You are on page 1of 7

Emerging Treatments and Technologies

O R I G I N A L A R T I C L E

Efficacy and Safety Comparison of Liraglutide,


Glimepiride, and Placebo, All in Combination
With Metformin, in Type 2 Diabetes
The LEAD (Liraglutide Effect and Action in Diabetes)-2 study
MICHAEL NAUCK, MD, PHD1 ISMAIL H. MITHA, MD6 pies, but both can elicit weight gain,
ANDERS FRID, MD2 MILAN ZDRAVKOVIC, MD, PHD7 whereas sulfonylureas can elicit hypogly-
KJELD HERMANSEN, MD3 MARIA DURING
, PHD7 cemia. Glucagon-like peptide-1 (GLP-1)
NALINI S. SHAH, MD4 DAVID R. MATTHEWS, MD8 has received attention as a therapeutic
TSVETALINA TANKOVA, MD5 FOR THE LEAD-2 STUDY GROUP*
agent to provide glycemic control while
avoiding hypoglycemia and weight gain.
GLP-1 is an incretin hormone pro-
OBJECTIVE The efficacy and safety of adding liraglutide (a glucagon-like peptide-1 re- duced in the gut that stimulates glucose-
ceptor agonist) to metformin were compared with addition of placebo or glimepiride to met- dependent endogenous insulin secretion,
formin in subjects previously treated with oral antidiabetes (OAD) therapy.
decreases glucagon secretion, slows gas-
RESEARCH DESIGN AND METHODS In this 26-week, double-blind, double- tric motility and emptying, and reduces
dummy, placebo- and active-controlled, parallel-group trial, 1,091 subjects were randomly appetite and food intake (2,3). Further-
assigned (2:2:2:1:2) to once-daily liraglutide (either 0.6, 1.2, or 1.8 mg/day injected subcutane- more, native GLP-1 stimulates -cell pro-
ously), to placebo, or to glimepiride (4 mg once daily). All treatments were in combination liferation in animal models and inhibits
therapy with metformin (1g twice daily). Enrolled subjects (aged 2579 years) had type 2 apoptosis in vitro, which may increase
diabetes, A1C of 711% (previous OAD monotherapy for 3 months) or 710% (previous OAD -cell mass and function (2). However,
combination therapy for 3 months), and BMI 40 kg/m2. native GLP-1 is not a viable therapeutic
agent because it has a short half-life of 2
RESULTS A1C values were significantly reduced in all liraglutide groups versus the pla-
min resulting from rapid degradation by
cebo group (P 0.0001) with mean decreases of 1.0% for 1.8 mg liraglutide, 1.2 mg liraglutide,
and glimepiride and 0.7% for 0.6 mg liraglutide and an increase of 0.1% for placebo. Body the enzyme dipeptidyl peptidase 4
weight decreased in all liraglutide groups (1.8 2.8 kg) compared with an increase in the (DPP-4) and rapid renal elimination (4).
glimepiride group (1.0 kg; P 0.0001). The incidence of minor hypoglycemia with liraglutide Liraglutide is a human GLP-1 analog
(3%) was comparable to that with placebo but less than that with glimepiride (17%; P with 97% homology to native GLP-1 (5).
0.001). Nausea was reported by 1119% of the liraglutide-treated subjects versus 3 4% in the The addition of a fatty acid (palmitate)
placebo and glimepiride groups. The incidence of nausea declined over time. side chain and a single amino acid substi-
tution produces self-association of the
CONCLUSIONS In subjects with type 2 diabetes, once-daily liraglutide induced similar molecule that prolongs absorption from
glycemic control, reduced body weight, and lowered the occurrence of hypoglycemia compared
the subcutaneous depot. The fatty acid
with glimepiride, when both had background therapy of metformin.
side chain also promotes albumin binding
Diabetes Care 32:8490, 2009 that renders the molecule resistant to deg-
radation by DPP-4. The resultant plasma
half-life of liraglutide is substantially pro-

T
ype 2 diabetes is a progressive dis- cations (1). An oral antidiabetes drug
longed to 13 h compared with the very
ease characterized by declining (OAD) such as metformin is commonly
short half-life of native human GLP-1
-cell function that, in concert with used in first-line pharmacotherapy for
(6,7). The pharmacokinetic profile makes
insulin resistance, leads to loss of glyce- type 2 diabetes. Sulfonylureas and thiazo-
liraglutide suitable for once-daily admin-
mic control and eventual diabetes compli- lidinediones are alternative OAD thera-
istration while achieving 24-h glycemic
control, in contrast to exenatide which
From the 1Diabeteszentrum Bad Lauterberg, Harz, Germany; the 2Oresund Diabetes Team AB, Lund, needs to be injected twice daily because it
Sweden; 3Aarhus University Hospital, Aarhus, Denmark; 4Seth G.S. Medical College and KEM Hospital, has a half-life of 2.4 h (5,8,9). Depend-
Mumbai, India; the 5Clinic of Diabetology, Sophia, Bulgaria; 6Benmed Hospital, Benoni, Republic of
South Africa; 7Novo Nordisk, Bagsvaerd, Denmark; and the 8Oxford Centre for Diabetes Endocrinology
ing upon dose and duration of treatment,
and Metabolism, Oxford, U.K. type 2 diabetic subjects receiving lira-
Corresponding author: Michael A. Nauck, nauck@diabeteszentrum.de. glutide had mean reductions in A1C of
Received 22 July 2008 and accepted 28 September 2008. 0.8 1.5%, reductions in fasting blood
Published ahead of print at http://care.diabetesjournals.org on 17 October 2008. DOI: 10.2337/dc08-1355.
Clinical trial reg. no. NCT00318461, clinicaltrials.gov.
glucose of 1.4 3.4 mmol/l, and reduc-
*A complete list of the LEAD-2 study investigators can be found in an online appendix available at http:// tions in body weight of 1.23.0 kg (10
dx.doi.org/10.2337/dc08-1355. 13). In a recent 52-week monotherapy
2009 by the American Diabetes Association. Readers may use this article as long as the work is properly study, 51% of subjects treated with 1.8
cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons. mg liraglutide reached the American Dia-
org/licenses/by-nc-nd/3.0/ for details.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby betes Association (ADA) target of A1C
marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 7.0% and had weight loss in the first 16

84 DIABETES CARE, VOLUME 32, NUMBER 1, JANUARY 2009


Nauck and Associates

weeks (2.4 kg) that was sustained groups received an injection of liraglutide bin Standardization Program. Subjects
throughout the remainder of the study placebo. were provided with MediSense Precision
(13). Randomization was performed using Xtra/MediSense Optium glucose meters
This current trial is part of a phase 3 a telephone-based or web-based random- (Abbott, Maidenhead, U.K.) calibrated
clinical development program for lira- ization system. Subjects were randomly to plasma glucose to determine self-
glutide. The trial investigates whether assigned to the lowest available random- measured plasma glucose and recorded
glycemic control (measured by A1C) ization number and stratified with respect values in their diaries. Serum insulin and
achieved by type 2 diabetic subjects us- to their previous use of OAD mono- C-peptide values were determined using a
ing combination therapy of liraglutide therapy or combination therapy. Subjects chemiluminescence immunoassay, and
and metformin is significantly better completing the study could enroll in an proinsulin was measured in serum using
than that achieved with metformin 18-month open-label extension period. an enzyme-linked immunosorbent assay.
monotherapy or at least as good as that Randomization to treatment occurred Safety variables included adverse
achieved with combination therapy of after a 3-week forced metformin titration events, vital signs, electrocardiogram,
metformin and glimepiride. period (dose increased up to 2,000 mg/ biochemical and hematology measures,
day: 1,000 mg in the morning and 1,000 and subject-reported hypoglycemic epi-
RESEARCH DESIGN AND mg in the evening) followed by a 3-week sodes (based on symptoms and plasma
METHODS Adult subjects with metformin maintenance period. Subjects glucose 3.1 mmol/l). Minor hypoglyce-
type 2 diabetes were screened and en- taking metformin at enrollment could go mic episodes were self-treated; major ep-
rolled if they were 18 80 years of age, through a modified titration period or ad- isodes required third-party assistance.
had A1C between 7 and 11% (prestudy vance directly to the metformin mainte-
OAD monotherapy for 3 months) or be- nance period. After randomization, Statistical analysis
tween 7 and 10% (prestudy combination subjects underwent a 2- and 3-week titra- The analyses of efficacy end points were
OAD therapy for 3 months), and had tion period for liraglutide (up to 0.6, 1.2, or based on the intent-to-treat population
BMI 40 kg/m2. Subjects were excluded 1.8 mg, as per randomization, at 0.6-mg defined as subjects who were exposed to
if they had used insulin during the previ- increases per week) and glimepiride (up to at least one dose of trial product and had
ous 3 months (except short-term treat- 4 mg, with 1-, 2-, and 4-mg doses at weeks one postbaseline measurement of the
ment). The protocol was approved by 1, 2, and 3). Glimepiride (active and pla- parameter. Each end point was analyzed
local institutional review boards, and all cebo) was taken orally once daily in the using an ANCOVA model with treatment,
subjects provided written informed con- morning. Liraglutide (active or placebo) country, and previous antidiabetic treat-
sent before initiation of any trial-related was injected subcutaneously once daily at ment as fixed effects and baseline as the
activities. The study was conducted in ac- any time of the day in the upper arm, abdo- covariate. Missing data were imputed as
cordance with the Declaration of Helsinki men, or thigh using a pen injector device. the last observation carried forward. Sam-
and Good Clinical Practice guidelines Subjects were encouraged to inject lira- ple size calculations were based on show-
(14). glutide at the same time each day. ing A1C and body weight differences of
In this 26-week, double-blind, double- The titration period was followed by a 0.5 and 3%, respectively, after 6 months
dummy, active-control, parallel-group, 23- or 24-week maintenance period dur- of treatment. The assumed standard de-
multicenter (170 sites), multinational (21 ing which the doses of study drugs were viation for A1C and the coefficient of
countries) trial, subjects were randomly to be maintained. However, metformin variance for weight were 1.2 and 3%,
assigned (2:2:2:1:2) to receive one of could be decreased to a minimum of respectively. The combined power (cal-
three once-daily doses of liraglutide (0.6, 1,500 mg/day in the case of unacceptable culated as the product of the marginal
1.2, or 1.8 mg/day; Novo Nordisk, Bags- hypoglycemia or other adverse events but powers for A1C and weight) was at least
vaerd, Denmark) injected subcutaneously had to be maintained between 1,500 and 85%.
in combination with metformin, to re- 2000 mg/day during the maintenance Superiority or noninferiority of glyce-
ceive liraglutide placebo with metformin period. mic control with liraglutide versus com-
monotherapy (placebo group), or to re- The primary outcome measure was parators was concluded if the upper limit
ceive combination therapy with glime- change in A1C at the end of the study. of the two-sided 95% CI for the treatment
piride and metformin (4 mg glimepiride Secondary end points included changes difference was 0%; noninferiority was
once daily with the first meal of the day). in body weight, fasting plasma glucose concluded if it was 0.4%. The propor-
The most relevant position for initiating (FPG), 7-point plasma glucose profiles tion of subjects achieving A1C targets
treatment with GLP-1 may be after met- (before each meal, 90 min after breakfast, (ADA target 7%; American Association
formin failure, but to facilitate recruit- lunch, and dinner, and at bedtime), and of Clinical Endocrinologists [AACE] tar-
ment into the trial other monotherapy or -cell function based on fasting insulin, get 6.5%) was compared between treat-
combination treatments were allowed. fasting C-peptide, fasting proinsulin-to- ments using a logistic regression model
Accordingly, the objective of this study insulin ratio, and the homeostasis model with treatment and baseline A1C as co-
was to compare the efficacy and safety of assessment index of -cell function variates. CIs for secondary end points
liraglutide with both placebo and another (HOMA-B) (15). Laboratory analyses were corrected using Dunnetts test.
commonly used therapeutic option were performed by a central laboratory Hypoglycemic episodes were analyzed
(glimepiride) after metformin failure. The (MDS Pharma Services in Canada, using a general linear model including
double-dummy design required that sub- France, Germany, Singapore, and Swit- treatment as a fixed effect. Values are ex-
jects in the liraglutide and placebo groups zerland and Laboratories Hildago in Ar- pressed as means SD unless otherwise
received a glimepiride placebo, whereas gentina). A1C was assayed by a method noted. The significance level was set at
subjects in the glimepiride and placebo certified by the National Glycohemoglo- P 0.05.

DIABETES CARE, VOLUME 32, NUMBER 1, JANUARY 2009 85


LEAD-2 study

Table 1Characteristics of enrolled population and subject disposition

Once-daily liraglutide
Once-daily 4 mg
0.6 mg 1.2 mg 1.8 mg glimepiride Placebo
Sex: male/female (%) 62/38 54/46 59/41 57/43 60/40
Age (years) 56 11 57 9 57 9 57 9 56 9
Race: C/B/A/O (%) 84/2/13/2 88/4/8/1 88/2/7/2 89/2/9/1 88/3/7/3
BMI (kg/m2) 30.5 4.8 31.1 4.8 30.9 4.6 31.2 4.6 31.6 4.4
Duration of diabetes (years) 75 75 85 85 86
Prestudy OAD treatment
Monotherapy 81 (34) 91 (38) 83 (34) 89 (37) 41 (34)
Metformin 70 (86) 78 (86) 72 (87) 82 (92) 38 (93)
Sulfonylurea 9 (11) 12 (13) 11 (13) 7 (8) 3 (7)
Repaglinide 2 (3) 1 (1) 0 0 0
Combination therapy 161 (67) 150 (62) 159 (66) 155 (63) 81 (66)
A1C (%) 8.4 0.9 8.3 1.0 8.4 1.0 8.4 1.0 8.4 1.1
FPG (mmol/l) 10.2 2.4 9.9 2.3 10.1 2.3 10.0 2.6 10.0 2.3
SBP (mmHg) 131 14 132 14 131 14 132 16 135 16
DBP (mmHg) 80 8 80 10 79 8 80 8 81 9
Randomized 242 241 242 244 122
Exposed (ITT and safety populations) 242 240 242 242 121
Completers 208 (86) 197 (82) 191 (79) 210 (86) 74 (61)
Withdrawals 34 (14) 44 (18) 51 (21) 34 (14) 48 (39)
Adverse events 11 (5) 23 (10) 29 (12) 8 (3) 2 (2)
Nausea/vomiting/diarrhea 3 (1) 13 (5) 20 (8) 0 0
Ineffective therapy 19 (8) 8 (3) 13 (5) 9 (4) 29 (24)
Noncompliance 2 (1) 4 (2) 4 (2) 5 (2) 4 (3)
Other 2 (1) 9 (4) 5 (2) 12 (5) 13 (11)
Data are means SD or n (%) unless otherwise noted. *Race: C, Caucasian; B, Black; A, Asian /Pacific Islander; O, other. DBP, diastolic blood pressure; ITT, intention
to treat.

RESULTS A total of 1,662 subjects liraglutide groups compared with the pla- ever, the baseline and end-of-study mean
were screened for the study; 571 subjects cebo group and the resulting 95% CIs A1C values in the monotherapy group
failed the screening criteria or withdrew demonstrated that liraglutide-treated were slightly less than those in the com-
consent, and the remaining 1,091 sub- subjects had superior glycemic control bination therapy group, and the resulting
jects were randomly assigned to treatment compared with those in the placebo change-from-baseline decreases appeared
after the metformin run-in period. Four group (0.6 mg liraglutide versus placebo to be slightly greater in the monotherapy
subjects were randomly assigned but 0.8% [95% CI 1.0 to 0.6]; 1.2 mg group than in the combination therapy
were withdrawn before receiving treat- liraglutide versus placebo 1.1% [1.3 group (Fig. 1D).
ment. Accordingly, the intent-to-treat to 0.9]; and 1.8 mg liraglutide versus pla- The percentages of subjects reaching
and safety (all subjects exposed to study cebo 1.1% [1.3 to 0.9]). Analysis of the ADA and AACE A1C goals were dose-
drug) populations comprised 1,087 sub- the estimated treatment difference in A1C dependent for liraglutide treatment in
jects and are summarized by treatment between liraglutide and glimepiride dem- overall subjects and in subjects with pre-
group in Table 1. Baseline characteristics onstrated that 1.2 and 1.8 mg liraglutide study OAD monotherapy or combination
of the study population appeared to be treatments were noninferior to treatment therapy (Fig. 1E and F). For the overall
balanced across treatment groups (Table with glimepiride (1.2 mg liraglutide versus population, a logistic regression analysis
1). The majority (65%) of the randomly glimepiride 0.0% [0.2 to 0.2] and 1.8 mg demonstrated that a significantly greater
assigned subjects were treated with two liraglutide versus glimepiride 0.0% percentage of subjects in all of the lira-
OADs before the study. [0.2 to 0.2]). glutide groups achieved the ADA and
Within the first 12 weeks of the study, AACE A1C goals than subjects in the re-
Efficacy mean A1C values for the overall popula- spective placebo groups (P 0.02 for all
At the end of the study, the mean A1C tion decreased from baseline for all lira- comparisons of liraglutide to placebo for
values for the overall population de- glutide treatment groups and for the both A1C goals). The percentages of sub-
creased by 0.7 0.1% for the 0.6 mg glimepiride group whereas a slight in- jects in the 1.2 and 1.8 mg liraglutide
liraglutide group and by 1.0 0.1% for crease was observed in the placebo group treatment groups achieving the ADA and
both the 1.2 and 1.8 mg liraglutide (Fig. 1A). The A1C profiles for subjects AACE A1C goals were comparable to
groups and for the glimepiride group and stratified by prestudy OAD therapy the percentage achieving goals in the
increased by 0.1 0.1% for the placebo (monotherapy or combination therapy) glimepiride group. However, for the over-
group (all values means SEM). The es- were similar in appearance to those of the all population, the ADA target was
timated treatment differences of all three overall population (Fig. 1B and C). How- achieved by significantly more subjects in

86 DIABETES CARE, VOLUME 32, NUMBER 1, JANUARY 2009


Nauck and Associates

Figure 1 A1C profiles for the overall study population (A), for subjects with prestudy oral monotherapy (B), and for subjects with prestudy oral
combination therapy (C). D: Change in A1C for the overall study population and for subjects treated by prestudy monotherapy or combination
therapy. Error bars in A, B, C, and D represent 2 SEM. Percentages of subjects achieving ADA (E) and AACE (F) A1C goals at the end of the study
were determined using a logistic regression analysis. Symbols for AC: pink boxes, 0.6 mg/day liraglutide; circles, 1.2 mg/day liraglutide; diamonds,
1.8 mg/day liraglutide; triangles, 4 mg/day glimepiride; gray boxes, placebo. QD, every day.

the 1.8 mg liraglutide group than in the FPG values decreased within the 2 increased in the placebo group, remain-
1.2 mg liraglutide group (42.4 vs. 35.3%, weeks of randomization in the liraglutide ing relatively stable thereafter. At the end
P 0.0265). groups and in the glimepiride group and of the study, FPG values were 9.1 2.5,

DIABETES CARE, VOLUME 32, NUMBER 1, JANUARY 2009 87


LEAD-2 study

8.5 2.6, 8.5 2.4, 8.9 2.5, and across all groups) for the three liraglutide was experienced by 10, 8, and 15% in the
10.7 3.2 mmol/l in the 0.6 mg lira- groups (decrease by 0.1) were compara- 0.6, 1.2, and 1.8 mg liraglutide groups,
glutide, 1.2 mg liraglutide, 1.8 mg lira- ble to those in the glimepiride group and respectively, and by 4% in the placebo
glutide, glimepiride, and placebo groups, were significantly different (P 0.0001) and glimepiride groups.
respectively. The decreases in FPG from than those in the placebo group, which The percentages of subjects with-
baseline for all of the liraglutide groups increased from baseline by 0.1. No signif- drawn because of adverse events were
(1.1, 1.6, and 1.7 mmol/l for 0.6, icant differences in the change-from- generally greater in the liraglutide groups
1.2, and 1.8 mg liraglutide groups, re- baseline fasting insulin and fasting than in the glimepiride or placebo groups
spectively) were significantly greater than C-peptide values were observed between (Table 1). Nausea, vomiting, and/or diar-
the increase observed for the placebo the liraglutide treatment groups com- rhea were the gastrointestinal events that
group (0.4 mmol/l, P 0.0001) but were pared with either the glimepiride or pla- led to the withdrawal of 36 liraglutide-
similar to the decrease observed for the cebo groups. The liraglutide treatment treated subjects (5% of all liraglutide-
glimepiride group (1.3 mmol/l). groups had improvements in HOMA-B of treated subjects) in a dose-dependent
Mean postprandial glucose values 63, 70, and 71% for the 0.6, 1.2, and 1.8 manner (Table 1). Most of these adverse
(mean of three meals), from self- mg liraglutide groups from baseline val- event withdrawals caused by gastrointes-
monitored 7-point plasma glucose mea- ues of 40, 47, and 43%, respectively. The tinal disorders occurred during the first
surements at the end of the study, glimepiride group had a similar improve- month of therapy. One subject in the 1.2
decreased from baseline in all treatment ment in the mean HOMA-B value to 68% mg liraglutide group and one in the
groups (1.7 mmol/l for 0.6 mg lira- from a baseline value of 43%. No im- glimepiride group were withdrawn for
glutide, 2.3 mmol/l for 1.2 mg lira- provement in HOMA-B was observed in acute pancreatitis during the study. Nei-
glutide, and 2.6 mmol/l for 1.8 mg the placebo group; baseline and end-of- ther subject had a prior history of pancre-
liraglutide, 2.5 mmol/l for glimepiride, study values were 45 and 43%, respec- atitis, and both subjects were hospitalized
and 0.6 mmol/l for placebo; P 0.001 tively. No significant differences were for 7 days and subsequently recovered.
for comparisons of all liraglutide groups observed between treatments for the ho- One death (cardiorespiratory arrest) was
to placebo; the decreases in the 1.2 and meostasis model assessment index of in- reported during the trial and occurred
1.8 mg liraglutide groups were compara- sulin resistance. during the metformin run-in period, be-
ble to those with glimepiride). The blood The 1.2 and 1.8 mg liraglutide groups fore randomization to treatment. A sec-
glucose values 90 min after breakfast, had significant reductions in systolic
ond subject had liver cirrhosis and
lunch, and dinner for the 1.8 mg lira- blood pressure (SBP) of 23 mmHg com-
hepatocellular carcinoma during the trial
glutide group at week 26 appeared to be pared with the increase in SBP of 0.4
and died after the trial had completed.
similar across the three meals (9.9, 9.5, mmHg observed in the glimepiride group
Both deaths were unrelated to liraglutide
and 9.7 mmol/l, respectively) and ap- (treatment difference compared with
treatment.
peared to be similar to the corresponding glimepiride: 1.2 mg liraglutide, 3.2
In general, minor hypoglycemia oc-
blood glucose values observed in the mmHg, P 0.0128; 1.8 mg liraglutide,
glimepiride group (10.0, 9.2, and 9.7 2.7 mmHg, P 0.0467). The decreases curred at low incidence (3% of subjects
mmol/l, respectively) and slightly less in SBP in the 0.6 mg liraglutide and pla- in the placebo and liraglutide groups and
than the corresponding values in the pla- cebo groups were 0.6 and 1.8 mmHg, re- 17% in the glimepiride group), resulting
cebo group (11.0, 10.7, and 10.9 mmol/l, spectively. On the basis of the SBP and in a relatively low rate of reported minor
respectively). weight profiles over time, the reduction in hypoglycemia (0.03 0.14 events/year for
Weight loss was dose dependent in SBP may not be fully explained by the the placebo and liraglutide groups and
the liraglutide treatment groups (1.8 reduction in body weight. Diastolic blood 1.23 events/year for the glimepiride
0.2, 2.6 0.2, and 2.8 0.2 kg for 0.6, pressure did not appear to change from group) that was significantly less for all
1.2, and 1.8 mg liraglutide groups, re- baseline for any groups. three liraglutide groups than for the
spectively) and was significantly different glimepiride group (P 0.001). No major
(P 0.0001) from the weight gain in the Safety hypoglycemic events were reported.
glimepiride group (1.0 0.2 kg). The Gastrointestinal disorders (nausea, vom- No clinically relevant between-
weight losses in the 1.2 and 1.8 mg lira- iting, and diarrhea) were the most fre- treatment differences were observed in
glutide groups were also significantly quently reported adverse events in the physical examination findings, laboratory
greater (P 0.01) than the weight loss in liraglutide groups and were reported dur- analyses (hematology and biochemistry
the placebo group (1.5 0.3 kg). The ing the course of the study by 35, 40, and analyses), electrocardiogram, or ophthal-
great majority of subjects either did not 44% of the subjects in the 0.6, 1.2, and moscopy. No significant differences in
report nausea or reported nausea for 7 1.8 mg liraglutide groups, respectively, calcitonin laboratory values were found
days during the first 8 weeks of treatment and by 17% in the placebo and between the liraglutide groups and either
and did not report nausea or reported glimepiride groups. Overall, nausea alone the placebo or glimepiride group. Slight
nausea for 7 days in weeks 8 26 of was experienced by 11, 16, and 19% of increases in pulse rate were observed in all
treatment (86 93% in the liraglutide the subjects in the 0.6, 1.2, and 1.8 mg treatment groups (23 bpm in the lira-
groups and 98 99% in the placebo and liraglutide groups, respectively; however, glutide groups and 1 bpm in the glime-
glimepiride groups). As such, the occur- 10% of the subjects were experiencing piride and placebo groups). The increases
rence of nausea did not appear to account nausea on a weekly basis by week 4. Vom- in pulse in the 0.6 and 1.2 mg liraglutide
for the weight loss. iting was experienced by 57% in the groups were significantly greater than that
The decreases in the proinsulin-to- liraglutide groups and by 1% in the pla- in the glimepiride group (P 0.012 and
insulin ratio from baseline (baseline of 0.4 cebo and glimepiride groups; diarrhea P 0.024, respectively).

88 DIABETES CARE, VOLUME 32, NUMBER 1, JANUARY 2009


Nauck and Associates

CONCLUSIONS This trial dem- therapy (66 and 39%, respectively, for 1.8 significantly greater rate of hypoglycemia.
onstrated that treatment with liraglutide mg liraglutide) (compare Fig. 1E and F). Thus, addition of once-daily liraglutide to
once daily (0.6, 1.2, or 1.8 mg) in combi- This finding probably results from the ad- metformin monotherapy is a viable treat-
nation with metformin provided im- dition of liraglutide as a second therapeu- ment option if weight gain and hypoglyce-
provement in A1C superior to that of tic agent in subjects whose type 2 diabetes mia are a concern.
metformin monotherapy (placebo group) may not be as advanced as that is those
and noninferior to that of combination subjects entering the study with prestudy
therapy of glimepiride and metformin combination therapy. Acknowledgments M.N. is a member on
(glimepiride group). The improvements In this study, liraglutide provided advisory boards for Amylin Pharmaceuticals,
in A1C for liraglutide-treated subjects re- 24-h glycemic control as demonstrated by ConjuChem, Eli Lilly, GlaxoSmithKline, Hoff-
ceiving prestudy OAD monotherapy were the similar postprandial blood glucose man-La Roche, Novartis Pharma, Novo Nor-
disk, Probiodrug, Restoragen, and sanofi-
greater (decrease of 1.3% for 1.8 mg lira- values after each of the three meals. The aventis; has received consulting fees from
glutide) than the improvements for sub- postprandial values of the 1.8 mg lira- AstraZeneca, Bayer Vital, Berlin-Chemie, Bio-
jects with prestudy oral combination glutide group also appeared to be similar vitrum, ConjuChem, Eli Lilly, Hoffman-La
therapy (decrease of 0.8%, 1.8 mg lira- to those of the glimepiride group and less Roche, Merck, Merck Sharp & Dohme, Novar-
glutide, significance not tested). The than those of the placebo group. tis Pharma, Novo Nordisk, Pfizer, Probiodrug,
monotherapy group probably had greater Nausea, vomiting, and diarrhea are Restoragen, sanofi-aventis, and Takeda; and
improvement because they added lira- known side effects of GLP-1 receptor has received research support from Amylin
glutide onto a monotherapy treatment agonists. The incidence of gastrointesti- Pharmaceuticals, Bayer Vital, Eli Lilly, Lifescan
whereas liraglutide was substituted for nal effects in this study increased dose Germany, Berlin-Chemie, Novartis Pharma,
one of the prestudy oral therapies (other dependently during the first 2 weeks of Novo Nordisk, sanofi-aventis, Probiodrug,
and Restoragen. K.H. has received honoraria
than metformin) in the subjects receiving liraglutide treatment but decreased for consultation and lecturing from Novo Nor-
prestudy combination therapy. The find- thereafter. In studies with treatment of disk, sanofi-aventis, and Merck. No other po-
ings of the current study demonstrate that exenatide, subjects reported nausea that tential conflicts of interest relevant to this
liraglutide is an effective treatment option also decreased in incidence over time article were reported.
for combination therapy with metformin (16,18,19). The dose dependence and We acknowledge the assistance of the
when subjects are not achieving glycemic transient nature of gastrointestinal side LEAD-2 study group, their staff, and clinical
control with metformin therapy alone. effects in the current trial suggest that in trial personnel and the subjects for participat-
The decrease in A1C observed in the clinical practice liraglutide should be ti- ing in the study. We also thank Brian Henni-
prestudy monotherapy group confirms trated from a starting dose of 0.6 mg/ gan, PhD, of Novo Nordisk for writing
the A1C decrease observed (1.45%) in day up to 1.2 mg/day and then up to 1.8 assistance.
Parts of this study were presented as a
subjects treated with liraglutide mono- mg/day. Apart from the gastrointestinal poster at the 68th annual meeting of the Amer-
therapy in a prior 14-week study (10). side effects, liraglutide treatment was gener- ican Diabetes Association, San Francisco, Cal-
Another recent liraglutide phase 3 study ally well tolerated and had a low incidence ifornia, 6 10 June 2008.
demonstrated that a subgroup of subjects of hypoglycemia that was comparable to
previously treated with diet and exercise that of the placebo group (metformin
and then treated with 1.8 mg liraglutide monotherapy). A reduction in SBP was also References
monotherapy had a mean A1C decrease of observed in the 1.2 and 1.8 mg liraglutide 1. UK Prospective Diabetes Study 16: over-
1.60% that was sustained over the 52- groups (23 mmHg) within 2 weeks of view of 6 years therapy of type II diabetes:
week trial period. In this subgroup, the treatment and could not be explained en- a progressive disease. Diabetes 44:1249
decrease in A1C was significantly greater tirely by weight loss that occurred over a 1258, 1995
than the decrease achieved in the corre- greater time frame. 2. Drucker DJ, Nauck MA: The incretin sys-
sponding glimepiride control group in Liraglutide was shown to be similar to tem: glucagon-like peptide-1 receptor
the same study (0.88%; P 0.05) (13). glimepiride in improving HOMA-B val- agonists and dipeptidyl petidase-4 inhib-
A1C decreases by subjects previously ues in the current study. Improvements in itors in type 2 diabetes. Lancet 368:1696
treated with metformin monotherapy -cell function with liraglutide treatment 1705, 2006
3. Holst JJ: The physiology of glucagon-like
have also been observed with the GLP-1 have been observed in several other stud- peptide 1. Physiol Rev 87:1409 1439,
agonist exenatide (decrease of 0.8%) or ies with liraglutide (10,20,21). Further- 2007
with the DPP-4 inhibitor sitagliptin (de- more, liraglutide has been shown to 4. Meier JJ, Nauck MA, Kranz D, Holst JJ,
crease of 0.65%) in populations of pa- increase -cell mass in animal models and Deacon CF, Gaeckle D, Schmidt WE,
tients with type 2 diabetes inadequately decrease -cell apoptosis in vitro (22,23). Gallwitz B: Secretion, degradation, and
controlled with previous metformin Such improvements in -cell function elimination of glucagon-like peptide 1
monotherapy (16,17). have the potential to delay type 2 diabetes and gastric inhibitory polypeptide in pa-
The percentages of liraglutide-treated progression. tients with chronic renal insufficiency and
subjects achieving A1C targets are reflec- The liraglutide-treated subjects gener- healthy control subjects. Diabetes 53:654
tive of the decreases in A1C values for the ally had improvements in glycemic param- 662, 2004
5. Knudsen LB, Nielsen PF, Huusfeldt PO,
overall population and for the subgroups eters (decreases in A1C, FPG, and Johansen NL, Madsen K, Pedersen FZ,
stratified by prior OAD therapy. Greater postprandial glucose from baseline) similar Thogersen H, Wilken M, Agerso H: Potent
percentages of subjects receiving pre- to those of glimepiride-treated subjects. derivatives of glucagon-like peptide-1
study monotherapy achieved the target However, the glimepiride group had a sig- with pharmacokinetic properties suitable
compared with their respective counter- nificant increase in weight compared with for once daily administration. J Med Chem
parts receiving prestudy combination the liraglutide groups and experienced a 43:1664 1669, 2000

DIABETES CARE, VOLUME 32, NUMBER 1, JANUARY 2009 89


LEAD-2 study

6. Deacon CF, Knudsen LB, Madsen K, a 12-week, double-blind, randomized, con- 2638 2643, 2006
Wiberg FC, Jacobsen O, Holst JJ: Dipep- trolled trial. Diabetes Care 27:13351342, 18. Buse JB, Henry RR, Han J, Kim DD, Fine-
tidyl peptidase IV resistant analogues of 2004 man MS, Baron AD: Effects of exenatide
GLP-1 which have extended metabolic 12. Nauck MA, Hompesch M, Filipczak R, Le (exendin-4) on glycemic control over 30
stability and improved biological activity. TD, Zdravkovic M, Gumprecht J: Five weeks in sulfonylurea-treated patients
Diabetologia 41:271278, 1998 weeks of treatment with the GLP-1 ana- with type 2 diabetes. Diabetes Care 27:
7. Agers H, Jensen LB, Elbrnd B, Rolan P, logue liraglutide improves glycemic con- 2628 2635, 2004
Zdravkovic M: The pharmacokinetics, trol and lowers body weight in patients 19. Kendall DM, Riddle MC, Rosenstock J,
pharmacodynamics, safety and tolerabil- with type 2 diabetes. Exp Clin Endocrinol Zhuang D, Kim DD, Fineman MS, Baron
ity of NN2211, a new long-acting GLP-1 Diabetes 114:417 423, 2006 AD: Effects of exenatide (exendin-4) on
derivative, in healthy men. Diabetologia 13. Garber A, Henry R, Ratner R, Garcia-Her-
glycemic control over 30 weeks in pa-
45:195202, 2002 nandez PA, Rodriguez-Pattzi H, Olvera-
tients with type 2 diabetes treated with
8. Elbrnd B, Jakobsen G, Larsen S, Agers Alvarez I, Hale PM, Zdravkovic M, Bode
H, Jensen LB, Rolan P, Sturis J, Hatorp V, B: Liraglutide versus glimepiride mono- metformin and a sulfonylurea. Diabetes
Zdravkovic M: Pharmacokinetics, phar- therapy for type 2 diabetes (LEAD-3 Care 28:10831091, 2005
macodynamics, safety and tolerability of a Mono): randomised, 52-week, phase III, 20. Mari A, Degn K, Brock B, Rungby J, Fer-
single dose of NN2211, a long-acting glu- double-blind, parallel-treatment trial. rannini E, Schmitz O: Effects of the long-
cagon-like peptide 1 derivative, in healthy Lancet. 2008 [Epub ahead of print]. DOI: acting human glucagon-like peptide-1
male patients. Diabetes Care 25:1398 10.1015/S0140-6736(08)61246-5 analog liraglutide on -cell function in
1404, 2002 14. World Medical Association: Declaration normal living conditions. Diabetes Care
9. Degn KB, Juhl CB, Sturis J, Jakobsen G, of Helsinki: recommendations guiding 30:20322033, 2007
Brock B, Chandramouli V, Rungby J, physicians in biomedical research involv- 21. Vilsbll T, Brock B, Perrild H, Levin K, Le-
Landau BR, Schmitz O: One weeks treat- ing human subjects. JAMA 277:925926, rvang H-H, Klendorf K, Krarup T, Schmitz
ment with the long-acting glucagon-like 1997 O, Zdravkovic M, Le-Thi T, Madsbad S: Li-
peptide 1 derivative liraglutide (NN2211) 15. Matthews DR, Hosker JP, Rudenski AS, raglutide, a once-daily human GLP-1 ana-
markedly improves 24-h glycemia and - Naylor BA, Treacher DF, Turner RC: Ho- logue, improves pancreatic B-cell function
and -cell function and reduces endoge- meostasis model assessment: insulin re- and arginine-stimulated insulin secretion
nous glucose release in patients with type sistance and -cell function from fasting during hyperglycaemia in patients with
2 diabetes. Diabetes 53:11871194, 2004 plasma glucose and insulin concentra- type 2 diabetes mellitus. Diabet Med
10. Vilsbll T, Zdravkovic M, Le-Thi T, Krarup tions in man. Diabetologia 28:412 419, 25:152156, 2008
T, Schmitz O, Courreges J-P, Verhoeven R, 1985 22. Rolin B, Larsen MO, Gotfredsen CF, Dea-
Buganova I, Madsbad S: Liraglutide, a long- 16. DeFronzo RA, Ratner RE, Han J, Kim DD,
con CF, Carr RD, Wilken M, Knudsen LB:
acting human glucagon-like peptide-1 ana- Baron AD: Effects of exenatide (ex-
The long-acting GLP-1 derivative NN2211
log, given as monotherapy significantly endin-4) on glycemic control and weight
improves glycemic control and lowers body over 30 weeks in metformin-treated pa- ameliorates glycemia and increases -cell
weight without risk of hypoglycemia in pa- tients with type 2 diabetes. Diabetes Care mass in diabetic mice. Am J Physiol Endocri-
tients with type 2 diabetes. Diabetes Care 28:10921100, 2007 nol Metab 283:E745E752, 2002
30:1608 1610, 2007 17. Charbonnel B, Karasik A, Liu J, Wu M, 23. Bregenholt S, Moldrup A, Blume N,
11. Madsbad S, Schmitz O, Ranstam J, Jakob- Meininger G, Sitagliptin Study 020 Group: Karlsen AE, Friedrichsen BN, Tornhave
sen G, Matthews DR: Improved glycemic Efficacy and safety of the dipeptidyl pepti- D, Knudsen LB, Petersen JS: The long-
control with no weight increase in pa- dase-4 inhibitor sitagliptin added to ongo- acting glucagon-like peptide-1 analogue,
tients with type 2 diabetes after once-daily ing metformin therapy in patients with type liraglutide, inhibits -cell apoptosis in
treatment with the long-acting glucagon- 2 diabetes inadequately controlled with vitro. Biochem Biophys Res Commun 330:
like peptide 1 analog liraglutide (NN2211): metformin monotherapy. Diabetes Care 29: 577584, 2005

90 DIABETES CARE, VOLUME 32, NUMBER 1, JANUARY 2009