DIABETICMedicine
`
`Research: Treatment
`
`DOI: 10.1111/j.1464-5491.2012.03745.x
`
`Monotherapy with the once-weekly GLP-1 analogue
`dulaglutide for 12 weeks in patients with Type 2
`diabetes: dose-dependent effects on glycaemic
`control in a randomized, double-blind, placebo-controlled
`study
`
`G. Grunberger1, A. Chang2, G. Garcia Soria3, F. T. Botros4, R. Bsharat4 and Z. Milicevic5
`
`1Grunberger Diabetes Institute, Bloomfield Hills, MI, 2John Muir Physician Network Clinical Research Center, Concord, CA, USA, 3IMIC Research, Mexico City,
`Mexico, 4Lilly Diabetes, Eli Lilly and Company, Indianapolis, IN, USA and 5Lilly Diabetes, Eli Lilly and Company, Vienna, Austria
`
`Accepted 26 June 2012
`
`Abstract
`
`Aims Evaluate dose-dependent effects of once-weekly dulaglutide, a glucagon-like peptide-1 analogue, on glycaemic control
`in patients with Type 2 diabetes treated with lifestyle measures with or without previous metformin.
`
`Methods This 12-week, double-blind, placebo-controlled, dose–response trial randomized 167 patients who were anti-
`hyperglycaemic medication-naı¨ve or had discontinued metformin monotherapy [mean baseline HbA1c 59  8 to
`61  8 mmol ⁄ mol (7.6  0.7 to 7.8  0.8%)] to once-weekly injections of placebo or dulaglutide (0.1, 0.5, 1.0 or 1.5 mg).
`Results A significant dose-dependent reduction in HbA1c (least squares mean  se) was observed across doses (P < 0.001).
`HbA1c reductions in the 0.5, 1.0 and 1.5 mg dulaglutide groups were greater than in the placebo group [)10  1, )11  1
`and )11  1 vs. 0  1 mmol ⁄ mol ()0.9  0.1, )1.0  0.1 and )1.0  0.1 vs. 0.0  0.1%), respectively, all P < 0.001].
`Dose-dependent reductions in fasting plasma glucose were also observed [least squares mean difference (95% CI) ranging
`from )0.43 ()1.06 to 0.19) mmol ⁄ l for dulaglutide 0.1 mg to )1.87 ()2.56 to )1.19) mmol ⁄ l for dulaglutide 1.5 mg,
`P < 0.001]. Dose-dependent weight loss was demonstrated across doses (P = 0.009), but none of the groups were different
`from placebo. The most common adverse events were nausea and diarrhoea.
`
`Conclusions The observed dulaglutide dose-dependent reduction in HbA1c and its acceptable safety profile support further
`clinical development for treatment of Type 2 diabetes.
`
`Diabet. Med. 29, 1260–1267 (2012)
`
`Introduction
`
`Glucagon-like peptide-1 (GLP-1) is an incretin hormone that
`causes increases in glucose-dependent insulin secretion, inhi-
`bition of glucagon secretion, slowing of gastric emptying, and
`increased satiety [1]. Several GLP-1 analogues have been
`developed or are in development for treatment of Type 2
`diabetes [2–7]. Dulaglutide (Dula; LY2189265; Eli Lilly and
`Company,
`Indianapolis,
`IN, USA), a long-acting GLP-1
`analogue, consists of two GLP-1 peptides covalently linked
`by a small peptide to a human IgG4-Fc heavy chain (Fig. 1).
`The GLP-1 moieties contain amino acid substitutions that
`
`Correspondence to: Zvonko Milicevic. E-mail: milicevic_zvonko@lilly.com
`
`protect from inactivation by dipeptidyl peptidase-4 (DPP-4),
`while the linker peptide maintains the potency of the GLP-1
`peptide. The IgG4-Fc is modified by substituting several
`amino acids
`to reduce interaction with high-affinity Fc
`receptors, cytotoxicity and immunogenicity [8]. The large
`molecule size is expected to limit its renal clearance. The
`resulting half-life is approximately 4 days and time to peak
`concentration is 12–72 h [9].
`in fasting plasma glucose,
`Dose-dependent
`reductions
`postprandial glucose and HbA1c were previously reported in
`patients with Type 2 diabetes (n = 43) receiving once-weekly
`dulaglutide (doses ranging from 0.05 to 8 mg) for 5 weeks [9].
`The objective of this Phase 2 study was to assess the dose–
`response relationship with respect to HbA1c across a narrower
`range of doses and a longer 12-week treatment period.
`
`1260
`
`ª 2012 Eli Lilly and Company.
`Diabetic Medicine ª 2012 Diabetes UK
`
`MPI EXHIBIT 1109 PAGE 1
`
`

`

`Research article
`
`DIABETICMedicine
`
`NH2
`
`NH2
`
`GLP-1 peptide
`
`Linker peptide
`
`IgG4-Fc domain
`
`HOOC
`
`COOH
`
`FIGURE 1 Dulaglutide; structure of the molecule.
`
`Research design and methods
`
`Study design
`
`This 12-week, double-blind, placebo-controlled, dose–response
`study assessed the safety and efficacy of dulaglutide in patients
`with Type 2 diabetes (n = 167). The study was conducted
`between November 2008 and January 2010 in 44 sites in seven
`countries. Eligible patients were anti-hyperglycaemic medica-
`tion-naı¨ve or on metformin monotherapy. Inclusion criteria
`were: age ‡ 18 and £ 75 years; BMI ‡ 23 to £ 40 kg ⁄ m2 for
`patients native to and residents of South and ⁄ or East Asia; ‡ 25
`to £ 40 kg ⁄ m2 for all other patients; stable weight for 3 months
`before screening; and HbA1c ‡ 53 to £ 80 mmol ⁄ mol (‡ 7.0 to
`£ 9.5%) for anti-hyperglycaemic medication-naı¨ve patients and
`> 48 to £ 75 mmol ⁄ mol
`(> 6.5 to £ 9.0%)
`[> 42 to
`£ 69 mmol ⁄ mol
`(> 6.0 to £ 8.5%) prior
`to a protocol
`amendment] for patients who were taking metformin. Exclu-
`sion criteria included treatment with any oral anti-diabetes
`drug other than metformin within 3 months or other GLP-1
`analogue within 6 months prior to screening, prior use of
`insulin for long-term glycaemic control, serious cardiovascular
`condition, liver disease, history of pancreatitis or serum creat-
`inine ‡ 1.5 mg ⁄ dl (men) or ‡ 1.4 mg ⁄ dl (women).
`Study periods included: 2-week screening, 4- to 8-week lead-
`in (8-week washout after discontinuing metformin was
`required prior to obtaining the qualifying HbA1c); 12-week
`treatment period; and 4-week safety follow-up. After lead-in,
`an HbA1c value ‡ 48 to £ 80 mmol ⁄ mol (‡ 6.5 to £ 9.5%)
`[‡ 53 to £ 80 mmol ⁄ mol (‡ 7.0 to £ 9.5%) prior to protocol
`amendment] was required for randomization. Patients were
`randomized (block sizes of 5) to one of five treatment arms:
`placebo, 0.1 mg, 0.5 mg, 1.0 mg or 1.5 mg dulaglutide
`(Dula 0.1, Dula 0.5, Dula 1.0 and Dula 1.5) in a 1:1:1:1:1
`ratio via an interactive voice-response system. In the original
`design, patients were randomized to placebo, 0.1 mg, 0.5 mg,
`1.0 mg or 3.0 mg dulaglutide. Based on recommendations
`from the data monitoring committee of another dulaglutide
`
`study, the Dula 3.0 arm was discontinued in May 2009 and the
`protocol was amended to replace the Dula 3.0 arm with the
`Dula 1.5 arm. A total of 17 patients had been randomized
`prior to protocol amendment;
`the three patients on the
`Dula 3.0 dose were discontinued and the other 14 patients
`continued on randomized treatment.
`Patients were stratified for randomization by country, BMI
`and pre-study therapy (metformin use or not). Study drug was
`administered once weekly by subcutaneous injection; as this
`was a placebo-controlled study, the use of additional oral anti-
`diabetes drugs was permitted only when needed for rescue
`therapy (according to pre-specified criteria). If rescued, patients
`continued to administer the study drug until the last on-treat-
`ment visit. Other GLP-1 agonists and DPP-4 inhibitors were
`not allowed at any time.
`A common protocol was approved at each site by an insti-
`tutional review board and was performed in accordance with
`the principles of the Declaration of Helsinki. Prior to partici-
`pation, all patients provided written informed consent.
`
`Study endpoints
`
`The primary efficacy measure was change from baseline in
`HbA1c at 12 weeks. Additional measures included changes in
`fasting plasma glucose (central laboratory), 7-point self-moni-
`tored plasma glucose, b-cell function and insulin sensitivity
`using the homeostasis model assessment 2 (HOMA2-%B and
`HOMA2-%S, respectively), body weight and proportion of
`patients achieving HbA1c < 53 or £ 48 mmol ⁄ mol (< 7 or
`£ 6.5%). Safety assessments included cardiovascular (pulse
`rate, blood pressure, electrocardiogram) and laboratory
`parameters, reported adverse events and anti-dulaglutide anti-
`bodies. Electrocardiograms were recorded in triplicate and
`tracings were over-read by a cardiologist at a centralized ven-
`dor (Biomedical Systems Corporation, Maryland Heights, MO,
`USA); this report was used for analysis. Referencing the
`American Diabetes Association definition, hypoglycaemia was
`defined as plasma glucose £ 3.9 mmol ⁄ l (£ 70 mg ⁄ dl) and ⁄ or
`symptoms and ⁄ or signs attributable to hypoglycaemia. Severe
`hypoglycaemia was defined as an episode requiring the assis-
`tance of another person to actively administer therapy [10].
`Patients with at least one pancreatic enzyme measurement ‡ 3
`times the upper limit of normal underwent a standardized
`diagnostic examination.
`Plasma analytes and HbA1c were quantified by Quintiles
`Laboratories
`(Smyrna, GA, USA). Electrochemilumines-
`cence immunosorbent assay was used for detection of
`anti-dulaglutide antibodies (Millipore, Billerica, MA, USA); po-
`sitive samples were titrated for titres. Fasting plasma glucose and
`insulin concentrations were used for HOMA2 calculations [11].
`
`Statistical analysis
`
`The target sample size of 36 patients per group was calculated
`to provide 90% power for detecting a linear dose–response,
`
`ª 2012 Eli Lilly and Company.
`Diabetic Medicine ª 2012 Diabetes UK
`
`1261
`
`
`
`e
`
`MPI EXHIBIT 1109 PAGE 2
`
`

`

`DIABETICMedicine
`
`Dulaglutide monotherapy and glycaemic control • G. Grunberger et al.
`
`excluding the placebo group, with a 0.60 slope in change from
`baseline HbA1c for each 1-mg change in dose. Assumptions
`included residual standard deviation (sd) of 1.2%, 0.61 mg sd
`of the doses, 2-sided 0.05 significance level and 20% dropout
`rate. With this sample size, a 0.9% difference in change from
`baseline in HbA1c could be detected between any dulaglutide
`group and the placebo group with 80% power.
`The primary and secondary analyses were performed on the
`intention-to-treat population (n = 167), defined as
`all
`randomized patients who received study therapy, including
`patients from the discontinued Dula 3.0 mg arm (n = 3).
`Changes from baseline were reported as least-squares mean and
`standard error (least-squares mean  se), although summary
`statistics were not provided for the discontinued Dula 3.0 arm
`because of the small number.
`A mixed-effects model for repeated measures (MMRM) was
`used for analyses of continuous variables. To evaluate the dose–
`response relationship on the change in HbA1c at 12 weeks, the
`model included: country, dose, pre-study therapy (metformin
`yes ⁄ no), visit and dose-by-visit interaction as the fixed effects;
`baseline BMI and ⁄ or baseline HbA1c as a covariate; and patient
`as a random effect. If baseline BMI and baseline HbA1c were
`significantly correlated at the 0.10 alpha level, the model
`included the one that had a higher correlation with the change
`in HbA1c at 12 weeks. Orthogonal contrasts considering the
`unequal spacing between doses were used to examine the linear
`and log linear dose–response without placebo at 12 weeks. The
`contrast with the smaller se, representing the better fit, was
`reported. Dunnett’s test was used to control the type I error
`when comparing placebo to the individual doses.
`
`A Cochran–Armitage trend test was used to assess categori-
`cal data, and a one-way ANOVA on the ranks with treatment
`as a fixed effect was conducted for laboratory data. All
`statistical analyses were performed using the SAS SystemÒ
`version 8.2 or higher (SAS Institute, Cary, NC, USA).
`
`Results
`
`Patients
`
`In total, 460 patients were screened; most frequent reasons for
`screen failure were not fulfilling inclusion ⁄ exclusion criteria
`(n = 244), patient decision (n = 36) and physician decision
`(n = 12). The three patients randomized to Dula 3.0 were
`discontinued at 1, 64 and 72 days post-randomization; 164
`patients were randomized to the other five treatment arms and
`153 completed the 12-week treatment (Fig. 2). Twelve patients
`discontinued before the last safety follow-up visit (Table 1 and
`Fig. 2). Three patients received rescue therapy (two in the
`placebo group and one in the Dula 1.0 group). Patient
`characteristics at entry were well balanced with no significant
`differences between groups (Table 1).
`
`Primary endpoint
`At randomization, baseline HbA1c (mean  sd) was com-
`parable among groups [60  9, 60  8, 59  8, 61  8
`and 60  7 mmol ⁄ mol
`(7.7  0.8, 7.6  0.7, 7.6  0.7,
`7.8  0.8 and 7.6  0.7%)] for placebo, Dula 0.1, Dula 0.5,
`Dula 1.0 and Dula 1.5, respectively (Fig. 3a). Dose-dependent
`
`293 patients failed screening or discontinued
`prior to randomization
`
`460 patients entered trial
`
`167 patients enrolled/randomized
`
`Dulaglutide 3.0 mg
`3 patients
`
`Dulaglutide 0.1 mg
`35 patients
`
`Dulaglutide 0.5 mg
`34 patients
`
`Dulaglutide 1.0 mg
`34 patients
`
`Dulaglutide 1.5 mg
`29 patients
`
`Placebo
`32 patients
`
`3 patients
`discontinued
`
`Dose discontinued
`from study
`
`Sponsor decision (3)
`
`2 patients
`discontinued
`treatment period
`
`Protocol violation (2)
`
`3 patients
`discontinued
`treatment period
`
`0 patients
`discontinued
`treatment period
`
`4 patients
`discontinued
`treatment period
`
`2 patients
`discontinued
`treatment period
`
`Adverse event (1)
`Protocol violation (1)
`Lost to follow-up (1)
`
`Adverse event (2)
`Protocol violation (1)
`Patient decision (1)
`
`Adverse event (1)
`Patient decision (1)
`
`0 patients completed
`
`33 patients completed
`treatment
`
`31 patients completed
`treatment
`
`34 patients completed
`treatment
`
`25 patients completed
`treatment
`
`30 patients completed
`treatment
`
`153 patients completed treatment period
`
`152 patients completed safety period
`
`1 patient in the placebo group discontinued
`(lost to follow-up) after treatment but before
`completing the safety period
`
`FIGURE 2 Patient disposition from entry to completing safety period throughout the study.
`
`1262
`
`ª 2012 Eli Lilly and Company.
`Diabetic Medicine ª 2012 Diabetes UK
`
`
`
`e
`
`MPI EXHIBIT 1109 PAGE 3
`
`

`

`Research article
`
`DIABETICMedicine
`
`Table 1 Patient characteristics at entry (screening) for intent-to-treat population and reasons for discontinuation
`
`Intent-to-treat
`population
`(mean  sd)
`
`Age (years)
`Sex
`Female, n (%)
`Male, n (%)
`Race, Caucasian
`⁄ Asian ⁄ Black
`or African
`American ⁄ Others (%)
`Body weight (kg)
`BMI (kg ⁄ m2)
`HbA1c (mmol ⁄ mol)
`HbA1c (%)
`Duration of
`diabetes (years)
`History of
`metformin, n (%)
`No
`Yes
`Systolic blood
`pressure (mmHg)
`Diastolic blood
`pressure (mmHg)
`Reason for
`discontinuation, n (%)
`Adverse events
`Lost to follow-up
`Protocol violation
`Subject decision
`Completers, n (%)
`
`Placebo
`(n = 32)
`
`Dula 0.1 
`(n = 35)
`
`Dula 0.5 
`(n = 34)
`
`Dula 1.0 
`(n = 34)
`
`Dula 1.5 
`(n = 29)
`
`Total
`(n = 164)
`
`P-value*
`
`55.0  9.3
`
`56.3  9.2
`
`56.9  9.1
`
`57.2  8.8
`
`57.5  7.9
`
`56.6  8.8
`
`14 (43.8)
`18 (56.3)
`78 ⁄ 16 ⁄ 3 ⁄ 3
`
`24 (68.6)
`11 (31.4)
`83 ⁄ 11 ⁄ 3 ⁄ 3
`
`18 (52.9)
`16 (47.1)
`82 ⁄ 15 ⁄ 3 ⁄ 0
`
`18 (52.9)
`16 (47.1)
`77 ⁄ 15 ⁄ 0 ⁄ 9
`
`16 (55.2)
`13 (44.8)
`83 ⁄ 14 ⁄ 3 ⁄ 0
`
`90 (54.9)
`74 (45.1)
`81 ⁄ 14 ⁄ 2 ⁄ 3
`
`90.9  18.9
`32.1  5.2
`57  7
`7.4  0.6
`3.9  4.7
`
`87.1  17.3
`32.9  4.8
`54  6
`7.1  0.6
`3.9  3.2
`
`90.2  21.3
`32.3  5.4
`55  7
`7.2  0.6
`3.7  3.8
`
`86.9  17.0
`32.2  4.5
`56  7
`7.3  0.7
`3.3  2.5
`
`85.8  18.6
`31.0  4.3
`56  5
`7.3  0.4
`4.6  4.1
`
`88.2  18.6
`32.1  4.8
`56  6
`7.2  0.6
`3.9  3.7
`
`6 (18.8)
`26 (81.3)
`128.5  12.3
`
`7 (20.0)
`28 (80.0)
`130.7  15.9
`
`6 (17.6)
`28 (82.4)
`129.6  16.1
`
`8 (23.5)
`26 (76.5)
`125.6  15.1
`
`4 (13.8)
`25 (86.2)
`127.3  14.4
`
`31 (18.9)
`133 (81.1)
`128.4  14.8
`
`77.9  10.5
`
`77.1  9.9
`
`75.7  8.9
`
`77.3  9.2
`
`76.8  9.2
`
`77.0  9.5
`
`1 (3.1)
`1 (3.1)
`0 (0.0)
`1 (3.1)
`29 (90.6)
`
`0 (0.0)
`0 (0.0)
`2 (5.7)
`0 (0.0)
`33 (94.3)
`
`1 (2.9)
`1 (2.9)
`1 (2.9)
`0 (0.0)
`31 (91.2)
`
`0 (0.0)
`0 (0.0)
`0 (0.0)
`0 (0.0)
`34 (100)
`
`2 (6.9)
`0 (0.0)
`1 (3.4)
`1 (3.4)
`25 (86.2)
`
`4 (2.4)
`2 (1.2)
`4 (2.4)
`2 (1.2)
`152 (92.7)
`
`0.830
`
`0.360
`
`0.980
`
`0.770
`0.657
`0.547
`
`0.722
`
`0.912
`
`0.662
`
`0.922
`
`0.254
`0.652
`0.562
`0.265
`0.236
`
`*P-values from analysis of variance or Fisher‘s exact test.
` 0.1 mg, 0.5 mg, 1.0 mg or 1.5 mg dulaglutide.
`
`reductions in HbA1c were observed across the dulaglutide
`groups (P < 0.001) at endpoint. Reductions in HbA1c were
`greater than placebo for each of
`the dulaglutide doses
`(P < 0.001) except the Dula 0.1 group (P = 0.069);
`least-
`squares mean difference (95% CI): Dula 0.1, )4 ()8 to) 1)
`mmol ⁄ mol [)0.37 ()0.69 to )0.06) %]; Dula 0.5, )10 ()13 to
`)6) mmol ⁄ mol [)0.89 ()1.21 to )0.57) %]; Dula 1.0, )11
`()15 to )8) mmol ⁄ mol [)1.04 ()1.36 to )0.72) %]; and
`Dula 1.5, )11 ()15 to )8) mmol ⁄ mol [)1.04 ()1.39 to )0.70)
`%] (Fig. 3b); change in the placebo group was least-squares
`mean  se: 0  1 mmol ⁄ mol (0.01  0.13%). HbA1c reduc-
`tions in Dula 0.5, Dula 1.0 and Dula 1.5 were greater than
`Dula 0.1 (P £ 0.001). There was no difference among the other
`dulaglutide dose groups.
`
`Secondary endpoints
`
`At endpoint, dose-dependent reductions in mean daily plasma
`glucose and fasting plasma glucose were observed across all
`doses (P < 0.001) (Fig. 3c and d). Changes in fasting plasma
`
`glucose were greater than placebo for each of the doses
`(P < 0.001), except for the Dula 0.1 group (P = 0.456); least-
`squares mean difference (95% CI): Dula 0.1, )0.43 ()1.06 to
`0.19) mmol ⁄ l; Dula 0.5, )1.47 ()2.12 to )0.83) mmol ⁄ l;
`Dula 1.0, )1.66 ()2.31 to )1.02) mmol ⁄ l; and Dula 1.5, )1.87
`()2.56 to )1.19) mmol ⁄ l (Fig. 3d); change in the placebo group
`was least-squares mean  se: )0.21  0.25 mmol ⁄ l. Dose-
`dependent reductions in mean pre-meal and postprandial plas-
`ma glucose from 7-point self-monitored plasma glucose were
`observed at endpoint in response to treatment with dulaglutide
`(P £ 0.003, data not shown). Additionally, decreases in mean
`pre-meal and postprandial plasma glucose in Dula 0.5,
`Dula 1.0 and Dula 1.5 groups were significantly greater than
`placebo (data not shown).
`There was an increasing trend across groups in the per cent
`of patients achieving HbA1c < 53 mmol ⁄ mol (< 7.0%) at
`(P < 0.001): placebo (21%), Dula 0.1 (47%),
`endpoint
`Dula 0.5 (73%), Dula 1.0 (75%) and Dula 1.5 (71%). There
`was also an increasing trend in the per cent of patients
`achieving HbA1c £ 48 mmol ⁄ mol (£ 6.5%) (P < 0.001): pla-
`
`ª 2012 Eli Lilly and Company.
`Diabetic Medicine ª 2012 Diabetes UK
`
`1263
`
`
`
`e
`
`MPI EXHIBIT 1109 PAGE 4
`
`

`

`DIABETICMedicine
`
`Dulaglutide monotherapy and glycaemic control • G. Grunberger et al.
`
`0.2
`0.0
`–0.2
`–0.4
`–0.6
`–0.8
`–1.0
`–1.2
`–1.4
`
`Δ HbA1c (%)
`
`HbA1c (mmol/mol)
`
`72
`
`66
`60
`54
`48
`42
`1
`
`03
`
`8
`4
`Time (weeks)
`
`12
`
`(a)
`
`9.0
`8.5
`8.0
`7.5
`7.0
`6.5
`6.0
`5.0
`0.0
`Baseline
`
`HbA1c (%)
`
`01
`
`–1
`
`–2
`
`–3
`
`Δ FPG (mmol/l)
`
`60
`50
`40
`30
`20
`10
`0
`–10
`–20
`
`Change from baseline
`
`HOMA2-%B (%)
`
`Δ mean daily PG (mg/dl)
`
`18
`
`0
`
`–18
`
`–36
`
`–54
`
`*
`
`†*
`
`Placebo
`Dula 0.1
`Dula 0.5
`Dula 1.0
`Dula 1.5
`
`(c)
`
`0
`
`12
`8
`4
`Time (weeks)
`
`01
`
`–1
`
`–2
`
`–3
`
`Δ mean daily PG (mmol/l)
`
`(e)
`
`100
`
`75
`
`50
`
`25
`
`0
`
`Per cent of patients achieving
`
`HbA1c target (%)
`
`0
`
`–3
`
`–6
`
`–9
`
`–12
`
`Δ HbA1c (mmol/mol)
`
`Δ FPG (mg/dl)
`
`–15
`
`18
`
`0
`
`–18
`
`–36
`
`–54
`
`*
`
`†*
`
`*
`
`†*
`
`*†
`
`(b)
`
`8
`4
`Time (weeks)
`
`12
`
`0
`
`(d)
`
`12
`8
`4
`Time (weeks)
`
`0
`
`(f)
`
`< 53 mmol/mol ≤ 48 mmol/mol
`(≤ 6.5%)
`(< 7.0%)
`HbA1c target
`Placebo
`Dula 0.1
`
`0
`
`8
`4
`Time (weeks)
`
`12
`
`Dula 0.5
`
`Dula 1.0
`
`Dula 1.5
`
`FIGURE 3 Glycaemic control in patients with Type 2 diabetes (intent-to-treat population, n = 164) in response to treatment with placebo (n = 32), Dula
`0.1 (n = 35), Dula 0.5 (n = 34), Dula 1.0 (n = 34) or Dula 1.5 mg (n = 29): (a) HbA1c by study visit (mean  sd); (b) least-squares mean change from
`baseline in HbA1c by study visit (least-squares mean  se); (c) least-squares mean change from baseline in mean daily plasma glucose (PG) from 7-point self
`monitored plasma glucose (7-point SMPG) profile by visit; (d) least-squares mean change from baseline in fasting plasma glucose (FPG) by study visit;
`(e) percentage of patients achieving HbA1c targets of < 53 mmol ⁄ mol (< 7.0%) and £ 48 mmol ⁄ mol (£ 6.5%) at week 12. Statistically significant dose
`effect is observed for both targets, P < 0.001 by Cochran–Armitage trend exact test; and (f) least-squares mean change in HOMA2-%B by visit. Glucose
`values in mg ⁄ dl were converted to mmol ⁄ l by dividing by 18. *P < 0.05 vs. baseline;  P < 0.05 vs. placebo.
`
`cebo (7%), Dula 0.1 (15%), Dula 0.5 (53%), Dula 1.0 (50%)
`and Dula 1.5 (52%) (Fig. 3e).
`At week 12, dose-dependent increases in the homeostasis
`model assessment of b-cell
`function (HOMA2-%B) were
`observed across the dulaglutide groups (P = 0.036). Increases
`were larger in each of the dulaglutide dose groups (P £ 0.013)
`except the Dula 0.1 group (P = 0.325) compared with placebo;
`least-squares mean difference (95% CI): Dula 0.1, 15.2 ()3.8
`to 34.3) %; Dula 0.5, 33.7 (14.2 to 53.2) %; Dula 1.0, 41.1
`(20.6 to 61.6) %; and Dula 1.5, 31.4 (10.4 to 52.3) %; change
`in the placebo group (least-squares mean  se) was
`)2.1  7.4%; (Fig. 3f). No significant changes were observed
`in any dulaglutide group for HOMA2-%S.
`Changes in body weight at week 12 (least-squares mean -
`se) were )1.4  0.5 kg for placebo, )0.2  0.4 kg for
`Dula 0.1, )0.3  0.4 kg for Dula 0.5, )1.1  0.4 kg for
`Dula 1.0 and )1.5  0.5 kg for Dula 1.5. Dose-dependent
`
`reductions in body weight were observed across the dulaglutide
`groups at week 12 (P = 0.009), but were not significant
`when compared with placebo. This outcome may be partially
`related to two patients in the placebo group who experienced
`weight loss of 11.2 and 11.3 kg as a result of haemorrhagic
`pancreatitis and participation in a weight-loss programme,
`respectively.
`
`Safety and tolerability
`
`Overall, 51.8% (n = 85) of patients reported ‡ 1 treatment-
`emergent adverse event during the treatment period, with no
`significant trend across groups (see also Supporting Informa-
`tion, Table S1). The most frequent treatment-emergent adverse
`events were nausea, diarrhoea, and nasopharyngitis, with
`overall incidences of 7.9% (n = 13), 6.1% (n = 10) and 5.5%
`(n = 9), respectively; there was no significant trend across
`
`1264
`
`ª 2012 Eli Lilly and Company.
`Diabetic Medicine ª 2012 Diabetes UK
`
`
`
`e
`
`MPI EXHIBIT 1109 PAGE 5
`
`

`

`Research article
`
`DIABETICMedicine
`
`groups (Table S1). Four patients (2.4%) discontinued because
`of adverse events (Table 1).
`Investigators reported four cases of serious adverse events
`(2.4%). Two were considered possibly related to study drug:
`haemorrhagic pancreatitis associated with cholelithiasis (pla-
`cebo) and abdominal pain ⁄ distension (Dula 1.5). The other
`two cases were breast cancer (Dula 0.5), diagnosed based on a
`mammogram completed 6 months prior to study enrolment,
`and atrial flutter (Dula 1.0). No deaths occurred during the
`study.
`There were no severe hypoglycaemic events reported. The
`overall incidence of hypoglycaemia was not different among
`groups (P = 0.822): placebo, 3.1% (n = 1 ⁄ 32); Dula 0.1, 5.7%
`(n = 2 ⁄ 35); Dula 0.5, 8.8% (n = 3 ⁄ 34); Dula 1.0, 5.9%
`(n = 2 ⁄ 34); and Dula 1.5, 10.3% (n = 3 ⁄ 29).
`Systolic and diastolic blood pressures were not different
`between dulaglutide groups and placebo. Changes from base-
`line in electrocardiogram-derived heart rate were not different
`between dulaglutide groups and placebo (see also Supporting
`Information, Table S2).
`There were no significant differences between the groups in
`levels of pancreatic enzymes (lipase, pancreatic amylase and
`total amylase) at endpoint. Two patients (Dula 1.5 mg) dem-
`onstrated increases > 3 times the upper limit of normal on
`consecutive testing during the study; both underwent computed
`tomography (CT) scans and results were within normal range.
`Administration of study drug was not interrupted and both
`patients showed improvement or normalization of laboratory
`findings during the trial.
`A treatment-emergent anti-dulaglutide antibody was reported
`in one patient in the Dula 1.0 group (4-fold or greater increase
`in antibody titre from baseline); antibody titre 1:64 at week 4
`and 1:8 at week 12. One patient (Dula 0.1) reported treatment-
`emergent skin rash and skin exfoliation (anti-dulaglutide
`antibody negative).
`
`Discussion
`
`This dose–response study assessed the effect of a range of doses
`of dulaglutide, a once-weekly administered GLP-1 analogue, on
`HbA1c over a 12-week treatment period. The results of the
`study demonstrate a significant dose–response effect on change
`in HbA1c from baseline to 12-week endpoint.
`The trial involved patients earlier in the course of disease,
`supported by the relatively short mean duration of diabetes and
`the relatively modest increase in mean HbA1c level at the end of
`the lead-in period after discontinuation of metformin. This
`population was selected because a less advanced b-cell secre-
`tory deficit and modest hyperglycaemia were characteristics
`suitable for a monotherapy, placebo-comparator study. The
`HbA1c-lowering effect of the higher dulaglutide doses was
`clinically relevant [up to )11  1 mmol ⁄ mol ()1.0  0.1%)
`in the Dula 1.5 group]. The magnitude of this effect, when
`compared with that reported for other GLP-1 analogues in
`similar patient populations, suggests that this compound is
`
`effective in reducing elevated glucose levels [2,7,12]. No signifi-
`cant difference in glucose lowering between the three higher dose
`groups (0.5, 1.0 and 1.5 mg) was observed, but relatively lower
`baseline HbA1c levels and near-normal levels at endpoint in all
`three groups may have decreased the ability to fully differentiate
`the effect of increasing doses. As maximal reductions in average
`and fasting plasma glucose were not observed until 2–4 weeks of
`completed treatment (Fig. 3c and d), it is also possible that a
`longer treatment period (> 12 weeks) is needed for the full effect
`on haemoglobin glycosylation to be demonstrated [13]. Treat-
`ment with dulaglutide increased b-cell function as measured by
`the increase in HOMA2-%B. Caution must be taken in the
`interpretation of this observation because this is a short study.
`The increase in HOMA2-%B may just reflect GLP-1 agonist-
`mediated increase in insulin secretion and may not translate into
`long-term improvement in b-cell function. These observations
`are consistent with the previous 16-week, Phase-2, placebo-
`controlled trial showing that treatment with dulaglutide caused
`significant decreases in HbA1c and blood glucose, and improve-
`ment in b-cell function in patients with Type 2 diabetes treated
`with two other oral anti-diabetes drugs [14].
`The effect of exenatide, liraglutide, albiglutide and exenatide
`once weekly on body weight has been studied as a secondary
`objective in several large clinical trials. When compared with
`placebo or other glucose-lowering agents, these medications
`are, on average, associated with no weight gain or with weight
`loss [2,3,5,7,12]. In this study, dulaglutide was associated with
`dose-dependent reductions in body weight similar to other
`GLP-1 analogues; however, no difference was observed in
`comparison with placebo. One possible reason for this outcome
`is the large change in body weight in two patients in the pla-
`cebo group. An exploratory, post-hoc analysis that excluded
`outliers from all treatment groups ( 3 sd) indicated that these
`two patients could explain, to a significant extent, the body
`weight reductions observed in the placebo group and the lack
`of difference between the placebo arm and each dulaglutide
`dose. More comprehensive assessments will be conducted in
`larger, ongoing, Phase 3 studies.
`The most commonly reported treatment-emergent adverse
`events were nausea and diarrhoea, consistent with the adverse
`event profile of other agents in this class [2,15,16]. The inci-
`dence of these events in dulaglutide-treated patients was not
`different from that in placebo-treated patients and the severity
`infrequently resulted in discontinuation (one patient from
`Dula 0.5; two patients from Dula 1.5), indicating acceptable
`gastrointestinal tolerability.
`Because of previously reported cases of acute pancreatitis in
`conjunction with the use of marketed GLP-1 analogues, mea-
`surements of serial pancreatic enzymes were included to assess
`the predictive value for pancreatic adverse events. As mild
`increases in these laboratory parameters, without any known
`clinical association, frequently occur in patients with Type 2
`diabetes [17], patients with greater changes (‡ 3 times the
`upper limit of normal) were of special interest. The six patients
`who were noted to have enzyme elevations above this threshold
`
`ª 2012 Eli Lilly and Company.
`Diabetic Medicine ª 2012 Diabetes UK
`
`1265
`
`
`
`e
`
`MPI EXHIBIT 1109 PAGE 6
`
`

`

`DIABETICMedicine
`
`Dulaglutide monotherapy and glycaemic control • G. Grunberger et al.
`
`were in the Dula 1.0 and Dula 1.5 groups, but none presented
`with clinical symptoms. In only two patients was an increase in
`pancreatic enzymes hyperenzymemia confirmed on subsequent
`testing (both had a normal CT scan of the pancreas). These
`laboratory findings did not require change in the treatment
`regimen. Importantly, the majority (four patients) demon-
`strated increased pancreatic enzyme levels before randomiza-
`tion. Of note,
`the patient who presented with acute
`haemorrhagic pancreatitis shortly after randomization to the
`placebo-treated group had normal pancreatic enzymes at
`baseline. Therefore,
`the observed changes
`in pancreatic
`enzymes do not appear to be predictive of clinical outcomes
`and their relevance in this setting remains to be determined.
`Changes in heart rate have been reported with the use of
`marketed GLP-1 analogues [16,18]. Administration of dula-
`glutide did not result in any significant difference in heart rate
`in comparison with non-exposed individuals. Similar to the
`assessment of the effects of dulaglutide on heart rate, no sig-
`nificant changes in systolic and diastolic blood pressures were
`observed. There were no significant changes observed in lipid
`values associated with dulaglutide treatment (see also Sup-
`porting Information, Table S3).
`In conclusion, once-weekly administration of dulaglutide for
`12 weeks in patients with Type 2 diabetes (who were anti-
`hyperglycemic medication-naı¨ve or had previously been treated
`with metformin monotherapy) resulted in dose-dependent
`improvement in glycaemic control, with a significant increase in
`b-cell function. Dose-dependent decrease in body weight was
`observed; however, the decrease in body weight in the placebo
`group attenuated the significance of this analysis. In this study,
`dulaglutide displayed an acceptable safety and tolerability
`profile. Further studies are needed to fully characterize the
`effects of dulaglutide on safety and efficacy parameters, and
`clinical investigation is ongoing in the large Phase 3 Assessment
`of Weekly AdministRation of LY2189265 in Diabetes
`(AWARD) programme.
`
`Competing interests
`
`GG has commonality of interest with Eli Lilly and Company,
`Amylin, Merck, Novo Nordisk, GSK, Takeda, Merck, Sanofi-
`Aventis, Astra Zeneca, Bristol Myers Squibb by the nature of
`research grants and ⁄ or
`speaking engagements. GGS has
`commonality of interest with Eli Lilly and Company, Amylin,
`Xoma, Takeda and Roche by the nature of research grants.
`FTB, RB and ZM are employees and shareholders of Eli Lilly
`and Company. AC has nothing to declare.
`
`Acknowledgements
`
`This study was supported by Eli Lilly and Company. The
`authors would like to thank the patients and investigators who
`participated in the study. The authors would also like to
`thank Sherry A. Martin MD for her critical review of the
`manuscript.
`
`References
`
`incretin
`therapeutics for type 2 diabetes:
`1 Verspohl EJ. Novel
`hormone mimetics (glucagon-like peptide-1 receptor agonists)
`and dipeptidyl peptidase-4 inhibitors. Pharmacol Ther 2009; 124:
`113–138.
`2 Moretto TJ, Milton DR, Ridge TD, Macconell LA, Okerson T,
`Wolka AM et al. Efficacy and tolerability of exenatide monotherapy
`over 24 weeks in antidiabetic drug-naive patients with type 2 dia-
`betes: a randomized, double-blind, placebo-controlled, parallel-
`group study. Clin Ther 2008; 30: 1448–1460.
`3 Vilsboll T, Zdravkovic M, Le-Thi T, Krarup T, Schmitz O, Courreges
`JP et al. Liraglutide, a long-acting human glucagon-like peptide-1
`analog, given as monotherapy, significantly improves glycemic con-
`trol and lowers body weight without risk of hypoglycemia in patients
`with type 2 diabetes. Diabetes Care 2007; 30: 1608–1610.
`4 Aschner P, Katzeff HL, Guo H, Sunga S, Williams-Herman D,
`Kaufman KD et al. Efficacy and safety of monotherapy of sitagliptin
`compared with metformin in patients with type 2 diabetes. Diabe-
`tes Obes Metab 2010; 12: 252–261.
`5 Kim D, MacConell L, Zhuang D, Kothare PA, Trautmann M, Fin-
`eman M et al. Effects of once-weekly dosing of a long-acting release
`formulation of exenatide on glucose control and body weight in
`subjects with type 2 diabetes. Diabetes Care 2007; 30: 1487–1493.
`6 Matthews JE, Stewart MW, De Boever EH, Dobbins RL, Hodge RJ,
`Walker SE et al. Pharmacodynamics, pharmacokinetics, safety, and
`tolerability of albiglutide, a long-acting glucagon-like peptide-1
`mimetic, in patients with type 2 diabetes. J Clin Endocrinol Metab
`2008; 93: 4810–4817.
`7 Rosenstock J, Reusch J, Bush M, Yang F, Stewart M. Potential
`