original article Diabetes, Obesity and Metabolism 17: 4251, 2015.

2014 John Wiley & Sons Ltd

Twice-daily dapagliflozin co-administered with metformin in type

2 diabetes: a 16-week randomized, placebo-controlled clinical
ORIGINAL
ARTICLE

trial
P.-M. Schumm-Draeger1 , L. Burgess2 , L. Kornyi3 , V. Hruba4 , J. E. Hamer-Maansson5 & T. W. A. de Bruin6
1 Clinic for Endocrinology, Diabetology, Angiology, Academic Teaching Hospital, Munich, Germany
2 TREAD Research, Cardiology Unit, Department of Internal Medicine, Tygerberg Hospital and Stellenbosch University, Parow, South Africa
3 Clinical Pharmacology Unit, Drug Research Center Gygyszervizsgl Kzpont, Balatonfred, Hungary
4 Global Medicines Department, AstraZeneca, Prague, Czech Republic
5 Biostatistics Department, AstraZeneca R&D, Wilmington, DE, USA
6 Global Medicines Department, AstraZeneca, Gaithersburg, MD, USA

Aims: To evaluate the efficacy and safety of twice-daily dosing of dapagliflozin and metformin, exploring the feasibility of a fixed-dose combination.
Methods: In this 16-week, phase III, randomized, double-blind placebo-controlled study, adults who were receiving metformin administered twice
daily (1500 mg/day) and had inadequate glycaemic control were randomized 1 : 1 : 1 : 1 to receive dapagliflozin twice daily (2.5 or 5 mg), placebo or
dapagliflozin 10 mg once daily (which was included as a benchmark). The primary endpoint was change from baseline glycated haemoglobin (HbA1c)
level. Secondary endpoints included changes in fasting plasma glucose (FPG) level and body weight.
Results: Four hundred adults were randomized to dapagliflozin (2.5 mg twice daily, 5 mg twice daily, 10 mg once daily) or placebo co-administered with
metformin twice daily. At 16 weeks, the adjusted mean change in HbA1c from baseline was significantly reduced in the dapagliflozin 2.5 mg twice daily
and 5 mg twice daily groups versus placebo (0.52 vs. 0.30%, p = 0.0106 and 0.65% vs. 0.30%, p < 0.0001). There were also significantly greater
improvements for dapagliflozin twice daily groups versus placebo in FPG body weight and achievement of HbA1c level of <7%. Efficacy outcomes for
dapagliflozin twice daily were numerically similar to those for dapagliflozin once daily. Dapagliflozin twice daily was well tolerated.
Conclusions: Dapagliflozin 2.5 or 5 mg twice daily added to metformin was effective in reducing glycaemic levels in patients with type 2 diabetes
inadequately controlled with metformin alone. This study supports the development of a fixed-dose combination regimen.
Keywords: metformin, phase III study, SGLT2 inhibitor, type 2 diabetes

Date submitted 15 May 2014; date of first decision 20 June 2014; date of final acceptance 30 August 2014

Introduction plasma glucose [9]. This approach reduces hyperglycaemia

Glycaemic control is the primary treatment goal for patients
with type 2 diabetes mellitus. Metformin, along with lifestyle
modification, is the recommended first-line therapy [1,2]. With
progressive deterioration of diabetes control it is frequently
necessary to add other antidiabetic agents on to existing therapy
[24]. Commonly prescribed add-on antidiabetic agents gener-
ally act by increasing insulin secretion or reducing insulin resis-
tance, but some regimens may cause weight gain, gastrointesti-
nal side effects, insulin resistance and hypoglycaemia [37]. An
add-on agent with a non-insulin-based mechanism of action
may complement metformin by providing improved glycaemic
control without increased risk of hypoglycaemia.
Sodium-glucose co-transporter-2 is the major sodium-
glucose co-transporter in the kidney [8]. Dapagliflozin
inhibits sodium-glucose co-transporter-2, resulting in direct
excretion of glucose in the urine and removal of excess
Correspondence to: P.-M. Schumm-Draeger, Klinikum Bogenhausen, Endocrinology,
Diabetology and Angiology Clinic, Englschalkinger Strasse 77, 81925 Munich, Germany.
E-mail: petra-maria.schumm-draeger@klinikum-muenchen.de
independently of insulin secretion or action. Dapagliflozin
has been shown to have antihyperglycaemic efficacy, both
alone and in combination with other commonly prescribed
antidiabetic agents, with optimum efficacy and safety being
observed at 10 mg/day [5,1014]. Results obtained in studies
of dapagliflozin co-administered with metformin have shown
significant reductions in glycated haemoglobin (HbA1c) and
fasting plasma glucose (FPG) [57,15,16]. A fixed-dose com-
bination of dapagliflozin and metformin is expected to have
complementary effects on glucose metabolism [5,17] and to
potentially support weight loss. It may also improve medication
compliance by simplifying treatment regimens [1822].
In phase III clinical trials, dapagliflozin has been evaluated
as a once-daily treatment, including in combination regimens
with metformin, dosed either twice daily or once daily as an
extended-release formulation [23]. Dapagliflozin inhibits renal
glucose reabsorption [5] and metformin decreases hepatic glu-
cose output [17]; therefore, based on the complementary action
on glucose metabolism in patients with type 2 diabetes mellitus,
it may be possible to use dapagliflozin in a fixed-dose combi-
nation with metformin immediate release, administered twice
DIABETES, OBESITY AND METABOLISM
daily. In the present study, we describe the results of a phase III
trial that was designed to determine whether twice-daily dosing
of dapagliflozin 2.5 or 5 mg had better efficacy than placebo as
an add-on to metformin in patients with type 2 diabetes with
inadequate glycaemic control on metformin alone. The safety
and tolerability of twice-daily dosing with dapagliflozin were
also assessed.
Materials and Methods
Trial Design
A 16-week, international, multicentre, randomized, double-
blind, double-dummy placebo-controlled phase III study was
undertaken at 53 sites in Europe and South Africa between
November 2010 and August 2011. This study complied with
the Declaration of Helsinki, International Conference on Har-
monisation/Good Clinical Practice and applicable regulatory
requirements. The study protocol was approved by the insti-
tutional review boards and independent ethics committees
of the participating institutions. All patients provided written
informed consent. The study is registered under clinical trial
registration number NCT01217892 (http://clinicaltrials.gov).
Patients were eligible for inclusion if they were 1877 years
old, had type 2 diabetes and had been treated with stable
doses of metformin 1500 mg/day monotherapy for 10 weeks
before enrolment and showed inadequate glycaemic control,
defined as an HbA1c level 6.7 and 10.5% at screening, or
HbA1c 6.5 and 10.0% 1 week before randomization. The
lower limit was chosen to ensure inclusion of patients with
a broad range of baseline HbA1c values, including those just
above the lower target value of 6.5% [24]. Exclusion crite-
ria included: symptoms of poorly controlled diabetes (includ-
ing weight loss, FPG >15.0 mmol/l); renal disorders, includ-
ing a calculated creatinine clearance (CrCl) <60 ml/min or
a measured serum creatinine value of 133 mol/l in men
and 124 mol/l in women; urine albumin : creatinine ratio
>1800 mg/g; hepatic disorders, including hepatotoxicity with
any medication; a recent cardiovascular event or New York
Heart Association class III or IV congestive heart failure; blood
pressure 160/100 mmHg at randomization; and clinically
significant haematological or oncological conditions.
Treatments
After initial screening and a 1-week enrolment period, there
was a 4-week lead-in period. During this period, patients con-
tinued their metformin treatment and received placebo twice
daily; metformin study medication was adjusted to 1500, 2000
or 2500 mg/day according to a prespecified adjustment table.
Diet and lifestyle advice was provided to enhance glycaemic
control; this was reinforced at each study visit.
Patients were stratified by HbA1c levels at randomization:
group 1 had HbA1c levels 6.5 and <7.0% (planned maximum
was no more than 20% of all participants); group 2 had HbA1c
levels 7.0 and 10.0%. This stratification was solely for the
purpose of limiting the proportion of patients with low baseline
HbA1c. Randomization was conducted via an Interactive Web
Volume 17 No. 1 January 2015
original article
Response System, which assigned a unique enrolment num-
ber according to a schedule with balanced blocks prepared by
AstraZeneca; this method was used to allocate patients to treat-
ment groups and maintain double-blind status.
Patients were randomized to one of four treatment groups:
dapagliflozin 2.5 mg twice daily, dapagliflozin 5 mg twice daily,
dapagliflozin 10 mg once daily and placebo twice daily, as
add-on to metformin. The dapagliflozin and metformin tablets
were taken at the same time for the twice-daily dosing regimens.
The twice-daily dapagliflozin doses were quantitatively equiva-
lent to the once-daily doses of 5 and 10 mg. The dapagliflozin
10 mg once daily treatment group was included as a mea-
sure of study sensitivity and a benchmark control; outcomes
for this dose were statistically tested versus placebo as a sec-
ondary objective, but were not formally compared with the
dapagliflozin twice-daily treatment groups.
Treatment-phase study visits took place at weeks 0, 1, 4, 8, 12
and 16. Patients were required to monitor and record their FPG
levels at least every other day. Patients who had poor glycaemic
control according to study-specific criteria (FPG >15.0 mmol/l
during first 4 weeks of treatment period, FPG >13.2 mmol/l
from weeks 4 to 8, FPG >11.1 mmol/l from weeks 12 to 16) were
discontinued from the study; an increase in metformin dosing
was not allowed. Concomitant, non-antidiabetic medications
(e.g. antihypertensive, diuretic and lipid-lowering drugs) were
allowed, but changes in dosing were not allowed during the trial
unless medically necessary. Upon completion of active study
treatment, patients were monitored for 4 weeks to evaluate
changes in physical signs, symptoms or laboratory variables that
might be related to treatment with dapagliflozin. The last study
visit followed this period.
Objectives and Assessments
The primary efficacy objective was to compare the change from
baseline in HbA1c achieved with dapagliflozin 2.5 mg twice
daily and 5 mg twice daily co-administered with metformin
versus placebo co-administered with metformin after 16 weeks
of treatment. Key secondary endpoints included the percent
change from baseline in body weight, change from baseline
in FPG at weeks 1 and 16 and proportion of patients with
HbA1c levels 7% at baseline achieving HbA1c levels <7%
after 16 weeks. Other secondary endpoints included change in
seated blood pressure and body weight. Secondary endpoints
for the benchmark treatment group with dapagliflozin 10 mg
once daily included identical objectives and assessments.
The safety and tolerability of dapagliflozin co-administered
with metformin were determined by assessment of adverse
events (AEs), including hypoglycaemic events, laboratory val-
ues, vital signs (electrocardiogram, blood pressure), calculated
CrCl, estimated glomerular filtration rate (eGFR) and physical
examination findings.
Statistical Analysis
All randomized patients who received at least one dose of
double-blind study medication, with a non-missing baseline
value and at least one post-baseline value for an efficacy vari-
able, were included in the full analysis set. Missing values at
doi:10.1111/dom.12387 43
original article
week 16 were to be replaced by last observation carried for-
ward (LOCF). The study had a power of 90% to detect a 0.5%
difference in mean change from baseline in HbA1c using a
two-sample t-test at a 0.025 two-sided significance level (in the
conservative case where only one treatment group reached sta-
tistical significance). The Hochberg procedure [25] was used
to control the overall type I error rate 0.050 for the com-
parisons of the two dapagliflozin twice daily groups versus
placebo + metformin for the primary efficacy variable. If the
test of the primary endpoint was significant for a treatment
group, statistical significance of secondary outcomes proceeded
for the same treatment group and was evaluated using a hier-
archical, fixed-sequence testing procedure (per order of end-
points listed above) for the treatment groups, with demon-
strated statistical significance for the primary outcome; this
was to protect the global type I error rate for each treatment
group. The primary efficacy variable and other continuous effi-
cacy variables were analysed using an analysis of covariance
(ancova) model with treatment group as a fixed effect and
baseline HbA1c as a covariate. Proportions were analysed using
the logistic regression methodology described by Zhang et al.
[26] and Tsiatis et al. [27], with adjustment for baseline value.
Sensitivity analyses included a completers analysis and one that
excluded outliers.
The safety analysis set included all patients who received
one or more doses of double-blind study medication. Safety
variables were to be summarized descriptively.
Results
Patients
A total of 520 patients were enrolled in the study (Figure 1); 400
were randomly assigned to one of four treatments as add-on
to metformin: placebo (n = 101), dapagliflozin 2.5 mg twice
daily (n = 100), dapagliflozin 5 mg twice daily (n = 100) or
dapagliflozin 10 mg once daily (n = 99). Baseline demographics
and other characteristics were generally similar across treat-
ment groups (Table 1). The median background dose of met-
formin at enrolment and at randomization was 2000 mg/day
across treatment groups. Completion rates were similar across
treatment groups [between 91% (for dapagliflozin 10 mg once
daily) and 94% (for dapagliflozin 5 mg twice daily)]. The most
frequent reasons for discontinuation were study-specific dis-
continuation criteria (n = 19), the most commonly reported
being decreased CrCl (n = 12; distributed evenly across the
treatment groups), and the other being inadequate glycaemic
control (n = 7; five from the placebo group).
Efficacy
The adjusted mean reduction in HbA1c level was signifi-
cantly greater for the dapagliflozin 2.5 and 5 mg twice daily
treatment groups than placebo: 0.52% (p = 0.0106), 0.65%
(p < 0.0001) and 0.30%, respectively (Table 2, Figure 2A).
Sensitivity analyses were similar to the main analysis; the
adjusted mean reduction in HbA1c level was within 0.03%
of each main result (data not shown). The adjusted mean
44 Schumm-Draeger et al.
DIABETES, OBESITY AND METABOLISM
reduction from baseline for dapagliflozin 10 mg once daily was
0.59% [95% confidence interval (CI) 0.70, 0.47].
There were significant differences in adjusted mean reduc-
tion from baseline in FPG at week 1 and week 16 for both
dapagliflozin 2.5 and 5 mg twice daily treatment groups ver-
sus placebo (p < 0.0001; Table 2, Figure 2B). Among patients
with HbA1c levels 7% at baseline, significantly more patients
in the dapagliflozin 2.5 and 5 mg twice daily treatment groups
achieved an HbA1c level <7% at week 16 than patients in the
placebo group (Table 2).
The adjusted mean percent change in body weight was
also significantly greater in the dapagliflozin 2.5 and 5 mg
twice daily treatment groups than placebo: 2.84, 3.20
and 1.04%, respectively (both p < 0.0001 vs. placebo). For
dapagliflozin 10 mg once daily, the adjusted mean percent
change in body weight was 2.76% (95% CI 3.36, 2.15).
The greatest decreases were seen in the first week of treat-
ment. Placebo-subtracted changes in total body weight for the
dapagliflozin 2.5 and 5 mg twice daily treatment groups were
1.62 and 1.88 kg, respectively (Figure 2C).
Compared with placebo, all dapagliflozin groups showed a
reduction in seated systolic blood pressure, but not in diastolic
blood pressure (Table S1, Supporting Information).
Safety
A summary of AEs reported during the present trial is shown in
Table 3. The proportions of patients with AEs were low and sim-
ilar across the dapagliflozin 2.5 mg twice daily and 5 mg twice
daily treatment groups (3340.0%); the proportion was slightly
higher in the dapagliflozin 10 mg once daily group (46.5%). The
rate of discontinuations attributable to AEs was also low and
similar across all treatment groups. Serious AEs (SAEs) were
rare and slightly more frequent in the dapagliflozin 2.5 and 5 mg
twice daily treatment groups (4 and 1%, respectively) than in
the placebo group (0%). No SAEs were treatment-related; SAEs
were evenly distributed across several system organ classes and
none led to discontinuation. No deaths or malignancies were
reported during the treatment period.
In total, only three hypoglycaemic events were reported
in dapagliflozin-treated patients. None of these events
was considered major [i.e. a symptomatic event requiring
third-party assistance with a capillary or plasma glucose value
(<3.0 mmol/l (<54 mg/dl))], and none led to discontinuation.
Other events of special interest generally occurred at low rates
in all patient groups. Proportions of patients reporting events
suggestive of urinary tract infection were low and similar
across treatment groups (5%), with all events considered
mild to moderate. There were no reports of kidney infection.
Events suggestive of genital infections were infrequent (5%)
and reported only in female patients; events were mild or
moderate in intensity. Reports of other AEs of special interest
were absent (e.g. hypotension, dehydration or hypovolaemia,
renal impairment, renal failure, fractures or kidney infection)
or indicated no increased risk associated with dapagliflozin
(AEs of renal impairment, renal failure or urinary stones).
Gastrointestinal AEs, which are associated with metformin
treatment, were similar across treatment groups and did not
increase with the addition of dapagliflozin.
Volume 17 No. 1 January 2015
DIABETES, OBESITY AND METABOLISM
Figure 1. Patient disposition. Incorrect enrolment included subjects not meeting inclusion criteria or developing an exclusion criterion after being
correctly enrolled. No longer met study criteria refers to protocol-specified discontinuation criteria (e.g. inadequate glycaemic control, decreased creatinine
clearance). BID, twice daily; MET-IR, metformin immediate release.
Laboratory evaluations are shown in Table S1. Dapagli-
flozin-treated patients showed a large mean increase in urine
glucose excretion as measured by spot urine tests, as expected
from the mechanism of action (glucuresis); this effect was evi-
dent at week 1. A mean increase in blood urea nitrogen was
observed from baseline to week 16 in dapagliflozin-treated
patients (0.180.64 mmol/l). Follow-up monitoring indicated
that these changes were reversible. No meaningful changes
in blood urea nitrogen were observed in the placebo group.
Patients in the dapagliflozin 2.5 and 5 mg twice daily groups
showed a slight mean decrease in eGFR from baseline to week
8 that remained stable until week 16 (change from baseline of
2.1 and 3.5 ml/min/1.73 m2 at week 16). In the dapagliflozin
10 mg once daily group, eGFR showed a slight recovery at
weeks 12 and 16 from the initial decrease. No hepatic impair-
ment or increase in liver enzymes was evident in any treatment
group. Marked abnormalities of electrolytes were rare. Three
patients receiving dapagliflozin 5 mg twice daily and one receiv-
ing placebo showed one or more marked abnormalities of crea-
tine kinase >5 the upper limit of normal. Two patients in the
dapagliflozin 5 mg twice daily group showed marked abnormal-
ities in creatine kinase >10 the upper limit of normal; both
patients reported excessive physical work or exercise.
Discussion
The present study was designed to determine whether
twice-daily dosing of dapagliflozin co-administered with
metformin showed better efficacy compared with placebo
co-administered with metformin in patients with type 2 dia-
betes with inadequate glycaemic control on metformin alone,
Volume 17 No. 1 January 2015
original article
and to evaluate the safety and tolerability of these treatment
regimens. Dapagliflozin 2.5 and 5 mg twice-daily doses, which
are equivalent to the daily doses of 5 and 10 mg, respectively,
co-administered with metformin, resulted in statistically
significant and clinically meaningful glycaemic reductions
compared with placebo, measured as reductions in HbA1c
levels and FPG levels after 1 week and 16 weeks, and a greater
proportion of patients achieving an HbA1c target of <7.0%.
Reductions in FPG levels at week 16 in both dapagliflozin
twice daily groups approached the recommended therapeutic
targets of 3.97.2 mmol/l [1,2]. Overall, the FPG outcomes for
dapagliflozin 2.5 and 5 mg twice daily were consistent with the
dapagliflozin 10 mg once daily treatment arm, and were similar
to those reported previously for dapagliflozin 10 mg once daily
in combination with metformin [5,15,16]. Statistically signifi-
cant reductions in body weight reported in both dapagliflozin
twice daily treatment groups were consistent with the weight
reduction reported with dapagliflozin 10 mg once daily plus
metformin [5,16].
Marked increases in urine glucose excretion levels were
measured in all dapagliflozin treatment groups, consistent
with dapagliflozins mechanism of action and providing
an explanation for the observed efficacy and weight loss.
Co-administration of dapagliflozin and metformin is expected
to exert a complementary therapeutic effect on hyperglycaemia
in type 2 diabetes because dapagliflozin inhibits renal glucose
reabsorption and metformin decreases hepatic glucose output,
which explains the glycaemic efficacy observed. Previously,
dapagliflozin 10 mg once daily co-administered with met-
formin in patients with a similar HbA1c range was found to
doi:10.1111/dom.12387 45
original article DIABETES, OBESITY AND METABOLISM

Table 1. Demographics and baseline characteristics of randomized patients.

Placebo + Dapagliflozin 2.5 mg Dapagliflozin 5 mg Dapagliflozin 10 mg

metformin, twice daily + metformin, twice daily + metformin, once daily + metformin,
n = 101 n = 100 n = 99 n = 99
Mean (s.d.) age, years 58.5 (9.4) 58.3 (9.0) 55.3 (9.3) 58.5 (9.8)
Age category, n (%)
<65 years 77 (76.2) 77 (77.0) 85 (85.9) 70 (70.7)
65 and <75 years 23 (22.8) 21 (21.0) 13 (13.1) 27 (27.3)
75 years 1 (1.0) 2 (2.0) 1 (1.0) 2 (2.0)
Sex, n (%)
Male 47 (46.5) 37 (37.0) 46 (46.5) 49 (49.5)
Female 54 (53.5) 63 (63.0) 53 (53.5) 50 (50.5)
Race, n (%)
White 82 (81.2) 79 (79.0) 84 (84.8) 81 (81.8)
Black 5 (5.0) 10 (10.0) 5 (5.1) 6 (6.1)
Asian 10 (9.9) 8 (8.0) 7 (7.1) 3 (3.0)
Other* 4 (4.0) 3 (3.0) 3 (3.0) 9 (9.1)
Mean (s.d.) body mass index (kg/m2 ) 31.74 (4.69) 33.16 (5.16) 33.09 (4.94) 32.25 (5.01)
Mean (s.d.) seated blood pressure (mmHg)
Systolic 133.4 (11.9) 132.4 (13.3) 130.3 (11.4) 132.2 (12.0)
Diastolic 81.5 (6.7) 80.5 (7.4) 81.3 (6.7) 79.4 (7.7)
Mean (s.d.) duration of type 2 diabetes (years) 5.53 (4.23) 4.80 (3.87) 5.12 (4.20) 5.45 (4.05)
Mean (s.d.) HbA1c, (%) 7.94 (0.85) 7.77 (0.75) 7.78 (0.76) 7.71 (0.71)
HbA1c category (%), n (%)
<7.0 13 (12.9) 10 (10.0) 7 (7.1) 17 (17.2)
7.0 and <8.0 42 (41.6) 58 (58.0) 62 (62.6) 51 (51.5)
8.0 and <9.0 27 (26.7) 25 (25.0) 22 (22.2) 25 (25.3)
9.0 19 (18.8) 7 (7.0) 8 (8.1) 6 (6.1)
Mean (s.d.) FPG (mmol/l) 8.76 (1.99) 8.51 (1.85) 8.62 (1.77) 8.62 (2.01)
eGFR, n (%)
<30 ml/min/1.73 m2 0 0 0 0
30 and <60 ml/min/1.73 m2 4 (4.0) 4 (4.0) 5 (5.1) 5 (5.1)
60 and <90 ml/min/1.73 m2 61 (60.4) 57 (57.0) 55 (55.6) 65 (65.7)
<90 ml/min/1.73 m2 36 (35.6) 39 (39.0) 39 (39.4) 29 (29.3)

eGFR was calculated using the Modification in Diet and Renal Disease formula and is adjusted for body surface area. eGFR, estimated glomerular filtration
rate; HbA1c, glycated haemoglobin; s.d., standard deviation.
*Includes Native Hawaiian or other Pacific Islander, American Indian/Alaska Native or other.
All patients had creatinine clearance 60 ml/min at randomization.

result in improvements in glycaemic control and reductions in
body weight, which were maintained over 102 weeks [15].
Compared with other dapagliflozin studies [5,10,13], there
was a lower magnitude of placebo-corrected HbA1c reduction
in the present study. The most likely reason for this is that
because dapagliflozins efficacy is dependent on renal filtration
of glucose, the relatively low baseline HbA1c of the patients
in the present study may have affected this result. Baseline
HbA1c has been strongly associated with reduction in HbA1c
in studies of oral agents, including dapagliflozin, for the treat-
ment of type 2 diabetes [28,29]. Moreover, the HbA1c reduc-
tion observed in the placebo group, perhaps attributable to the
reinforcement of diet and lifestyle guidelines, tends to dimin-
ish over time, whereas the benefit from dapagliflozin does not
[14,15]. Nonetheless, compared with placebo, a significantly
higher percentage of patients in both dapagliflozin twice daily
groups achieved an HbA1c concentration of <7%, which is the
recommended target for HbA1c. Other possible contributing
factors to the smaller difference in HbA1c reduction between
dapagliflozin- and placebo-treated patients are the relatively
brief lead-in period of 4 weeks, which may have limited opti-
mum stabilization of the background metformin treatment and
delayed capture of HbA1c response, and natural study-to-study
variation in outcomes [30,31].
Dapagliflozin 2.5 and 5 mg twice-daily doses were both
well tolerated, and the incidence of AEs was low. Although
the relatively short trial duration may limit the extrapolation
of safety assessments to long-term studies, the reported out-
comes were in keeping with those documented throughout
the dapagliflozin clinical development programme, especially
long-term studies of >2 years of dapagliflozin co-administered
with metformin. Markedly fewer AEs were reported with
dapagliflozin 5 mg twice daily than with dapagliflozin 10 mg
once daily (33.0 and 46.5%, respectively). Discontinuation
rates were low across treatment groups. Despite glycaemic
reductions in a type 2 diabetes mellitus patient population with
a relatively low HbA1c at baseline, hypoglycaemic events were
rare, and none were major. SAEs were also rare, with no specific
pattern or trend being evident. In the present study, we closely
monitored gastrointestinal events, as metformin is associated

46 Schumm-Draeger et al. Volume 17 No. 1 January 2015

DIABETES, OBESITY AND METABOLISM original article
Table 2. Primary and secondary efficacy results at week 16 (last observation carried forward).

Dapagliflozin 2.5 mg Dapagliflozin 5 mg Dapagliflozin 10 mg

Placebo + twice daily + twice daily + once daily +
metformin, metformin, metformin, metformin,
n = 101 n = 100 n = 99 n = 99
HbA1c (%)
n 100 99 97 98
Mean (s.d.) baseline 7.94 (0.852) 7.78 (0.748) 7.79 (0.764) 7.71 (0.713)
Mean (s.d.) week 16 7.59 (0.895) 7.27 (0.777) 7.15 (0.704) 7.16 (0.625)
Mean (s.d.) change from baseline 0.35 (0.679) 0.50 (0.663) 0.65 (0.736) 0.55 (0.546)
Mean (s.e.) adjusted mean change from baseline 0.30 (0.0593) 0.52 (0.0594) 0.65 (0.0600) 0.59 (0.0598)
95% CI for adjusted mean change from baseline 0.42, 0.18 0.63, 0.40 0.77, 0.53 0.70, 0.47
Mean difference versus placebo + MET-IR*(s.e.) 0.22 (0.0840) 0.35 (0.0843) 0.29 (0.0844)
95% CI for mean difference versus placebo + MET-IR 0.38, 0.05 0.52, 0.18 0.45, 0.12
p value versus placebo + MET-IR 0.0106 <0.0001 0.0007
Body weight (kg)
n 101 100 99 99
Mean (s.d.) baseline 88.82 (15.327) 92.49 (18.632) 93.62 (16.641) 90.58 (15.929)
Mean (s.d.) week 16 87.89 (15.474) 89.95 (18.661) 90.82 (17.070) 88.17 (16.195)
Mean (s.e.) % change from baseline 1.12 (0.2508) 2.83 (0.2339) 3.16 (0.3009) 2.79 (0.2422)
Adjusted mean (s.e.) % change from baseline 1.04 (0.3105) 2.84 (0.3099) 3.20 (0.3125) 2.76 (0.3086)
95% CI for adjusted mean % change from baseline 1.65, 0.43 3.45, 2.23 3.82, 2.59 3.36, 2.15
Mean (s.d.) difference versus placebo + MET-IR* 1.82 (0.3630) 2.18 (0.3636) 1.73 (0.3635)
95% CI for mean difference versus placebo + MET-IR 2.53, 1.10 2.89, 1.46 2.44, 1.01
p value versus placebo + MET-IR <0.0001 <0.0001 <0.0001
FPG week 1 (mmol/l)
n 101 99 99 98
Mean (s.d.) baseline 8.76 (1.99) 8.52 (1.85) 8.62 (1.77) 8.62 (2.02)
Mean (s.d.) week 1 8.97 (2.08) 7.96 (1.56) 7.96 (1.47) 7.92 (1.63)
Mean (s.d.) change from baseline 0.21 (1.36) 0.57 (1.21) 0.65 (1.47) 0.70 (1.47)
Adjusted mean (s.e.) change from baseline 0.11 (0.14) 0.76 (0.15) 0.82 (0.15) 0.86 (0.15)
95% CI for adjusted mean change from baseline 0.17, 0.39 1.05, 0.47 1.11, 0.53 1.15, 0.57
Mean difference versus placebo + MET-IR* 0.87 (0.17) 0.93 (0.17) 0.97 (0.17)
95% CI for mean difference versus placebo + MET-IR 1.20, 0.54 1.26, 0.59 1.30, 0.64
p value versus placebo + MET-IR <0.0001 <0.0001 <0.0001
FPG week 16 (mmol/l)
n 101 100 99 98
Mean (s.d.) baseline 8.76 (1.99) 8.51 (1.85) 8.62 (1.77) 8.62 (2.02)
Mean (s.d.) week 16, LOCF 8.20 (1.86) 7.50 (1.52) 7.29 (1.25) 7.58 (1.55)
Mean (s.d.) change from baseline 0.56 (1.71) 1.00 (1.67) 1.33 (1.46) 1.04 (1.32)
Adjusted mean (s.e.) change from baseline 0.58 (0.15) 1.15 (0.15) 1.42 (0.15) 1.13 (0.15)
95% CI for adjusted mean change from baseline 0.87, 0.28 1.45, 0.85 1.73, 1.12 1.43, 0.84
Mean difference versus placebo + MET-IR* 0.58 (0.17) 0.85 (0.17) 0.56 (0.17)
95% CI for mean difference versus placebo + MET-IR 0.92, 0.23 1.19, 0.51 0.90, 0.22
p value versus placebo + MET-IR 0.0010 <0.0001 0.0015
Subjects achieving HbA1c <7% at week 16
n/N 17/87 32/89 37/90 22/81
% 19.5 36.0 41.1 27.2
% adjusted (s.e.) 21.4% (4.170) 33.6% (4.567) 38.2% (4.651) 28.1% (4.619)
95% CI for % adjusted 13.2, 29.6 24.6, 42.5 29.1, 47.3 19.0, 37.1
Mean (s.e.) difference versus placebo + MET-IR (%) 12.2% (6.097) 16.8% (6.153) 6.7% (6.145)
95% CI for mean difference versus placebo + MET-IR 0.2, 24.1 4.8, 28.9 5.3, 18.7
p value versus placebo + MET-IR 0.0455 0.0062 0.2755

CI, confidence interval; FPG, fasting plasma glucose; HbA1c, glycated haemoglobin; LOCF, last observation carried forward; MET-IR, metformin
immediate release; s.d., standard deviation; s.e., standard error.
*Difference in adjusted mean change from baseline versus placebo + MET-IR.
Population with HbA1c >7% at baseline.
Nominal p values; formal statistical comparisons versus placebo were not conducted.

Volume 17 No. 1 January 2015 doi:10.1111/dom.12387 47

original article DIABETES, OBESITY AND METABOLISM

Figure 2. Adjusted mean changes over 16 weeks in: (A) glycated haemoglobin (HbA1c) %; (B) fasting plasma glucose (FPG), mmol/l; and (C) total body
weight (kg). Data are adjusted mean changes from baseline over time (last observation carried forward) for the 16-week double-blind treatment period
(full analysis set). Mean value based on analysis of covariance model with treatment group and stratum as effects and baseline value as covariate. Error
bars represent 95% confidence intervals (CIs) for the adjusted mean change from baseline. Treatment symbols shifted horizontally to prevent error bars
overlapping. BID, twice daily; DAPA, dapagliflozin; HbA1c, glycated haemoglobin; MET, metformin; PBO, placebo; QD, once daily.

48 Schumm-Draeger et al. Volume 17 No. 1 January 2015

DIABETES, OBESITY AND METABOLISM original article
Table 3. Number and percentage of patients with adverse events over the 16-week treatment period (safety analysis set).

Placebo + Dapagliflozin 2.5 mg Dapagliflozin 5 mg Dapagliflozin 10 mg

metformin, twice daily + metformin, twice daily + metformin, once daily + metformin,
n = 101, n = 100, n = 100, n = 99,
n (%) n (%) n (%) n (%)
1 AE 37 (36.6) 40 (40.0) 33 (33.0) 46 (46.5)
1 treatment-related AE 10 (9.9) 5 (5.0) 7 (7.0) 8 (8.1)
1 AE leading to discontinuation 3 (3.0) 5 (5.0) 3 (3.0) 4 (4.0)
1 SAE 0 4 (4.0) 1 (1.0) 2 (2.0)
1 treatment-related SAE 0 0 0 0
1 SAE leading to discontinuation 0 0 0 0
Deaths 0 0 0 0
AEs occurring in 5% of patients in any group
Influenza 3 (3.0) 5 (5.0) 0 2 (2.0)
Hypertension 6 (5.9) 3 (3.0) 0 4 (4.0)
AEs of special interest
Hypoglycaemia 0 1 (1.0) 0 2 (2.0)
Hypotension/dehydration/hypovolaemia 0 0 0 0
Renal impairment/failure 4 (4.0) 5 (5.0) 3 (3.0) 3 (3.0)
Fractures 0 0 0 0
Urinary stones 1 (1.0) 0 0 1 (1.0)
Events suggestive of urinary tract infection 3 (3.0) 2 (2.0) 5 (5.0) 3 (3.0)
Events suggestive of genital infection 1 (1.0) 0 5 (5.0) 3 (3.0)
Kidney infection 0 0 0 0

Includes non-serious AEs with onset on or after the first date of double-blind treatment and on or before the last day of double-blind treatment plus 4 days
or up to follow-up visit if earlier. Includes SAEs with onset on or after the first date of double-blind treatment and on or before the last day of double-blind
treatment plus 30 days or up to follow-up visit if earlier. Events reported based on preferred terms from Medical Dictionary for Regulatory Activities, version
14.0. AE terms of renal impairment/failure were prespecified in the protocol. Hypoglycaemia was defined as a low blood glucose reading [<3.5 mmol/l
(<63 mg/dl)], with or without symptoms; a symptomatic event without a blood glucose reading was considered suggestive of hypoglycaemia. AE, adverse
event; SAE, serious adverse event.

with gastrointestinal AEs; there was no increase reported with
the co-administration of dapagliflozin 2.5 or 5 mg twice daily
with metformin.
The decrease in CrCl reported in the dapagliflozin treatment
groups may be a reflection of weight as a factor in the calcu-
lation. A treatment that produces weight loss would result in
a reduction in calculated CrCl. It has been shown previously
that dapagliflozin is associated with a small decrease in eGFR
at week 1 that returns to baseline by week 24 and is then main-
tained during further treatment [32]. Similarly, in the present
study, a small initial decrease in eGFR with dapagliflozin was
observed for all treatment groups and this did not appear to
affect renal function. Follow-up data beyond week 16 were not
available for eGFR in the present study, although follow-up data
for calculated CrCl indicated that the observed initial decreases
were reversible. Genital infections and urinary tract infections
would be expected to increase with dapagliflozin treatment
because of the resultant glucosuria; however, the incidences of
each were low overall and did not exceed incidences in other
dapagliflozin studies. No events of kidney infection or upper
urinary tract infection were reported, and there were no reports
of renal failure or injury.
The present population of patients with type 2 diabetes melli-
tus treated with metformin monotherapy represents an impor-
tant target group for the evaluation of combination therapies in
light of the frequency of suboptimum or inadequate glycaemic
control. The greater proportion of patients in the present trial
treated with dapagliflozin and metformin who achieved HbA1c
<7%, a goal recently emphasized in treatment guidelines [1],
shows that this combination therapy can be of clinical value.
In summary, the data from the present study provide evi-
dence that twice-daily co-administration of dapagliflozin and
metformin is effective in reducing HbA1c, FPG and body
weight to levels numerically similar to those obtained with
dapagliflozin once daily added to metformin. The improvement
in glycaemic control observed with dapagliflozin twice daily
co-administered with metformin shows that the combination
can offer better outcomes to patients without concern for hypo-
glycaemic events.
Acknowledgements
The present study was funded by Bristol-Myers Squibb and
AstraZeneca. Medical writing assistance was provided by Ann
L. Davis, MPH, CMPP, an employee of Bristol-Myers Squibb,
and Caroline Allinson, BSc, of PPSI (a PAREXEL company).
Medical writing assistance was funded by Bristol-Myers Squibb
and AstraZeneca.
Conflict of Interest
P.-M. S.-D. has received funds for speaking from Sanofi, Novo-
Nordisk, Eli Lilly, Novartis, Bristol-Myers Squibb, AstraZeneca,
MSD and Boehringer Ingelheim. She has also received sup-
port for advisory boards and clinical trials from Sanofi,
NovoNordisk, Bristol-Myers Squibb, AstraZeneca, Boehringer

Volume 17 No. 1 January 2015 doi:10.1111/dom.12387 49

original article
Ingelheim, Novartis, GlaxoSmithKline and Eli Lilly. L. B. and
L. K. do not have any conflicts of interest to declare. T. W.
A. d. B. is an employee of AstraZeneca. J. E. H.-M. and V.
H. were employees of AstraZeneca at the time the study was
conducted.
T. W. A. d. B., J. E. H.-M. and V. H. participated in the study
concept and design. P.-M. S.-D., L. B. and L. K. participated in
acquisition of data. All authors participated in the analysis and
interpretation of data. J. E. H.-M. participated in the statistical
verification of data. V. H. participated in study supervision.
All authors provided critical review of the manuscript and
approved the final version for submission.
Supporting Information
Additional Supporting Information may be found in the online
version of this article:
Table S1. Changes from baseline in laboratory measurements
and blood pressure at week 16 (safety analysis set).
References
1. Inzucchi SE, Bergenstal RM, Buse JB et al. Management of hyperglycemia
in type 2 diabetes: a patient-centered approach: position statement of the
American Diabetes Association (ADA) and the European Association for the
Study of Diabetes (EASD). Diabetes Care 2012; 35: 13641379.
2. American Diabetes Association. Standards of medical care in diabetes 2013.
Diabetes Care 2013; 36(Suppl. 1): S11S66.
3. Turner R, Cull C, Holman R. United Kingdom Prospective Diabetes Study 17:
a 9-year update of a randomized, controlled trial on the effect of improved
metabolic control on complications in non-insulin-dependent diabetes melli-
tus. Ann Intern Med 1996; 124: 136145.
4. Stratton IM, Adler AI, Neil HA et al. Association of glycaemia with macrovascular
and microvascular complications of type 2 diabetes (UKPDS 35): prospective
observational study. BMJ 2000; 321: 405412.
5. Bailey CJ, Gross JL, Pieters A, Bastien A, List JF. Effect of dapagliflozin in
patients with type 2 diabetes who have inadequate glycaemic control with
metformin: a randomised, double-blind, placebo-controlled trial. Lancet 2010;
375: 22232233.
6. Henry RR, Murray AV, Marmolejo MH, Hennicken D, Ptaszynska A, List JF.
Dapagliflozin, metformin XR, or both: initial pharmacotherapy for type 2
diabetes, a randomised controlled trial. Int J Clin Pract 2012; 66: 446456.
7. Nauck MA, Del PS, Meier JJ et al. Dapagliflozin versus glipizide as add-on ther-
apy in patients with type 2 diabetes who have inadequate glycemic control with
metformin: a randomized, 52-week, double-blind, active-controlled noninferi-
ority trial. Diabetes Care 2011; 34: 20152022.
8. Rahmoune H, Thompson PW, Ward JM, Smith CD, Hong G, Brown J. Glucose
transporters in human renal proximal tubular cells isolated from the urine of
patients with non-insulin-dependent diabetes. Diabetes 2005; 54: 34273434.
9. Komoroski B, Vachharajani N, Feng Y, Li L, Kornhauser D, Pfister M. Dapagliflozin,
a novel, selective SGLT2 inhibitor, improved glycemic control over 2 weeks in
patients with type 2 diabetes mellitus. Clin Pharmacol Ther 2009; 85: 513519.
10. Ferrannini E, Ramos SJ, Salsali A, Tang W, List JF. Dapagliflozin monotherapy in
type 2 diabetic patients with inadequate glycemic control by diet and exercise:
a randomized, double-blind, placebo-controlled, phase 3 trial. Diabetes Care
2010; 33: 22172224.
11. List JF, Woo V, Morales E, Tang W, Fiedorek FT. Sodium-glucose cotransport
inhibition with dapagliflozin in type 2 diabetes. Diabetes Care 2009; 32:
650657.
50 Schumm-Draeger et al.
DIABETES, OBESITY AND METABOLISM
12. Rosenstock J, Vico M, Wei L, Salsali A, List JF. Effects of dapagliflozin, an SGLT2
inhibitor, on HbA(1c), body weight, and hypoglycemia risk in patients with type
2 diabetes inadequately controlled on pioglitazone monotherapy. Diabetes Care
2012; 35: 14731478.
13. Strojek K, Yoon K, Hruba V, Elze M, Langkilde A, Parikh S. Effect of dapagliflozin
in patients with type 2 diabetes who have inadequate glycaemic control with
glimepiride: a randomised, 24-week, double-blind, placebo-controlled trial.
Diabetes Obes Metab 2011; 13: 928938.
14. Wilding JP, Woo V, Soler NG et al. Long-term efficacy of dapagliflozin in patients
with type 2 diabetes mellitus receiving high doses of insulin: a randomized trial.
Ann Intern Med 2012; 156: 405415.
15. Bailey CJ, Gross JL, Hennicken D, Iqbal N, Mansfield TA, List JF. Dapagliflozin
add-on to metformin in type 2 diabetes inadequately controlled with met-
formin: a randomized, double-blind, placebo-controlled 102-week trial. BMC
Med 2013; 11: 43.
16. Bolinder J, Ljunggren O, Kullberg J et al. Effects of dapagliflozin on body weight,
total fat mass, and regional adipose tissue distribution in patients with type
2 diabetes mellitus with inadequate glycemic control on metformin. J Clin
Endocrinol Metab 2012; 97: 10201031.
17. Garber AJ, Duncan TG, Goodman AM, Mills DJ, Rohlf JL. Efficacy of metformin
in type II diabetes: results of a double-blind, placebo-controlled, dose-response
trial. Am J Med 1997; 103: 491497.
18. Dickson M, Plauschinat CA. Compliance with antihypertensive therapy in the
elderly: a comparison of fixed-dose combination amlodipine/benazepril versus
component-based free-combination therapy. Am J Cardiovasc Drugs 2008; 8:
4550.
19. Claxton AJ, Cramer J, Pierce C. A systematic review of the associations
between dose regimens and medication compliance. Clin Ther 2001; 23:
12961310.
20. Bangalore S, Kamalakkannan G, Parkar S, Messerli FH. Fixed-dose combina-
tions improve medication compliance: a meta-analysis. Am J Med 2007; 120:
713719.
21. Melikian C, White TJ, Vanderplas A, Dezii CM, Chang E. Adherence to oral antidi-
abetic therapy in a managed care organization: a comparison of monotherapy,
combination therapy, and fixed-dose combination therapy. Clin Ther 2002; 24:
460467.
22. Ho PM, Rumsfeld JS, Masoudi FA et al. Effect of medication nonadherence
on hospitalization and mortality among patients with diabetes mellitus. Arch
Intern Med 2006; 166: 18361841.
23. Glucophage [prescribing information]. Geneva, Switzerland: Merck Serono.
Available from URL: http://www.medicines.org.uk/emc/medicine/1043/SPC/
Glucophage+500+mg+and+850+mg+film+coated+tablets/, 2010. Accessed
18 August 2014.
24. Nathan DM, Buse JB, Davidson MB et al. Medical management of hyperglycemia
in type 2 diabetes: a consensus algorithm for the initiation and adjustment
of therapy: a consensus statement of the American Diabetes Association and
the European Association for the Study of Diabetes. Diabetes Care 2009; 32:
193203.
25. Hochberg Y. A Sharper Bonferroni procedure for multiple tests of significance.
Biometrika 1988; 75: 800802.
26. Zhang M, Tsiatis AA, Davidian M. Improving efficiency of inferences in
randomized clinical trials using auxiliary covariates. Biometrics 2008; 64:
707715.
27. Tsiatis AA, Davidian M, Zhang M, Lu X. Covariate adjustment for two-sample
treatment comparisons in randomized clinical trials: a principled yet flexible
approach. Stat Med 2008; 27: 46584677.
28. Esposito K, Chiodini P, Capuano A, Maiorino MI, Bellastella G, Giugliano D.
Baseline glycemic parameters predict the hemoglobin A response to DPP-4
inhibitors: meta-regression analysis of 78 randomized controlled trials with
20,053 patients. Endocrine 2014; 46: 4351.
Volume 17 No. 1 January 2015
DIABETES, OBESITY AND METABOLISM
29. Hardy E, Salsali A, Hruba V et al. Efficacy increases with increasing baseline
HbA1c category with dapagliflozin therapy. Diabetes 2012; 61(Suppl. 1): A23
Abstract 82-OR.
30. Bloomgarden ZT, Dodis R, Viscoli CM, Holmboe ES, Inzucchi SE. Lower
baseline glycemia reduces apparent oral agent glucose-lowering efficacy: a
meta-regression analysis. Diabetes Care 2006; 29: 21372139.
Volume 17 No. 1 January 2015
original article
31. Defronzo RA, Stonehouse AH, Han J, Wintle ME. Relationship of baseline
HbA1c and efficacy of current glucose-lowering therapies: a meta-analysis of
randomized clinical trials. Diabet Med 2010; 27: 309317.
32. Ptaszynska A, Johnsson K, Apanovitch AM, Sugg J, Parikh SJ, List JF. Effect of
dapagliflozin on renal function. Diabetes 2012; 61(Suppl. 1): A283 Abstract
1098-P.
doi:10.1111/dom.12387 51