the primary endpoint; the phase II study by Kapitza
`et al. [61], however, used changes in PPG exposure as the pri-
`mary endpoint.
`In general, baseline characteristics were similar across trial
`populations and between treatment groups within individual
`trials (Table 3). The mean age of participants ranged from 55
`to 61 years across the studies, with mean duration of diabetes
`ranging from 6 to 9 years. Mean baseline HbA1c levels were
`in the range of 8.0 (64 mmol/mol) to 8.7% (72 mmol/mol)
`across the studies, except for the phase II study (in which
`HbA1c levels were lower) [61]. Glucose concentrations in the
`range of 9.1–9.9 mmol/l were determined in plasma and serum
`samples, and mean baseline weight was consistently in the range
`91–102 kg, except in an Asian study, in which mean weight
`was lower [56] (Table 3). Differences in key study outcomes are
`discussed in the following sections.
`
`Glycaemic Measurements
`In all of the head-to-head trials, the GLP-1RAs studied led to
`notable reductions in HbA1c.
`These reductions ranged between 0.3 (3 mmol/mol) and
`1.9% (21 mmol/mol). Although data are not comparable across
`studies because of differences in study design and patient
`cohorts, there were some important differences between treat-
`ment arms in the magnitude of HbA1c reductions (Figure 2).
`In particular, in the DURATION-1, DURATION-5 studies and
`the study by Ji et al. [54–56], exenatide once weekly produced
`both more consistent and greater reductions in HbA1c levels
`than did exenatide twice daily (p ≤ 0.0023). In the GetGoal-X
`study [12], meanwhile, exenatide twice daily showed a numeri-
`cally greater HbA1c reduction than lixisenatide. Liraglutide, in
`
`318 Madsbad
`
`Volume 18 No. 4 April 2016
`
`
`
`MPI EXHIBIT 1111 PAGE 2
`
`

`

`DIABETES, OBESITY AND METABOLISM
`
`review article
`
`participants[40]
`Detectedin49%of
`
`Unpublishedresults
`respectively[51]
`30mgdoses,
`
`4,6and8hat1,8and
`
`165h[51]
`
`binding[45]
`allowsgreaterreceptor
`proteasesandalso
`DPP-4andbyserine
`tohindercleavageby
`Modificationsdesigned
`
`*Valuebasedonsimilaritytoexenatide.
`AE,adverseevent;DPP-4,dipeptidylpeptidase-4;GLP-1,glucagon-likepeptide-1;GLP-1RA,GLP-1receptoragonist;IgG4,immunoglobulin4;Tmax,timetomaximumplasmaconcentration.
`
`ofpatients[13]
`antidrugantibodiesin1%
`dulaglutideneutralising
`participantsand
`antibodiesin1%of
`Dulaglutideantidrug
`
`0.107l/h,respectively[29]
`statewas0.073and
`
`0.75mgand1.5mgatsteady
`
`albiglutide[14]
`ofparticipantstreatedwith
`Antibodiesdevelopedin3.7%
`
`[59]
`with20μgtwicedaily
`dailyand71%treated
`treatedwith10μgonce
`•43%ofparticipants
`[58]
`with20μgoncedaily
`participantstreated
`•56–60%of
`
`after26weeks[57]
`1.8mg,respectively,
`liraglutide1.2and
`low-titreantibodiesto
`participantshad
`8.7and8.3%of
`
`Antibodiesdevelopedin:
`
`Fromsixphase3studies,
`
`reportedAEs[54–56].
`
`[54–56].
`onceweeklycomparedwithexenatidetwicedaily
`commonandtitreswerehigherwithexenatide
`•Inhead-to-headstudies,antibodiesweremore
`
`•Antibodiesdidnotcorrelatewithratesof
`
`Antibodyformation
`
`67ml/h[53]
`
`21.2–28.5l/h[49]
`
`1.2l/h[52]
`
`Unpublishedresults
`
`9.1l/h[52]
`
`Clearance
`
`24–72h[39]
`
`72–96h[50]
`
`1.25–2.25h[49]
`
`≈9–12h[48]
`
`2.1–5.1hduringthefirst
`
`2.1h[46]
`
`48h[47]
`
`Tmax
`
`≈5days[39]
`
`6–8days[50]
`
`2.7–4.3h[49]
`
`11–15h[48]
`
`Half-lifeisan
`
`2.4h[46]
`
`Half-life
`
`prolongingactivity[39]
`renalclearanceand
`IgG4molecule,limiting
`moleculeisfusedtoan
`TheGLP-1portionofthe
`
`albumin[44]
`
`theC-terminal[43]
`sixlysineresiduesat
`residueandadditionof
`deletionofoneproline
`modifiedbythe
`
`GLP-1dimerfusedto
`
`Basedonexenatide,butis
`
`[38]
`heptamerformation
`albuminbindingand
`C-16fattyacidconfers
`
`microspheres[42]
`biodegradablepolymer
`
`position2byglycine[41]
`substitutionofalaninein
`largelyduetothe
`
`Encapsulatedin
`
`ResistanttoDPP-4cleavage,
`
`Propertiesofthedrug
`
`6–7weeks[47]
`concentrationsat
`steadystate
`unpublishedresultbut
`
`93%[40]
`
`onceweekly
`Taspoglutide
`
`90%[39]
`
`onceweekly
`Dulaglutide
`
`95%[38]
`
`onceweekly
`Albiglutide
`
`≈50%*
`oncedaily
`Lixisenatide
`
`97%[37]
`
`oncedaily
`Liraglutide
`
`53%[36]
`
`onceweekly
`Exenatide
`
`53%[36]
`
`Exenatidetwicedaily
`
`nativeGLP-1
`sequencesimilarityto
`
`Percentageaminoacid
`
`Table1.Comparativecharacteristicsoftheglucagon-likepeptide-1receptoragonists.
`
`Volume 18 No. 4 April 2016
`
`doi:10.1111/dom.12596 319
`
`
`
`MPI EXHIBIT 1111 PAGE 3
`
`

`

`review article
`
`DIABETES, OBESITY AND METABOLISM
`
`Figure 1. Gastric-emptying effects of short-acting versus longer-acting glucagon-like peptide-1 receptor agonists (GLP-1RAs). (A) Short-acting GLP-1RAs
`suppress gastric emptying, which prolongs the presence of food in the stomach and upper small intestine; the reduced transpyloric flow causes delayed
`intestinal glucose absorption and diminished postprandial insulin secretion. Short-acting GLP-1RAs may also directly suppress glucagon secretion. (B)
`Longer-acting GLP-1RAs do not significantly affect gastric motility, because of tachyphylaxis. Instead, longer-acting GLP-1RAs exert more of their effect
`via the pancreas, increasing insulin secretion, and inhibiting glucagon secretion via paracrine release of somatostatin. By targeting the central nervous
`system, both shorter- (A) and longer (B) -acting GLP-1RAs increase satiety and also may induce nausea. Adapted from Meier [22]. Adapted by permission
`from Macmillan Publishers Ltd: Nature Reviews Endocrinology 2012;8(12):728–42, copyright 2012.
`
`the DURATION-6 and LEAD-6 studies, led to greater HbA1c
`reductions than both exenatide twice daily and once weekly
`(p ≤ 0.02) [66,67]. Liraglutide also led to greater reductions
`in HbA1c than lixisenatide (p < 0.01) in the phase II study
`(although it was not the primary endpoint and the study dura-
`tion was short [61]). Likewise, in the HARMONY 7 study [14],
`liraglutide led to greater reductions in HbA1c than albiglutide
`(however, the predefined non-inferiority criteria for albiglutide
`were not met). In the AWARD-6 study [13], the reduction in
`HbA1c did not differ between liraglutide and dulaglutide after
`26 weeks of treatment. In the T-emerge 2 study, taspoglutide at
`10 and 20 mg led to greater reductions in HbA1c than exenatide
`10 μg twice daily (p < 0.0001) [40].
`In addition, PPG and FPG were assessed in many of these
`trials. As expected, based on the delayed gastric emptying seen
`with the short-acting GLP-1RAs, exenatide twice daily and
`lixisenatide had greater effects on PPG than the longer-acting
`GLP-1RAs and this improvement was observed after the meal
`that followed the injection. For example, in the phase II study,
`lixisenatide administered before breakfast was associated with
`significantly greater reductions in maximum PPG excursion
`than liraglutide (−3.9 mmol/l vs −1.4 mmol/l, respectively;
`p < 0.0001), resulting in PPG of 7.3 and 10.1 mmol/l, respec-
`tively, 2 h after starting breakfast
`[61]. The differential
`effect on PPG is evident in the 24-h plasma glucose pro-
`files shown in Figure 3. These data are supported by results
`from another phase II study, which showed that lixisenatide
`
`had a significantly greater effect than liraglutide in reducing
`area under the PPG curve after a standardized solid breakfast.
`This difference was thought to be attributable to significant
`delays in gastric emptying with lixisenatide versus liraglutide,
`which reduced post-breakfast blood glucose exposure [68].
`Similarly, in a comparison of exenatide twice daily and exe-
`natide once weekly, the mean change from baseline in 2-h
`PPG was significantly greater with the twice-daily versus the
`once-weekly formulation (−6.9 mmol/l vs −5.3 mmol/l, respec-
`tively; p = 0.0124), and the delay in gastric emptying was more
`pronounced with exenatide twice daily than with exenatide
`once weekly [54]. Furthermore, in a study conducted in Asian
`participants by Ji et al. [56], exenatide twice daily produced
`significantly greater reductions in postprandial blood glucose
`than exenatide once weekly based on assessments 2 h after
`each of the morning and evening meals (p < 0.001); however,
`in a sub-analysis of participants in the T-emerge 2 study, the
`longer-acting GLP-1RA taspoglutide had similar effects to exe-
`natide twice daily on postprandial metabolism, although the
`mechanisms underlying this effect are not entirely clear [69].
`Generally, the longer-acting GLP-1RAs improve glucose
`control via a downward shift of the whole 24-h glucose curve,
`which explains the greater overall efficacy compared with the
`short-acting exenatide twice daily and lixisenatide once daily.
`While the short-acting GLP-1RAs typically have an advantage
`with respect to PPG, the situation is reversed with FPG. Here,
`the longer-acting GLP-1RAs resulted in greater improvements.
`
`320 Madsbad
`
`Volume 18 No. 4 April 2016
`
`
`
`MPI EXHIBIT 1111 PAGE 4
`
`

`

`DIABETES, OBESITY AND METABOLISM
`
`review article
`
`Safetyandtolerability
`ChangesinFPGandweight
`
`(53mmol/mol)and≤6.5%(48mmol/mol)]
`ParticipantsachievingHbA1ctargets[<7.0%
`
`ChangeinHbA1c
`
`HbA1c7.0–10.0%
`Treatedwithmetformin
`21–84yearsofage
`
`24weeks
`
`lixisenatideoncedaily
`Exenatidetwicedailyvs
`
`GetGoal-X[12]
`
`ChangesinFPG,SMPG,weight,BP,lipids,glucagonand
`
`Safetyandtolerability
`
`HOMA-𝛽
`
`(53mmol/mol)and≤6.5%(48mmol/mol)]
`
`ParticipantsreachingHbA1ctargets[<7.0%
`
`ChangeinHbA1c
`
`Exenatidepharmacokineticsandparacetamol
`ChangesinFPG,PPG,weight,BP,lipidsandglucagon
`
`Safetyandtolerability
`Patient-reportedoutcomes
`ChangesinFSG,weight,BPandlipids
`ParticipantsachievingHbA1ctarget(<7.0%)
`Safetyandtolerability
`
`ChangesinFSG,SMPG,weight,lipids,HOMA-𝛽and
`
`insulinsensitivity
`
`(42mmol/mol)]
`(53mmol/mol),≤6.5%(48mmol/mol)and≤6.0%
`
`ParticipantsachievingHbA1ctargets[≤7.0%
`
`Safetyandtolerability
`ChangesinFPG,weight,BPandlipids
`PatientsachievingFPGtarget(≤7.0mmol/l)
`
`(53mmol/mol)and≤6.5%(48mmol/mol)]
`ParticipantsachievingHbA1ctargets[<7.0%
`Safetyandtolerability
`
`absorption
`
`ChangeinHbA1c
`
`ChangeinHbA1c
`
`ChangeinHbA1c
`
`(42mmol/mol)]
`(53mmol/mol),≤6.5%(48mmol/mol)and≤6.0%
`
`ParticipantsachievingHbA1ctargets[≤7.0%
`Keysecondaryendpoints
`
`ChangeinHbA1c
`Primaryendpoint
`
`≥16yearsofage
`Inclusioncriteria
`
`30weeks*
`Duration
`
`Table2.DesignofpublishedphaseIII(andonekeyphaseII)randomizedhead-to-headstudiesofglucagon-likepeptide-1receptoragonistsintype2diabetes.
`
`BMI≤45kg/m2
`HbA1c7.0–11.0%
`Therapywithmetformin,SUorboth
`18–80yearsofage
`
`Stablebodyweight
`BMI≤45kg/m2
`HbA1c7.1–11.0%
`
`TreatedwithdietandexerciseandOADs
`≥18yearsofage
`
`metformin+pioglitazone)
`(metformin,SU,metformin+SU,or
`
`BMI21–35kg/m2
`HbA1c7.0–11.0%
`
`Treatedwith1–3OADs(metformin,SU,
`≥20yearsofage
`BMI25–45kg/m2
`FPG<15.5mmol/l
`HbA1c7.1–11.0%
`
`TZD)
`
`metformin,SU,TZDoracombination
`
`Treatedwithdietandexercise,orwith
`≥18yearsofage
`BMI25–45kg/m2
`FPG<16mmol/l
`HbA1c7.1–11.0%
`
`Therapywithdietandexercise,orwith1–2
`
`OADs(metformin,SUand/orTZD)
`
`26weeks*
`
`liraglutideoncedaily
`Exenatidetwicedailyvs
`
`LEAD-6[66]
`
`26weeks
`
`Exenatideonceweeklyvs
`
`liraglutideoncedaily
`
`DURATION-6[67]
`
`26weeks
`
`Exenatideonceweeklyvs
`
`exenatidetwicedaily
`
`Jietal.[56]
`
`24weeks
`
`Exenatideonceweeklyvs
`
`exenatidetwicedaily
`
`DURATION-5[55]
`
`exenatidetwicedailyvs
`Exenatideonceweeklyvs
`Treatmentarms
`
`DURATION-1[54]
`Studyname
`
`Volume 18 No. 4 April 2016
`
`doi:10.1111/dom.12596 321
`
`
`
`MPI EXHIBIT 1111 PAGE 5
`
`

`

`review article
`
`DIABETES, OBESITY AND METABOLISM
`
`†PhaseIIstudy(n=148).
`*Alsoincludedanextensionphaseupto52weeks.
`modelassessmentof𝛽-cellfunction;OAD,oralantidiabeticdrug;PPG,postprandialglucose;SMPG,self-measuredplasmaglucose;SU,sulphonylurea;T2D,type2diabetes;TZD,thiazolidinedione.
`BMI,bodymassindex;BP,bloodpressure;FPG,fastingplasmaglucose;FSG,fastingserumglucose;GLP-1RA,glucagon-likepeptide-1receptoragonist;HbA1c,glycatedhaemoglobinHOMA-𝛽,homeostasis
`Allstudieswereopen-label.
`
`Safetyandtolerability
`MeanHbA1c
`24-hplasmaglucoseprofile
`
`Changesinpremealseruminsulin,serumC-peptide
`
`andplasmaglucagonlevels
`
`ChangesinmaximumPPGexcursionaftera
`
`standardizedbreakfasttestmeal
`
`Safetyandtolerability
`
`52weeks)
`pro-insulin/insulinratioandHOMA-𝛽(assessedover
`
`Changesinfastingpro-insulin,fasting
`ChangesinFPGandweight(assessedover52weeks)
`
`Safetyandtolerability
`ChangeinFSG,SMPG,weight,BMIandHOMA-𝛽
`
`(53mmol/mol)and≤6.5%(48mmol/mol)]
`ParticipantsachievingHbA1ctargets[<7.0%
`
`ChangeinPPGexposure
`
`ChangeinHbA1c
`
`ChangeinHbA1c
`
`Safetyandtolerability
`ChangesinFPGandweight
`
`(53mmol/mol)and<6.5%(48mmol/mol)]
`ParticipantsachievingHbA1ctargets[<7.0%
`
`Keysecondaryendpoints
`
`ChangeinHbA1c
`
`Primaryendpoint
`
`HbA1c6.5–9.0%
`Treatedwithmetformin
`37–74yearsofage
`
`Stablebodyweight
`BMI25–45kg/m2
`HbA1c7.0–10.0%
`Treatedwithmetforminand/orTZD
`18–75yearsofage
`
`28days
`
`Lixisenatideoncedailyvs
`
`liraglutideoncedaily
`
`Kapitzaetal.†[61]
`
`24weeks*
`
`taspoglutideonceweekly
`
`Exenatidetwicedailyvs
`
`T-emerge2[40]
`
`HbA1c7.0–10.0%
`Treatedwithmetformin
`≥18yearsofage
`
`26weeks
`
`Dulaglutideonceweeklyvs
`
`liraglutideoncedaily
`
`AWARD-6[13]
`
`BMI20–45kg/m2
`HbA1c7.0–10.0%
`
`Treatedwithmetformin,SU,TZDora
`≥18yearsofage
`
`combination
`
`32weeks
`
`Albiglutideonceweeklyvs
`
`liraglutideoncedaily
`
`HARMONY7[14]
`
`Inclusioncriteria
`
`Duration
`
`Treatmentarms
`
`Studyname
`
`Table2.Continued
`
`322 Madsbad
`
`Volume 18 No. 4 April 2016
`
`
`
`MPI EXHIBIT 1111 PAGE 6
`
`

`

`DIABETES, OBESITY AND METABOLISM
`
`review article
`
`31(4)
`N/A
`
`—
`
`—
`
`—
`
`—
`
`100
`
`70/30
`60(9)
`
`31(4)
`N/A
`
`—
`
`—
`
`—
`
`1
`
`99
`
`64/36
`61(8)
`
`33(5)
`21
`
`N/A
`15
`N/A
`N/A
`85
`
`52/48
`56(10)
`
`34(5)
`18
`
`N/A
`15
`N/A
`N/A
`85
`
`34(5)
`20
`
`N/A
`15
`N/A
`N/A
`85
`
`58/42
`56(10)
`
`49/51
`55(10)
`
`(n=71)
`oncedaily
`0.6–1.8mg
`Liraglutide
`
`(n=77)
`oncedaily
`10–20𝛍g
`Lixisenatide
`
`(n=392)
`onceweekly
`20mg
`Taspoglutide
`
`(n=384)
`onceweekly
`10mg
`Taspoglutide
`
`(n=373)
`twicedaily
`10𝛍g
`Exenatide
`
`Kapitzaetal.(28days)§[61]
`
`T-emerge2(24weeks)[40]
`
`34(5)
`24
`
`8
`
`1
`
`—
`
`5
`
`86
`
`50/50
`57(10)
`
`(n=300)
`oncedaily
`1.8mg
`Liraglutide
`
`80(8)
`132(16)
`9(6)
`9.8(2.5)
`66(11)
`8.2(1.0)
`93(20)
`33(6)
`14
`
`—
`
`1
`
`<1
`
`6
`
`93
`
`79(9)
`134(17)
`8(6)
`9.5(2.4)
`65(11)
`8.1(1.0)
`93(20)
`33(6)
`11
`
`—
`
`5
`
`2
`
`2
`
`91
`
`80(9)
`134(14)
`9(6)
`9.8(2.6)
`68(11)
`8.4(1.0)
`91(19)
`32(5)
`22
`
`4
`
`1
`
`13
`
`1
`
`82
`
`79(9)
`132(14)
`8(6)
`9.6(2.5)
`69(11)
`8.5(1.0)
`91(20)
`32(6)
`21
`
`<1
`
`4
`
`12
`
`1
`
`83
`
`—
`
`—
`
`—
`
`100
`
`—
`
`—
`
`49/51
`56(10)
`
`55/45
`57(11)
`
`54/46
`57(10)
`
`55/45
`57(9)
`
`54/46
`56(10)
`
`(n=233)
`oncedaily
`Liraglutide1.8mg
`
`(n=231)
`twicedaily
`Exenatide10𝛍g
`
`(n=450)
`oncedaily
`1.8mg
`Liraglutide
`
`(n=461)
`onceweekly
`2mg
`Exenatide
`
`(n=338)
`twicedaily
`Exenatide10𝛍g
`
`LEAD-6(26weeks)[66]
`
`DURATION-6(26weeks)[67]
`
`7
`
`34(5)
`25
`
`<1
`
`<1
`
`7
`
`86
`
`46/54
`57(9)
`
`(n=299)
`onceweekly
`1.5mg
`Dulaglutide
`
`AWARD-6(26weeks)[13]
`80(1)‡
`79(1)‡
`132(1)‡
`131(1)‡
`9(6)
`8(5)
`9.4(2.7)
`9.1(2.4)
`72(11)
`72(11)
`8.7(1.0)
`8.7(1.0)
`70(12)
`70(12)
`27(3)
`26(4)
`
`—
`
`—
`
`—
`
`100
`
`—
`
`—
`
`—
`
`33
`
`7
`
`4
`
`55
`
`—
`
`29
`
`5
`
`4
`
`63
`
`—
`
`14
`
`1
`
`13
`73
`
`—
`
`11
`
`6
`
`0
`
`83
`
`54/46
`55(11)
`
`(n=340)
`onceweekly
`Exenatide2mg
`
`55/45
`55(10)
`
`60/40
`56(11)
`
`(n=123)
`twicedaily
`10𝛍g
`Exenatide
`
`(n=129)
`onceweekly
`2mg
`Exenatide
`
`51/49
`55(10)
`
`55/45
`55(10)
`
`(n=147)
`twicedaily
`10𝛍g
`Exenatide
`
`(n=148)
`onceweekly
`2mg
`Exenatide
`
`Jietal.(26weeks)[56]
`
`DURATION-5(24weeks)[55]
`
`DURATION-1(30weeks)[54]
`
`33(6)
`N/A
`
`N/A
`N/A
`N/A
`N/A
`N/A
`
`33(6)
`N/A
`
`N/A
`N/A
`N/A
`N/A
`N/A
`
`53/47
`56(10)
`
`47/53
`55(10)
`
`(n=408)
`oncedaily
`1.8mg
`Liraglutide
`
`(n=404)
`onceweekly
`50mg
`Albiglutide
`
`HARMONY7(32weeks)[14]
`78(1)‡
`130(1)‡
`7(5)
`9.6(2.6)
`69(12)
`8.5(1.1)
`97(21)
`34(6)
`29
`
`77(1)‡
`128(1)‡
`7(5)
`9.3(2.6)
`68(13)
`8.4(1.2)
`94(19)
`33(5)
`33
`
`34(6)
`N/A
`
`—
`
`3
`
`1
`
`4
`
`93
`
`48/53
`57(9)
`
`34(7)
`N/A
`
`BMI,kg/m2
`
`Hispanic/LatinAmerican,%
`
`—
`
`3
`
`1
`
`3
`
`92
`
`59/41
`58(11)
`
`Ethnicorigin
`Multiple†
`Other*
`Asian/PacificIslander
`Black/AfricanAmerican
`White
`Race,%
`Male/female,%
`Age,years
`
`(n=318)
`oncedaily
`Lixisenatide20𝛍g
`
`(n=316)
`twicedaily
`Exenatide10𝛍g
`
`GetGoal-X(24weeks)[12]
`80(1)‡
`130(1)‡
`6(5)
`9.2(2.3)
`67(11)
`8.3(1.0)
`102(21)
`35(5)
`14
`
`78(1)‡
`128(1)‡
`7(6)
`9.6(2.4)
`67(11)
`8.3(1.0)
`102(19)
`35(5)
`11
`
`DBP,mmHg
`SBP,mmHg
`Durationofdiabetes,years
`FSG/FPG,mmol/l
`HbA1c,mmol/mol
`HbA1c,%
`Weight,kg
`BMI,kg/m2
`
`HispanicorLatinAmerican,%
`
`Ethnicorigin
`Multiple†
`Other*
`Asian/PacificIslander
`Black/AfricanAmerican
`White
`Race,%
`Male/female,%
`Age,years
`
`Table3.BaselinecharacteristicsofparticipantpopulationsinpublishedphaseIII(andonephaseII)randomizedhead-to-headstudiesofglucagon-likepeptide-1receptoragonistsintype2diabetes.
`
`Volume 18 No. 4 April 2016
`
`doi:10.1111/dom.12596 323
`
`
`
`MPI EXHIBIT 1111 PAGE 7
`
`

`

`DIABETES, OBESITY AND METABOLISM
`
`For example, in DURATION-1, changes in FPG were sig-
`nificantly greater after 30 weeks with exenatide once weekly
`than with exenatide twice daily (−2.3 mmol/l vs −1.4 mmol/l,
`respectively; p < 0.0001) [54]. Similarly,
`in DURATION-5,
`the mean change in FPG at 24 weeks was significantly greater
`with exenatide once weekly than with exenatide twice daily
`(−1.9 mmol/l vs −0.7 mmol/l, respectively; p = 0.0008) [55].
`The longer-acting GLP-1RA taspoglutide was also associated
`with a significantly greater reduction in FPG than short-acting
`exenatide twice daily at 24 weeks [40]. Accordingly, in the
`phase II study, changes in FPG were greater with liraglutide
`than with lixisenatide (−1.3 mmol/l vs −0.3 mmol/l, respec-
`tively; p < 0.0001) [61]. Likewise, the long-acting liraglutide
`once daily demonstrated greater improvements than exenatide
`twice daily (−1.6 mmol/l vs −0.6 mmol/l, respectively) [66].
`Comparisons of once-daily liraglutide with once-weekly
`formulations produced a mixed pattern: liraglutide demon-
`strated superiority to exenatide once weekly (−2.1 mmol/l vs
`−1.8 mmol/l, respectively; p = 0.02) [67] and albiglutide once
`weekly (−1.7 mmol/l vs −1.2 mmol/l, respectively; p = 0.0048)
`[14] in lowering fasting serum glucose and FPG, respectively,
`but no significant difference compared with dulaglutide once
`weekly (1.90 mmol/l vs 1.93 mmol/l, respectively) [13].
`
`Effects on Weight
`As a class, the GLP-1RAs have all been shown to have a
`weight-reduction effect (Figure 4), and this effect is significantly
`greater than is typically seen with most other therapeutic classes
`[70]. Indeed, a systematic review of clinical trials involving exe-
`natide twice daily, liraglutide and exenatide once weekly in par-
`ticipants with a body mass index (BMI) ≥25 kg/m2 (with or
`without T2D) found a greater reduction in weight with these
`compounds versus non-GLP-1RA-treated control compounds
`(weighted mean difference: −2.9 kg) [70].
`The weight benefit varies among GLP-1RAs and studies. For
`example, Ji et al. [56] found that exenatide twice daily was asso-
`ciated with a significantly greater reduction in weight than
`exenatide once weekly (p < 0.001) but, in DURATION-1 and
`DURATION-5 [54,55], weight loss was not significantly differ-
`ent between the two exenatide formulations. In the T-emerge 2
`study, exenatide twice daily showed a greater (non-significant)
`reduction in weight than taspoglutide 10 mg once weekly but
`showed no difference in weight loss compared with taspoglu-
`tide 20 mg once weekly [40]. In the LEAD-6 study, exenatide
`twice daily and liraglutide treatment led to similar levels of
`weight loss (3.2 and 2.9 kg, respectively; p = 0.2235) [66]. Exe-
`natide twice daily was associated with greater (non-significant)
`weight loss than lixisenatide in the GetGoal-X study [12], and
`liraglutide treatment led to greater weight loss than lixisen-
`atide in the study by Kapitza et al. [61] (p < 0.01; Figure 3).
`Other head-to-head trials revealed significantly greater reduc-
`tions in weight with liraglutide than the once-weekly treatments
`exenatide (p = 0.0005), albiglutide (p < 0.0001) and dulaglutide
`(p = 0.011; Figure 4) [13,14,67]. Between-treatment differences
`were 0.9, 1.6 and 0.7 kg, respectively [13,14,67].
`In the Ji et al. [56] and DURATION-6 studies [67], the
`greatest weight loss was observed in participants treated with
`
`§PhaseIIstudy.
`‡Standarderrorused.
`†Asdefinedbystudy.
`*Othernon-white.
`Alldataareroundedtothenearestinteger,exceptforHbA1candFPG/FSG,whicharepresentedtothenearestdecimalpoint.Dataaremean(standarddeviation)unlessstatedotherwise.
`available;SBP,systolicbloodpressure;T2D,type2diabetes.
`BMI,bodymassindex;DBP,diastolicbloodpressure;FPG,fastingplasmaglucose;FSG,fastingserumglucose;GLP-1RA,glucagon-likepeptide-1receptoragonist;HbA1c,glycatedhaemoglobin;N/A,not
`
`review article
`
`N/A
`N/A
`7(N/A)
`N/A
`55(7)
`7.2(0.6)
`91(15)
`(n=77)
`oncedaily
`10–20𝛍g
`Lixisenatide
`Kapitzaetal.(28days)§[61]
`
`N/A
`N/A
`7(N/A)
`N/A
`57(9)
`7.4(0.8)
`93(17)
`(n=71)
`oncedaily
`0.6–1.8mg
`Liraglutide
`
`80(0)‡
`132(1)‡
`7(6)
`9.8(2.4)
`65(10)
`8.1(0.9)
`93(19)
`(n=392)
`onceweekly
`20mg
`Taspoglutide
`
`N/A
`80(0)‡
`N/A
`132(1)‡
`8(6)
`6(5)
`N/A
`9.9(2.6)
`66(10)
`65(10)
`8.2(0.9)
`8.1(0.9)
`92(21)
`96(20)
`(n=404)
`(n=384)
`onceweekly
`onceweekly
`50mg
`10mg
`Albiglutide
`Taspoglutide
`HARMONY7(32weeks)[14]AWARD-6(26weeks)[13]T-emerge2(24weeks)[40]
`
`N/A
`N/A
`8(6)
`N/A
`66(9)
`8.2(0.8)
`93(22)
`(n=408)
`oncedaily
`1.8mg
`Liraglutide
`
`N/A
`N/A
`7(6)
`9.7(2.0)
`64(9)
`8.0(0.8)
`94(20)
`(n=318)
`oncedaily
`Lixisenatide20𝛍g
`
`9.7(2.3)
`64(9)
`8.0(0.8)
`96(23)
`(n=316)
`twicedaily
`Exenatide10𝛍g
`
`N/A
`DBP,mmHg
`N/A
`SBP,mmHg
`Durationofdiabetes,years7(5)
`FPG/FSG,mmol/l
`HbA1c,mmol/mol
`HbA1c,%
`Weight,kg
`
`GetGoal-X(24weeks)[12]
`
`Table3.Continued.
`
`79(0)‡
`131(1)‡
`7(5)
`9.9(2.7)
`65(10)
`8.1(0.9)
`95(19)
`(n=373)
`twicedaily
`10𝛍g
`Exenatide
`
`79(9)
`131(15)
`7(5)
`9.2(2.3)
`65(9)
`8.1(0.8)
`94(19)
`(n=300)
`oncedaily
`1.8mg
`Liraglutide
`
`80(10)
`132(15)
`7(5)
`9.3(2.2)
`65(9)
`8.1(0.8)
`94(18)
`(n=299)
`onceweekly
`1.5mg
`Dulaglutide
`
`324 Madsbad
`
`Volume 18 No. 4 April 2016
`
`
`
`MPI EXHIBIT 1111 PAGE 8
`
`

`

`DIABETES, OBESITY AND METABOLISM
`
`review article
`
`Figure 2. Reductions in glycated haemoglobin (HbA1c) in published phase III (and one phase II) randomized head-to-head studies of glucagon-like
`peptide-1 receptor agonists in type 2 diabetes. *Non-inferiority criteria met. †Non-inferiority criteria not met. ‡Phase II study. §A 1% change in HbA1c
`corresponds to a 10.93 mmol/mol change in The International Federation of Clinical Chemistry units.
`
`Figure 3. Mean 24-h postprandial plasma glucose profiles at baseline and day 28. Data are mean ± standard error of the mean. Adapted from Kapitza
`et al. [61].
`
`exenatide (once weekly or twice daily) or liraglutide with the
`highest baseline BMI. Furthermore, in a retrospective analysis
`of seven phase III trials from the liraglutide diabetes devel-
`opment programme, a slightly greater weight reduction was
`observed in participants treated with liraglutide, with a longer
`duration of GI AEs [71].
`The exact mechanism by which GLP-1 exerts its anorectic
`effects is a matter of contention, but both peripheral and brain
`GLP-1 receptors seem to be involved [72,73]. It is also unclear
`whether the reduced weight loss with the large molecules,
`albig