Review
Annals of Internal Medicine
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
A Systematic Review and Meta-analysis
Despoina Vasilakou, MD, MSc; Thomas Karagiannis, MD, MSc; Eleni Athanasiadou, MSc; Maria Mainou, MD; Aris Liakos, MD;
Eleni Bekiari, MD, PhD; Maria Sarigianni, MD, PhD, MSc; David R. Matthews, MD, DPhil; and Apostolos Tsapas, MD, PhD, MSc
Background: Sodium–glucose cotransporter 2 (SGLT2) inhibitors
are a new class of antidiabetic drugs.
Purpose: To assess the efficacy and safety of SGLT2 inhibitors in
adults with type 2 diabetes.
Data Sources: MEDLINE, EMBASE, and the Cochrane Library from
inception through April 2013 without language restrictions; regulatory authorities’ reports; and gray literature.
Study Selection: Randomized trials comparing SGLT2 inhibitors
with placebo or other medication for type 2 diabetes.
Data Extraction: Three reviewers extracted or checked data for
study characteristics, outcomes of interest, and risk of bias, and 3
reviewers summarized strength of evidence using the Grading of
Recommendations Assessment, Development and Evaluation
approach.
Data Synthesis: Sodium–glucose cotransporter 2 inhibitors were
compared with placebo in 45 studies (n ⫽ 11 232) and with active
comparators in 13 studies (n ⫽ 5175). They had a favorable effect
on hemoglobin A1c level (mean difference vs. placebo, ⫺0.66%
[95% CI, ⫺0.73% to ⫺0.58%]; mean difference vs. active com-
S
odium–glucose cotransporter 2 (SGLT2) inhibitors are
a new class of antidiabetic drugs that reduce renal glucose reabsorption in the proximal convoluted tubule, leading to increased urinary glucose excretion (1). The SGLT2
is a high-capacity, low-affinity transporter that is overexpressed and overactivated in patients with type 2 diabetes
and is responsible for 80% to 90% of renal glucose reabsorption (2). Current guidelines do not include SGLT2
inhibitors in treatment recommendations (3). In 2011, a
U.S. Food and Drug Administration (FDA) Advisory
Committee voted against approval of dapagliflozin because
of concerns about increased risk for bladder and breast
cancer (4, 5). The European Medicines Agency (EMA)
recently approved dapagliflozin for treatment of type 2 diabetes, either as monotherapy or as add-on treatment (6).
In March 2013, the FDA approved canagliflozin for use in
patients with type 2 diabetes (7, 8).
See also:
Print
Related article. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Web-Only
Supplements
CME quiz
262 © 2013 American College of Physicians
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
parators, ⫺0.06% [CI, ⫺0.18% to 0.05%]). Sensitivity analyses
incorporating unpublished data showed similar effect estimates.
Compared with other agents, SGLT2 inhibitors reduced body
weight (mean difference, ⫺1.80 kg [CI, ⫺3.50 to ⫺0.11 kg]) and
systolic blood pressure (mean difference, ⫺4.45 mm Hg [CI, ⫺5.73
to ⫺3.18 mm Hg]). Urinary and genital tract infections were more
common with SGLT2 inhibitors (odds ratios, 1.42 [CI, 1.06 to 1.90]
and 5.06 [CI, 3.44 to 7.45], respectively). Hypoglycemic risk was
similar to that of other agents. Results for cardiovascular outcomes
and death were inconclusive. An imbalance in incidence of bladder
and breast cancer was noted with dapagliflozin compared with
control.
Limitation: Most trials were rated as high risk of bias because of
missing data and last-observation-carried-forward methods.
Conclusion: Sodium–glucose cotransporter 2 inhibitors may improve short-term outcomes in adults with type 2 diabetes, but
effects on long-term outcomes and safety are unclear.
Primary Funding Source: None.
Ann Intern Med. 2013;159:262-274.
For author affiliations, see end of text.
www.annals.org
Previous systematic reviews explored the efficacy and
safety of SGLT2 inhibitors but focused primarily on studies comparing dapagliflozin with placebo, thus missing extension studies, comparative effectiveness trials, and studies
assessing newer SGLT2 inhibitors (9, 10). To update and
clarify the evidence base of the efficacy and safety of
SGLT2 inhibitors, we conducted a systematic review and
meta-analysis, based on published and unpublished randomized trials of adult patients with type 2 diabetes, that
compared SGLT2 inhibitors with placebo or other antidiabetic agents, either as monotherapy or as add-on
treatment.
METHODS
We prespecified objectives and methods, revised some
methods in response to peer review and editor comments,
and report the review in accordance with the Preferred
Reporting Items for Systematic Reviews and MetaAnalyses statement (11).
Data Sources and Searches
We identified eligible studies by searching
MEDLINE, EMBASE, and the Cochrane Library from
inception to 13 April 2013 without language restrictions.
Our search strategy included relevant substance names;
Medical Subject Heading and Emtree terms; and a filter
for identifying randomized, controlled trials (see Supplement 1, available at www.annals.org) (12, 13). We also
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
hand-searched abstracts from meetings of relevant associations (American Diabetes Association, European Association for the Study of Diabetes, International Diabetes
Federation, and American Association of Clinical Endocrinologists) from 2009 to 2012 and, in 2013, scanned Web
sites of relevant pharmaceutical companies, retrieved reports from regulatory authorities (FDA and EMA) (4, 6,
7), and searched clinical trial registries (ClinicalTrials.gov).
We perused reference lists of eligible articles and relevant
reviews, including 2 systematic reviews (9, 10). Finally, we
conducted a rapid search of MEDLINE via PubMed in
April 2013 using relevant keywords for long-term observational studies addressing potential harms or adverse effects.
Study Selection
We included randomized, controlled trials that compared an SGLT2 inhibitor with placebo or another antidiabetic medication in adults with type 2 diabetes. We
included trials regardless of language, year of publication,
or publication status. Publications retrieved from electronic
databases were imported into reference management
software. After deduplication, 2 reviewers independently
screened titles and abstracts and subsequently examined
the full text of potentially eligible reports. Two different
reviewers independently screened reports retrieved from
regulatory databases, conference abstracts, Web sites of
pharmaceutical companies, and trial registries. Disagreements at each stage of selection were arbitrated by a third
reviewer and resolved by consensus. Eligible reports were
juxtaposed against each other to remove duplicates and
maximize information yield.
Data Extraction and Risk-of-Bias Assessment
Data extraction was performed independently by 2 reviewers using a predesigned data collection form and was
checked by a third reviewer. For each eligible trial, we
extracted data on study characteristics, participants’ baseline characteristics, and efficacy and safety outcomes. Efficacy outcomes included change from baseline in hemoglobin A1c (HbA1c) level (primary outcome), body weight,
and systolic and diastolic blood pressures. Clinical outcomes of interest included all-cause mortality and cardiovascular events (myocardial infarction, stroke, death due
to cardiovascular disease, or hospitalization for unstable
angina). Information about potential harms that was extracted included incidence of any hypoglycemia, urinary
tract infections, genital tract infections, hypotension, any
serious adverse event, bladder cancer, or breast cancer.
Data about renal and bone safety and liver toxicity were
taken from information in regulatory authorities’ reports.
Hypoglycemia and other safety outcome data were extracted on the basis of definitions used in each study.
Missing data were requested via e-mails to corresponding authors or pharmaceutical companies. Data from multiple reports for the same study were collated. In cases of
contradictory material, we used data from regulatory docwww.annals.org
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
Review
uments and published articles rather than reports from
conference abstracts or Web sites.
Two reviewers independently assessed risk of bias of
each study using the Cochrane Collaboration risk-of-bias
tool (14) (Supplement 2, available at www.annals.org).
Disagreements were resolved by consensus. We explored
risk of bias across studies (publication bias) by using the
Egger statistical test (15).
Data Synthesis, Grading of Evidence, and Analysis
We used the Grading of Recommendations Assessment, Development and Evaluation approach to summarize the strength of evidence and determine confidence in
summary estimates for clinically relevant comparisons and
outcomes (16, 17). Three reviewers graded inconsistency,
risk of bias, indirectness, imprecision, and publication bias
for evidence related to each of the following outcomes:
change in HbA1c level, change in body weight, change in
systolic blood pressure, incidence of hypoglycemia, incidence of cardiovascular events, and incidence of urinary
and genital tract infections.
We conducted meta-analyses when 3 or more studies
provided relevant data. For efficacy outcomes, we analyzed
only trials of at least 12 weeks’ duration. For safety outcomes, we used eligible trials regardless of duration of the
intervention. In studies with extension periods, we used
the report with the longest intervention. We calculated
weighted mean differences (WMDs) and 95% CIs for continuous outcomes using an inverse variance random-effects
model. For dichotomous outcomes, we calculated odds ratios (ORs) and 95% CIs by using the fixed-effects Mantel–
Haenszel approach with a treatment group continuity correction for zero events, including trials with zero events in
both groups. We verified robustness of findings across different methods (Peto OR; constant correction; or treatment group correction for continuity, including and
excluding studies with zero events) (18). For SGLT2 inhibitors that had received approval, we used data for patients randomly assigned to the highest available approved
dose (5 or 10 mg for dapagliflozin and 100 or 300 mg
for canagliflozin). For other SGLT2 inhibitors, we used
data from the group allocated to the highest, most common dose. Data on incidence of all-cause mortality, any
serious adverse event, cardiovascular events, bladder cancer,
and breast cancer were extracted for all treatment groups
regardless of SGLT2 inhibitor dose.
In our main analyses, we used only data published in
journals or identified in regulatory authorities’ reports and
excluded data retrieved only from conference abstracts or
Web sites. When available, data for intention-to-treat populations were used. We performed separate analyses for
placebo-controlled trials and those with active controls and
subgroup analyses for use of SGLT2 inhibitors as monotherapy or add-on treatment. We also conducted sensitivity
analyses using data from all eligible trials regardless of information source. Heterogeneity was assessed with the I2
20 August 2013 Annals of Internal Medicine Volume 159 • Number 4 263
Review
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
statistic, with values greater than 60% representing high
heterogeneity (17). We planned to explore heterogeneity
with a sensitivity analysis that included only trials at low
risk of bias for our primary outcome. Finally, we explored
potential differences among individual SGLT2 inhibitors
by conducting separate analyses for each substance when
sufficient data were available. All statistical analyses were
done using Stata, release 11.2 (StataCorp, College Station,
Texas), and RevMan 5.2 (Nordic Cochrane Center, Copenhagen, Denmark [19]).
Role of the Funding Source
This study received no funding.
RESULTS
Search Results and Study Characteristics
We found 49 primary and 9 extension studies that met
eligibility criteria (Appendix Figure, available at www
.annals.org). Our searches of ClinicalTrials.gov and Web
sites of pharmaceutical companies identified an additional
25 completed eligible trials that could not be included
because of pending or undisclosed results (Supplement 3,
available at www.annals.org). We identified no long-term
observational studies reporting on potential harms or adverse effects.
The Appendix Table (available at www.annals.org)
shows characteristics of included trials (20 –72). Almost all
were sponsored by pharmaceutical companies. Most trials
had a double-blind design, except for 3 studies that had a
double-blind primary phase (21, 23, 60) and an open-label
extension phase (22, 24). The SGLT2 inhibitors that were
used included dapagliflozin (21 trials), canagliflozin (12
trials), ipragliflozin (8 trials), empagliflozin (3 trials),
luseogliflozin (2 trials), tofogliflozin (1 trial), ertugliflozin
(1 trial), and remogliflozin (1 trial). Sodium–glucose
cotransporter 2 inhibitors were compared with placebo in
45 studies (n ⫽ 11 232) and with active comparators in 13
studies (n ⫽ 5175) as either monotherapy or add-on treatment. Among the 13 studies with active controls, SGLT2
inhibitors were compared with metformin in 6 studies (22,
23, 25, 30, 48), sitagliptin in 5 studies (7, 59, 60, 62, 63),
and a sulfonylurea in 2 studies (43, 57). Duration of the
intervention ranged from 12 days to 104 weeks. Five studies had less than 12 weeks’ duration, 32 studies lasted between 12 and 26 weeks, and 5 trials had a duration between 48 and 52 weeks. Study duration was at least 90
weeks in 7 studies. Mean HbA1c level at baseline was available for 41 studies, ranged from 6.9% to 9.2%, and was
balanced between treatment groups.
Patients with severe renal impairment were excluded
in almost all studies, except for 1 canagliflozin trial (55).
Dapagliflozin trials included 684 patients with moderate
renal impairment (6). Two canagliflozin trials allowed enrollment of patients with moderate renal impairment (34,
38), whereas 1 dapagliflozin trial (52) and 1 canagliflozin
trial (71) involved such patients exclusively. Three trials
264 20 August 2013 Annals of Internal Medicine Volume 159 • Number 4
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
(42, 54, 55) recruited patients at high risk for cardiovascular disease.
Data Collection and Risk-of-Bias Assessment of
Included Trials
We requested additional data for 2 published studies
and 29 abstracts from corresponding authors or relevant
pharmaceutical companies. Of these, 12 responded to our
e-mails, but only 2 provided the requested data.
Overall risk of bias for the primary outcome was high
in almost all studies, primarily because of incomplete outcome data (high discontinuation rate or use of inadequate
imputation method to handle missing data) (Supplement
2). Attrition rates were high (ⱖ20%) or unbalanced between treatment groups in 14 studies and unclear in 26
studies (primarily reported as abstracts). The method of
imputation of missing data was unclear in 11 studies identified in abstract form. In most published trials, postrescue
data were excluded and missing data were imputed using a
last-observation-carried-forward (LOCF) approach. The
FDA report that was reviewed (4) included a sensitivity
analysis on different imputation methods for 2 dapagliflozin trials (21, 36). The analysis showed overstated results
when the LOCF method was used. The Egger test did not
reveal any evidence of publication bias (P ⫽ 0.89).
Glycemic Efficacy
Compared with placebo, SGLT2 inhibitors reduced
HbA1c levels when used as monotherapy (WMD, ⫺0.79%
[95% CI, ⫺0.96% to ⫺0.62%]; I2 ⫽ 71%) or add-on
treatment (WMD, ⫺0.61% [CI, ⫺0.69% to ⫺0.53%];
I2 ⫽ 73%) (Figure 1). Compared with other hypoglycemic
agents, SGLT2 inhibitors had similar glycemic efficacy
when used as monotherapy (WMD, 0.05% [CI, ⫺0.06%
to 0.16%]; I2 ⫽ 0%) or add-on treatment (WMD,
⫺0.16% [CI, ⫺0.32% to 0.00%]; I2 ⫽ 82%) (Figure 2).
When each SGLT2 inhibitor was analyzed separately,
changes in HbA1c level versus placebo were ⫺0.59% (CI,
⫺0.67% to ⫺0.50%) for dapagliflozin and ⫺0.78% (CI,
⫺0.90% to ⫺0.66%) for canagliflozin (Table 1).
Overall results were similar for both comparisons in
sensitivity analyses that included all eligible studies regardless of information source (SGLT2 inhibitor vs. placebo
WMD, ⫺0.69% [CI, ⫺0.78% to ⫺0.61%; I2 ⫽ 84%];
SGLT2 inhibitor vs. active comparator WMD, ⫺0.11%
[CI, ⫺0.21% to ⫺0.01%; I2 ⫽ 59%]) (Figures 1 and 2 of
Supplement 4, available at www.annals.org). We did not
perform a sensitivity analysis that included only trials at
low risk of bias because overall risk of bias was high for all
studies.
Body Weight
Body weight was measured as absolute change from
baseline in most trials; however, some trials reported only
the percentage of change. Thus, we conducted separate
analyses based on unit of measurement. Compared with
placebo, treatment with an SGLT2 inhibitor resulted in
reductions in absolute change (WMD, ⫺1.74 kg [CI,
www.annals.org
Review
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
Figure 1. Weighted mean difference in change in hemoglobin A1c level from baseline: SGLT2 inhibitors versus placebo.
Study, Year (Reference)
Monotherapy
SGLT2 Inhibitor
Mean Total, n
Change
(SD), %
Placebo
Weight, % Mean Difference
Mean Total, n
IV, Random (95% CI)
Change
(SD), %
Bailey et al, 2012 (20)
–0.82 (0.99) 66
0.02 (0.99)
68
2.7
–0.84 (–1.18 to –0.50)
Ferrannini et al, 2010 (21)
–0.84 (0.88) 141
–0.23 (0.89) 72
3.5
–0.61 (–0.86 to –0.36)
Ferrannini et al, 2013 (23)
–0.63 (0.82) 82
0.09 (0.82)
82
3.5
–0.72 (–0.97 to –0.47)
Fonseca et al, 2013 (25)
–0.39 (0.79) 67
0.26 (0.73)
69
3.5
–0.65 (–0.91 to –0.39)
0.37 (0.51)
Kaku et al, 2013 (27)
–0.44 (0.5)
54
4.2
–0.81 (–1.00 to –0.62)
List et al, 2009 (30)
–0.85 (0.75) 47
–0.18 (0.73) 54
3.2
–0.67 (–0.96 to –0.38)
Stenlöf et al, 2013 (34)
–1.03 (0.84) 197
0.14 (0.97) 192
4.3
–1.17 (–1.35 to –0.99)
652
591
24.9
–0.79 (–0.96 to –0.62)
Subtotal
52
Mean Difference
IV, Random (95% CI)
Heterogeneity: tau-square = 0.04; chi-square = 20.42; P = 0.002; I 2 = 71%
Test for overall effect: Z = 9.18 (P < 0.001)
Add-on Treatment
Bailey et al, 2013 (37)
–0.78 (1.1) 135
0.02 (1.28) 137
3.2
–0.80 (–1.08 to –0.52)
Bode et al, 2012 (38)
–0.73 (0.91) 229
–0.03 (0.91) 232
4.4
–0.70 (–0.87 to –0.53)
Cefalu et al, 2012 (42)
–0.38 (0.85) 448
0.08 (0.85) 451
5.0
–0.46 (–0.57 to –0.35)
Henry et al, 2012 (study 1) (48)* –2.05 (1.21) 185
–1.35 (1.25) 195
3.6
–0.70 (–0.95 to –0.45)
Henry et al, 2012 (study 2) (48)* –1.98 (1.09) 202
–1.44 (1.09) 203
3.9
–0.54 (–0.75 to –0.33)
Kohan et al, 2011 (52)
–0.44 (1.54) 82
–0.32 (1.54) 82
1.8
–0.12 (–0.59 to 0.35)
Leiter et al, 2012 (54)
–0.33 (0.94) 474
0.07 (0.94) 471
4.9
–0.40 (–0.52 to –0.28)
Ljunggren et al, 2012 (40)
–0.38 (0.54) 86
0.02 (0.56)
90
4.5
–0.40 (–0.56 to –0.24)
Matthews et al, 2012 (55)
–0.72 (0.74) 611
0.02 (0.72) 586
5.2
–0.74 (–0.82 to –0.66)
Rosenstock et al, 2012 (61)
–1.21 (0.83) 140
–0.54 (0.94) 139
4.0
–0.67 (–0.88 to –0.46)
Rosenstock et al, 2012 (62)
–0.92 (0.7)
–0.22 (0.7)
65
3.6
–0.70 (–0.94 to –0.46)
Strojek et al, 2011 (64)
–0.82 (0.74) 150
–0.13 (0.74) 143
4.4
–0.69 (–0.86 to –0.52)
Wilding et al, 2009 (66)
–0.61 (0.61) 23
0.09 (0.67)
19
2.3
–0.70 (–1.09 to –0.31)
Wilding et al, 2012 (67)
–1.01 (0.8) 194
–0.47 (0.8) 193
4.5
–0.54 (–0.70 to –0.38)
Wilding et al, 2012 (69)
–1.06 (0.99) 152
–0.13 (0.98) 150
3.8
–0.93 (–1.15 to –0.71)
Wilding et al, 2013 (70)
–0.65 (0.75) 66
–0.31 (0.76) 65
3.5
–0.34 (–0.60 to –0.08)
Yale et al, 2013 (71)
–0.44 (0.85) 89
–0.03 (0.84) 87
3.5
–0.41 (–0.66 to –0.16)
NCT01106677, 2013 (7)
–0.94 (0.76) 360
–0.17 (0.81) 181
4.7
–0.77 (–0.91 to –0.63)
NCT01106690, 2013 (7)
–1.03 (0.74) 112
–0.26 (0.75) 114
4.1
–0.77 (–0.96 to –0.58)
3802
3603
75.1
–0.61 (–0.69 to –0.53)
100.0
–0.66 (–0.73 to –0.58)
Subtotal
64
Heterogeneity: tau-square = 0.02; chi-square = 67.36; P < 0.001; I 2 = 73%
Test for overall effect: Z = 14.79 (P < 0.001)
Total
4454
4194
Heterogeneity: tau-square = 0.03; chi-square = 106.92; P < 0.001; I 2 = 77%
Test for overall effect: Z = 16.40 (P < 0.001)
Test for subgroup differences: chi-square = 3.64; P = 0.056; I 2 = 72.5%
–1.0
–0.5
Favors SGLT2
inhibitor
0
0.5
1.0
Favors placebo
Results are from IV random-effects meta-analysis. IV ⫽ inverse variance; SGLT2 ⫽ sodium–glucose cotransporter 2.
* Reference 48 includes 2 randomized trials of dapagliflozin at doses of 5 mg (study 1) and 10 mg (study 2).
⫺2.03 to ⫺1.45 kg]; I2 ⫽ 47%) (Figure 3 of Supplement
4) and percentage of change (WMD, ⫺2.37% [CI,
⫺2.73% to ⫺2.02%]; I2 ⫽ 65%) in body weight. Similarly, SGLT2 inhibitors had a favorable effect compared
www.annals.org
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
with other antihyperglycemic agents in absolute change
(WMD, ⫺1.80 kg [CI, ⫺3.50 to ⫺0.11 kg]; I2 ⫽ 97%)
(Figure 4 of Supplement 4) and percentage of change
(WMD, ⫺2.14% [CI, ⫺3.02% to ⫺1.25%]; I2 ⫽ 67%)
20 August 2013 Annals of Internal Medicine Volume 159 • Number 4 265
Review
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
Figure 2. Weighted mean difference in change in hemoglobin A1c level from baseline: SGLT2 inhibitors versus other antidiabetic
drugs.
Study, Year (Reference)
Monotherapy
SGLT2 Inhibitor
Mean Total, n
Change
(SD), %
Active Comparator Weight, %
Mean Difference
Mean Total, n
IV, Random (95% CI)
Change
(SD), %
Ferrannini et al, 2013 (23)
–0.63 (0.82) 82
–0.75 (0.76) 80
9.6
0.12 (–0.12 to 0.36)
Fonseca et al, 2013 (25)
–0.39 (0.79) 67
–0.46 (0.73) 69
9.1
0.07 (–0.19 to 0.33)
Henry et al, 2012 (study 1) (48)*
–1.19 (1.21) 196
–1.35 (1.25) 195
9.5
0.16 (–0.08 to 0.40)
Henry et al, 2012 (study 2) (48)*
–1.45 (1.05) 216
–1.44 (1.09) 203
10.9
–0.01 (–0.22 to 0.20)
List et al, 2009 (30)
–0.85 (0.75) 47
–0.73 (0.75) 56
8.0
–0.12 (–0.41 to 0.17)
608
603
47.2
0.05 (–0.06 to 0.16)
Subtotal
Mean Difference
IV, Random (95% CI)
Heterogeneity: tau-square = 0.00; chi-square = 2.76; P = 0.60; I 2 = 0%
Test for overall effect: Z = 0.88 (P = 0.38)
Add-on Treatment
Cefalu et al, 2012 (43)
–0.93 (0.87) 474
–0.82 (0.87) 473
14.6
–0.11 (–0.22 to 0.00)
Nauck et al, 2011 (57)
–0.52 (0.82) 400
–0.52 (0.82) 401
14.5
0.00 (–0.11 to 0.11)
Rosenstock et al, 2012 (62)
–0.92 (0.7)
64
–0.74 (0.62) 65
10.1
–0.18 (–0.41 to 0.05)
Schernthaner et al, 2013 (63)
–1.03 (0.96) 365
–0.66 (0.97) 374
13.5
–0.37 (–0.51 to –0.23)
1303
1313
52.8
–0.16 (–0.32 to 0.00)
100.0
–0.06 (–0.18 to 0.05)
Subtotal
Heterogeneity: tau-square = 0.02; chi-square = 16.72; P < 0.001; I 2 = 82%
Test for overall effect: Z = 1.92 (P = 0.05)
Total
1916
1911
Heterogeneity: tau-square = 0.02; chi-square = 27.70; P < 0.001;
I2
= 71%
Test for overall effect: Z = 1.07 (P = 0.28)
Test for subgroup differences: chi-square = 4.35; P = 0.037;
–1.0
I2
= 77.0%
–0.5
Favors SGLT2
inhibitor
0
0.5
1.0
Favors active
comparator
Results are from IV random-effects meta-analysis. IV ⫽ inverse variance; SGLT2 ⫽ sodium–glucose cotransporter 2.
* Reference 48 includes 2 randomized trials of dapagliflozin at doses of 5 mg (study 1) and 10 mg (study 2).
in body weight. Of note, absolute body weight reduction
for SGLT2 inhibitors versus other active comparators was
less evident and heterogeneity was eliminated in a post hoc
sensitivity analysis that excluded 1 sulfonylurea-controlled
study (57) (WMD, ⫺1.11 kg [CI, ⫺1.46 to ⫺0.76 kg];
I2 ⫽ 0%). Overall risk of bias for body weight analyses was
high.
Blood Pressure
Sodium–glucose cotransporter 2 inhibitors were associated with a reduction in systolic blood pressure compared
with placebo (WMD, ⫺3.77 mm Hg [CI, ⫺4.65 to
⫺2.90 mm Hg]; I2 ⫽ 44%) (Figure 5 of Supplement 4)
and active comparators (WMD, ⫺4.45 mm Hg [CI,
⫺5.73 to ⫺3.18 mm Hg]; I2 ⫽ 34%) (Figure 6 of Supplement 4). Diastolic blood pressure was also reduced with
SGLT2 inhibitors compared with placebo (WMD, ⫺1.75
mm Hg [CI, ⫺2.27 to ⫺1.23 mm Hg]; I2 ⫽ 0%) (Figure
7 of Supplement 4) and other antidiabetic agents (WMD,
⫺2.01 mm Hg [CI, ⫺2.62 to ⫺1.39 mm Hg]; I2 ⫽ 0%)
(Figure 8 of Supplement 4). Risk of bias was high for both
systolic and diastolic blood pressure analyses.
266 20 August 2013 Annals of Internal Medicine Volume 159 • Number 4
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
Hypoglycemia
Incidence of hypoglycemia was low in most treatment
groups, except for among patients receiving a sulfonylurea
or insulin as allocation treatment or background therapy.
The OR for any hypoglycemia with SGLT2 inhibitors was
1.28 (CI, 0.99 to 1.65; I2 ⫽ 0%) (Table 1 and Figure 9 of
Supplement 4) compared with placebo and 0.44 (CI, 0.35
to 0.54; I2 ⫽ 93%) (Figure 10 of Supplement 4) compared with other antidiabetic medications. However, exclusion of 1 sulfonylurea-controlled study (57) in a post hoc
sensitivity analysis resulted in similar hypoglycemic risk
compared with other antidiabetic agents and removed heterogeneity (OR, 1.01 [CI, 0.77 to 1.32]; I2 ⫽ 0%). Across
all studies analyzed, severe hypoglycemia (defined as an
episode requiring assistance from another person) was rare
in all treatment groups and was seen primarily in participants already receiving a sulfonylurea.
Genitourinary Tract Infections and Hypotension
Urinary tract infections were more common among
patients treated with SGLT2 inhibitors than among those
receiving placebo (OR, 1.34 [CI, 1.03 to 1.74]; I2 ⫽ 0%)
www.annals.org
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
Review
Table 1. Findings of Subgroup Analyses for Efficacy and Safety Outcomes*
Outcome
Comparison†
Studies
Contributing
Data, n
Participants
Analyzed, n
Participants With
Outcome, n
Effect Estimate‡
(95% CI)
I 2,
%
SGLT2
Comparator SGLT2
Comparator
Inhibitor
Inhibitor
Mean change in HbA1c level (%)
from baseline
Mean absolute change in body
weight (kg) from baseline
Mean percentage of change in body
weight from baseline
Mean change in systolic blood
pressure (mm Hg) from baseline
Mean change in diastolic blood
pressure (mm Hg) from baseline
Hypoglycemia
Urinary tract infection
Genital tract infection
Hypotension
SGLT2 inhibitor vs. placebo
Dapagliflozin vs. placebo
Canagliflozin vs. placebo
SGLT2 inhibitor vs. active
agent
Dapagliflozin vs. active
agent
Canagliflozin vs. active
agent
SGLT2 inhibitor vs. placebo
Dapagliflozin vs. placebo
SGLT2 inhibitor vs. active
agent§
SGLT2 inhibitor vs. placebo
Dapagliflozin vs. placebo
Canagliflozin vs. placebo
SGLT2 inhibitor vs. active
agent
SGLT2 inhibitor vs. placebo
Dapagliflozin vs. placebo
Canagliflozin vs. placebo
SGLT2 inhibitor vs. active
agent
Dapagliflozin vs. active
agent
SGLT2 inhibitor vs. placebo
Dapagliflozin vs. placebo
Canagliflozin vs. placebo
SGLT2 inhibitor vs. active
agent
Dapagliflozin vs. active
agent
SGLT2 inhibitor vs. placebo
Dapagliflozin vs. placebo
Canagliflozin vs. placebo
Ipragliflozin vs. placebo
SGLT2 inhibitor vs. active
agent§
Dapagliflozin vs. active
agent§
SGLT2 inhibitor vs. placebo
Dapagliflozin vs. placebo
Canagliflozin vs. placebo
Ipragliflozin vs. placebo
SGLT2 inhibitor vs. active
agent
Dapagliflozin vs. active
agent
SGLT2 inhibitor vs. placebo
Dapagliflozin vs. placebo
Canagliflozin vs. placebo
SGLT2 inhibitor vs. active
agent
Dapagliflozin vs. active
agent
SGLT2 inhibitor vs. placebo
Dapagliflozin vs. placebo
Canagliflozin vs. placebo
SGLT2 inhibitor vs. active
agent
Dapagliflozin vs. active
agent
26
15
8
9
4454
2425
1814
1911
4194
2371
1607
1916
NA
NA
NA
NA
NA
NA
NA
NA
⫺0.66 (⫺0.73 to ⫺0.58)
⫺0.59 (⫺0.67 to ⫺0.50)
⫺0.78 (⫺0.90 to ⫺0.66)
⫺0.06 (⫺0.18 to 0.05)
77
60
77
71
4
859
855
NA
NA
0.01 (⫺0.08 to 0.10)
0
3
903
912
NA
NA
⫺0.22 (⫺0.40 to ⫺0.04)
76
15
11
4
1702
1398
571
1638
1332
577
NA
NA
NA
NA
NA
NA
⫺1.74 (⫺2.03 to ⫺1.45)
⫺1.92 (⫺2.23 to ⫺1.60)
⫺1.11 (⫺1.46 to ⫺0.76)
47
35
0
8
3
5
3
2222
982
1240
488
2209
994
1215
499
NA
NA
NA
NA
NA
NA
NA
NA
⫺2.37 (⫺2.73 to ⫺2.02)
⫺2.06 (⫺2.38 to ⫺1.74)
⫺2.61 (⫺3.09 to ⫺2.13)
⫺2.14 (⫺3.02 to ⫺1.25)
65
0
66
67
21
14
6
6
3666
2273
1327
1240
3548
2180
1303
1247
NA
NA
NA
NA
NA
NA
NA
NA
⫺3.77 (⫺4.65 to ⫺2.90)
⫺3.20 (⫺4.20 to ⫺2.21)
⫺4.79 (⫺6.39 to ⫺3.18)
⫺4.45 (⫺5.73 to ⫺3.18)
44
29
53
34
4
799
804
NA
NA
⫺3.95 (⫺5.57 to ⫺2.33)
38
16
12
3
6
1762
1348
348
1240
1652
1242
345
1247
NA
NA
NA
NA
NA
NA
NA
NA
⫺1.75 (⫺2.27 to ⫺1.23)
⫺1.74 (⫺2.35 to ⫺1.13)
⫺1.84 (⫺2.96 to ⫺0.72)
⫺2.01 (⫺2.62 to ⫺1.39)
0
0
0
0
4
799
804
NA
NA
⫺1.72 (⫺2.48 to ⫺0.96)
0
21
12
4
3
7
1920
1315
360
147
1059
1857
1255
356
148
1058
204
144
55
5
169
174
128
42
2
169
1.28 (0.99 to 1.65)
1.20 (0.88 to 1.64)
1.53 (0.93 to 2.54)
2.02 (0.49 to 8.24)
1.01 (0.77 to 1.32)
0
0
0
0
0
3
469
465
5
11
0.49 (0.18 to 1.39)
0
21
12
3
3
8
2059
1455
350
147
1465
1944
1393
347
148
1465
139
108
19
11
116
103
75
17
10
84
1.34 (1.03 to 1.74)
1.43 (1.05 to 1.94)
1.12 (0.57 to 2.19)
1.12 (0.47 to 2.68)
1.42 (1.06 to 1.90)
0
0
0
0
25
4
875
873
89
55
1.69 (1.19 to 2.40)
7
20
13
3
8
2049
1466
350
1465
1981
1401
347
1465
137
108
17
150
37
30
5
32
3.50 (2.46 to 4.99)
3.48 (2.33 to 5.20)
3.26 (1.23 to 8.61)
5.06 (3.44 to 7.45)
0
0
0
0
4
875
873
93
21
4.81 (2.97 to 7.81)
0
12
9
3
5
1535
1239
296
1252
1468
1177
291
1251
14
10
4
18
7
6
1
6
1.57 (0.74 to 3.35)
1.38 (0.59 to 3.24)
2.49 (0.47 to 13.27)
2.68 (1.14 to 6.29)
0
0
0
2
4
875
873
12
5
2.14 (0.83 to 5.53)
10
Continued on following page
www.annals.org
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
20 August 2013 Annals of Internal Medicine Volume 159 • Number 4 267
Review
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
Table 1—Continued
Outcome
Comparison†
Studies
Contributing
Data, n
Participants
Analyzed, n
Participants With
Outcome, n
Effect Estimate‡
(95% CI)
I 2,
%
SGLT2
Comparator SGLT2
Comparator
Inhibitor
Inhibitor
Cardiovascular event㥋
Serious adverse event
All-cause mortality
Bladder cancer
Breast cancer
SGLT2 inhibitor vs. placebo
or active agent
Dapagliflozin vs. placebo or
active agent
Canagliflozin vs. placebo or
active agent
SGLT2 inhibitor vs. placebo
or active agent
Dapagliflozin vs. placebo or
active agent
Canagliflozin vs. placebo or
active agent
Ipragliflozin vs. placebo or
active agent
SGLT2 inhibitor vs. placebo
or active agent
Dapagliflozin vs. placebo or
active agent
Canagliflozin vs. placebo or
active agent
SGLT2 inhibitor vs. placebo
or active agent
Dapagliflozin vs. placebo or
active agent
Canagliflozin vs. placebo or
active agent
SGLT2 inhibitor vs. placebo
or active agent
Dapagliflozin vs. placebo or
active agent
Canagliflozin vs. placebo or
active agent
25
11 372
5808
182
101
0.89 (0.70 to 1.14)
0
14
4359
1941
48
30
0.73 (0.46 to 1.16)
0
10
6769
3705
132
71
0.95 (0.71 to 1.26)
0
24
6324
3051
226
145
0.90 (0.72 to 1.13)
0
14
4362
1865
164
98
0.91 (0.70 to 1.20)
0
5
1094
799
49
39
0.90 (0.58 to 1.39)
0
3
597
217
10
5
0.86 (0.29 to 2.51)
0
21
5771
2989
23
8
NE
NE
17
4803
2320
19
6
NE
NE
4
968
669
4
2
1.18 (0.29 to 4.90)
0
25
12 149
6824
14
5
NE
NE
16
5501
3184
9
1
NE
NE
9
6648
3640
5
4
NE
NE
25
8328
4685
21
7
NE
NE
16
5501
3184
9
1
NE
NE
9
2827
1501
12
6
NE
NE
HbA1c ⫽ hemoglobin A1c; NA ⫽ not applicable; NE ⫽ not estimable; SGLT2 ⫽ sodium–glucose cotransporter 2.
* Based on data from individual study reports and reports from the U.S. Food and Drug Administration Advisory Committee (4, 7) and European Medicines Agency (6).
† We synthesized data when ⱖ3 studies provided relevant data.
‡ For change in HbA1c level, body weight, and systolic and diastolic blood pressures, the effect estimate is the weighted mean difference, calculated using an inverse
variance–weighted random-effects model. For the remaining outcomes, the effect estimate is the odds ratio, calculated using the fixed-effects Mantel–Haenszel approach with
a treatment group continuity correction for zero events, including trials with zero events in both groups (18). Effect estimates could not be calculated when data were based
only on pooled analyses from regulatory authorities’ reports and we could not reproduce data from the original trials.
§ Results from sensitivity analysis, excluding 1 sulfonylurea-controlled study (57).
㛳 Cardiovascular death, myocardial infarction, nonfatal stroke, or hospitalization for unstable angina.
or other hypoglycemic agents (OR, 1.42 [CI, 1.06 to
1.90]; I2 ⫽ 25%). We also found an increased incidence of
genital tract infections with SGLT2 inhibitors compared
with placebo (OR, 3.50 [CI, 2.46 to 4.99]; I2 ⫽ 0%) and
active comparators (OR, 5.06 [CI, 3.44 to 7.45]; I2 ⫽ 0%)
and a higher risk for hypotension with SGLT2 inhibitors
than with other antidiabetic medications (OR, 2.68 [CI,
1.14 to 6.29]; I2 ⫽ 2%) (Table 1).
Death and Serious Adverse Events
All-cause mortality did not differ between SGLT2 inhibitors and placebo or active comparators, although relatively few deaths have been reported in trials. Twenty-three
deaths were reported among patients treated with SGLT2
inhibitors (n ⫽ 5771), and 8 were reported among patients
receiving either placebo (4 of 1738 patients) or an active
comparator (4 of 1251 patients). The OR for any serious
268 20 August 2013 Annals of Internal Medicine Volume 159 • Number 4
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
adverse event with SGLT2 inhibitors versus comparators
was 0.90 (CI, 0.72 to 1.13; I2 ⫽ 0%) (Table 1).
Cardiovascular Outcomes
Our meta-analysis of cardiovascular outcomes for dapagliflozin, which was based on 14 trials (n ⫽ 6300),
yielded an OR of 0.73 (CI, 0.46 to 1.16; I2 ⫽ 0%) compared with control. This estimate was consistent with the
FDA report and the more recent EMA report (hazard ratio
[HR], 0.82 [CI, 0.58 to 1.15]). In a pooled analysis of 2
dapagliflozin trials in patients with established cardiovascular disease (42, 54), the HR for the composite cardiovascular end point (cardiovascular death, myocardial infarction, stroke, and hospitalization for unstable angina) was
1.07 (CI, 0.64 to 1.72) versus placebo (6).
Canagliflozin was not associated with an increased risk
for the composite cardiovascular outcome compared with
www.annals.org
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
placebo or active comparator on the basis of data from 10
trials that included a total of 10 474 patients (OR, 0.95
[CI, 0.71 to 1.26]; I2 ⫽ 0%), although CIs were wide. In
the FDA report (7), the HR for nonfatal stroke was higher
in patients receiving canagliflozin (6876 patient-years) than
in the control groups (3470 patient-years) (HR, 1.46 [CI,
0.83 to 2.58]). In addition, an imbalance in incidence of
cardiovascular events observed during the first 30 days of
the dedicated cardiovascular trial (55) between canagliflozin (13 of 2886 patients) and placebo (1 of 1441 patients) resulted in an HR of 6.50 (CI, 0.85 to 49.66),
possibly due to volume depletion after canagliflozin initiation. This imbalance was not evident after 30 days.
Bladder and Breast Cancer
For dapagliflozin, data on bladder and breast cancer
that were retrieved from regulatory databases and other
sources produced a pool of 5501 patients with at least
5000 patient-years of exposure who were treated with dapagliflozin and a total of 3184 patients with at least 2350
patient-years of exposure who received placebo or an active
comparator (4, 6). Nine cases of bladder cancer were identified in patients treated with dapagliflozin as opposed to 1
case in patients receiving placebo. All patients were men,
with a median time of diagnosis of 399 days (range, 43 to
727 days). Baseline characteristics for bladder cancer risk
factors were similar between dapagliflozin and placebo
groups. As noted in the FDA report, included trials were
not powered to distinguish statistically between the 9 cases
and 1 case of bladder cancer; however, the observed event
rate of 9 cases exceeds the expected rate of only 2 cases in
an age-matched reference male population of patients with
diabetes (4).
Breast cancer was reported in 9 women aged 53 to 74
years who were treated with dapagliflozin and in 1 patient
in the control group. Study day of diagnosis ranged from
day 6 to day 334, which is much shorter than the average
of at least 5 years of exposure regarded as sufficient for
breast cancer to be detectable (6).
Data on cancer in patients treated with canagliflozin
are based primarily on an FDA report (7). A pooled analysis of 9 trials with approximately 8000 person-years of
exposure did not show any difference in incidence of bladder cancer between canagliflozin (5 of 6648 patients) and
control (4 of 3640 patients) groups. Similarly, incidence of
breast cancer did not differ between canagliflozin (12 of
2827 patients) and comparators (6 of 1501 patients).
Renal and Bone Safety and Liver Toxicity
Data on renal and bone safety and liver toxicity were
retrieved from regulatory authorities’ reports (4, 6, 7). In
patients with moderate renal impairment, incidence of
renal-related adverse events was increased with dapagliflozin and canagliflozin compared with placebo. In patients
with normal or mildly impaired renal function, high doses
of canagliflozin (300 mg) were associated with increased
www.annals.org
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
Review
incidence of renal-related adverse events (14 of 834 patients) compared with placebo (4 of 646 patients).
In a pooled analysis, there was no imbalance in fractures between dapagliflozin and comparator groups (overall
incidence, ⬍1.6%). However, in patients with moderate
renal impairment, incidence of fractures was higher in dapagliflozin recipients (4 and 8 events in the 5- and 10-mg
groups, respectively) than in placebo recipients (no events).
An updated safety analysis of canagliflozin trials noted a
nonsignificant imbalance in fracture incidence in patients
treated with canagliflozin compared with control patients.
Regarding liver-related adverse events, regulatory authorities’ reports concluded that slight imbalances among patients treated with dapagliflozin or canagliflozin and control groups were probably not associated with the study
drug.
Grading of Evidence
Quality of evidence was downgraded to low for glycemic efficacy, percentage of change in body weight, incidence of any hypoglycemia, and cardiovascular outcomes
because of high risk of bias and inconsistency or imprecision. Quality was downgraded to moderate for effect on
systolic blood pressure, incidence of urinary and genital
tract infections, and absolute change in body weight because of high or unclear risk of bias (Table 2).
DISCUSSION
Sodium– glucose cotransporter 2 inhibitors were associated with a 0.66% reduction in HbA1c level and had
glycemic efficacy similar to that of other antidiabetic
agents. They also had a favorable effect on body weight
and blood pressure. Risk for hypoglycemia was similar to
that of metformin or sitagliptin and lower than that of
sulfonylureas. Increased incidence of urinary and genital
tract infections was probably due to glucosuria associated
with the use of SGLT2 inhibitors. In patients with moderate renal impairment, use of dapagliflozin or high doses
of canagliflozin was associated with increased incidence of
renal-related adverse events. Data on cardiovascular outcomes and death were inconclusive. A numerical imbalance
in nonfatal stroke events among patients treated with canagliflozin needs clarification and confirmation. We also
noted a numerical imbalance of bladder and breast cancer
cases between patients treated with dapagliflozin and control patients. The number of observed cases for these types
of cancer exceeds the expected number of cases in the general diabetic population, as reported in epidemiologic reports (4). However, early detection after short exposure
and potential detection bias due to frequent urinalysis mitigate against a causative relationship. Hence, no robust
conclusions can be drawn pending accumulation of longterm data.
To our knowledge, this systematic review provides the
most up-to-date and comprehensive summary of the benefits and risks of SGLT2 inhibitors as of April 2013. We
20 August 2013 Annals of Internal Medicine Volume 159 • Number 4 269
Review
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
Table 2. Quality of Evidence for Clinically Relevant Outcomes*
Outcome
Follow-up,
wk
Illustrative Comparative Risks†
Assumed Risk (Active Comparator)
Corresponding Risk (SGLT2 Inhibitor)
Mean change in HbA1c level
(%) from baseline
12–52
The mean change in HbA1c level ranged across
control groups from ⫺0.37% to 0.16%
The mean change in HbA1c level in the intervention groups
was 0.06% lower§ (95% CI, 0.18% lower to 0.05%
higher)
Mean absolute change in body
weight (kg) from baseline
12–24
The mean change in body weight ranged across
control groups from ⫺1.37 to ⫺0.71 kg
The mean change in body weight in the intervention groups
was 1.11 kg lower§ (CI, 1.46 to 0.76 kg lower)
Mean percentage of change in
body weight from baseline
12–52
The mean change in body weight ranged across
control groups from ⫺2.80% to ⫺1.00%
Mean change in systolic blood
pressure (mm Hg) from
baseline
Incidence of any hypoglycemia
12–52
12–52
The mean change in systolic blood pressure
ranged across control groups from ⫺6.00 to
⫺2.40 mm Hg
16 cases per 100 patients
The mean change in body weight in the intervention groups
was 2.14 percentage points lower§ (CI, 3.02 to 1.25
percentage points lower)
The mean change in systolic blood pressure in the
intervention groups was 4.45 mm Hg lower§ (CI, 5.73 to
3.18 mm Hg lower)
16 cases per 100 patients (CI, 13 to 20 cases per 100
patients)
12–102
2 cases per 100 patients
2 cases per 100 patients (CI, 1 to 2 cases per 100 patients)
12–52
6 cases per 100 patients
12–52
2 cases per 100 patients
8 cases per 100 patients (CI, 6 to 10 cases per 100
patients)
10 cases per 100 patients (CI, 7 to 14 cases per 100
patients)
Incidence of cardiovascular
events
Incidence of urinary tract
infections
Incidence of genital tract
infections
HbA1c ⫽ hemoglobin A1c; SGLT2 ⫽ sodium–glucose cotransporter 2.
* Among studies that compared SGLT2 inhibitors with active comparators (any antidiabetic medication) in adults with type 2 diabetes mellitus.
† The assumed risk is based on the median risk in the control group across studies. The corresponding risk is based on the assumed risk in the comparison group and the
relative effect of the intervention.
‡ Evidence was graded using the Grading of Recommendations Assessment, Development and Evaluation guidelines (16, 17). Evidence could be rated as high quality (further
research is very unlikely to change our confidence in the estimate of effect), moderate quality (further research is likely to have an important effect on our confidence in the
estimate of effect and may change the estimate), low quality (further research is very likely to have an important effect on our confidence in the estimate of effect and is likely
to change the estimate), or very low quality (we are very uncertain about the estimate).
§ Lower change indicates better outcome.
㛳 Downgraded for inconsistency due to heterogeneity of effect estimate.
¶ Downgraded because most of the studies had high risk of bias.
** The monotherapy subgroup included SGLT2 inhibitors as first-line antidiabetic treatment. The add-on therapy subgroup included SGLT2 inhibitors as add-on therapy
to existing antidiabetic treatment.
†† Downgraded because most of the studies had unclear risk of bias.
‡‡ Downgraded for imprecision due to wide CIs in results.
identified 2 pertinent systematic reviews through a rapid
search of MEDLINE. Musso and colleagues (10) documented the favorable effect of dapagliflozin on glycemic
control, blood pressure, and weight on the basis of data
retrieved before December 2010 from electronic databases
and conference abstracts. Their meta-analysis included 13
placebo-controlled trials (primarily for dapagliflozin), 5 of
which had a duration of less than 28 days. In a more recent
systematic review, Clar and associates (9) examined the
efficacy and safety of SGLT2 inhibitors in dual or triple
antidiabetic therapy. Their search, which was done in July
2012, provided a total of 8 trials with a duration between
12 and 52 weeks, 7 of which were for dapagliflozin and 1
of which was for canagliflozin. Thus, conclusions from
prior meta-analyses were based on a small number of trials
of short duration that primarily compared dapagliflozin
with placebo. Our review included a much larger number
of both placebo-controlled trials and those with active controls, including extensions and studies for newer SGLT2
inhibitors, identified from multiple electronic databases
and gray literature. We also used data from regulatory databases to summarize information about bladder and breast
270 20 August 2013 Annals of Internal Medicine Volume 159 • Number 4
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
cancer, cardiovascular outcomes, liver toxicity, and renal
and bone safety. Finally, we rated the overall strength of
evidence using the Grading of Recommendations Assessment, Development and Evaluation approach.
We acknowledge several limitations of the body of
evidence and review process. Most included studies used
LOCF methods to impute missing data, which can introduce significant bias into the results (73, 74). The combination of LOCF imputation with exclusion of postrescue
data can lead to overstated results, as noted in the FDA
report for dapagliflozin (4). Unfortunately, this approach
to handling missing data was used in recently published
trials of SGLT2 inhibitors (23, 25, 27, 34, 63, 70, 71).
Most studies received industry funding, which introduced
further bias into the results (75). No conclusions about
differences among individual SGLT2 inhibitors could be
made because of a lack of head-to-head trials. Results for
cardiovascular outcomes, death, and incidence of cancer
are based primarily on data from trials designed to assess
short-term efficacy outcomes and should therefore be interpreted with caution. Limitations at the review level
are related to the high degree of heterogeneity observed in
www.annals.org
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
Review
Table 2—Continued
Odds Ratio (95% CI)
Participants, n
Studies, n
Quality of Evidence‡
Comments
In the monotherapy subgroup, the mean difference was 0.05%
(CI, ⫺0.06% to 0.16%) with no heterogeneity; in the add-on
therapy subgroup, the mean difference was ⫺0.16%
(CI, ⫺0.32% to 0.00%) with high heterogeneity**
Excludes 1 sulfonylurea-controlled study (57) with no heterogeneity
of results; when the study was included, the mean difference
was ⫺1.80 kg (CI, ⫺3.50 to ⫺0.11 kg) with high heterogeneity
–
–
3827
9
Low㛳¶
–
1128
4
Moderate¶
–
987
3
Low㛳¶
–
2487
6
Moderate¶
1.01 (0.77–1.32)
2117
7
Low††‡‡
0.89 (0.70–1.14)
17 180
25
Low††‡‡
1.42 (1.06–1.90)
2930
8
Moderate††
Excludes 1 sulfonylurea-controlled study (57) with no heterogeneity
of results; when the study was included, the odds ratio was
0.44 (CI, 0.35 to 0.54) with high heterogeneity
Results for cardiovascular events refer to SGLT2 inhibitors vs. any
other antidiabetic medication or placebo
–
5.06 (3.44–7.45)
2930
8
Moderate††
–
analyses of HbA1c level, body weight, and incidence of
hypoglycemia. We explained heterogeneity of body
weight and hypoglycemia by sensitivity analyses excluding
sulfonylurea-controlled trials. For HbA1c level, heterogeneity could be attributed to a combination of factors,
including differences in individual SGLT2 inhibitors,
background antidiabetic treatment, or class of active
comparator.
Future research should explore differences among
individual SGLT2 inhibitors and differences between
SGLT2 inhibitors and other antidiabetic agents. We identified several ongoing trials, some of which have been designed to assess safety outcomes (Supplement 3). On completion, they are expected to provide adequate data to draw
safer inferences about long-term safety and cardiovascular
outcomes. Moreover, we identified 25 eligible completed
trials that could not be included in our analysis because of
undisclosed results (Supplement 3). Most of these trials
were completed in 2012 and will probably be published in
the near future. Incomplete reporting and inappropriate
analysis plans are consistently criticized, and many organizations urge the need to rectify this problem (76). Timely
disclosure of trial results (77), judicious analysis plans (73,
74), and access to raw data (78) are essential to ensure valid
results and guide evidence-based therapeutic decision making. In the meantime, it seems justifiable for systematic
reviews to include data from regulatory authorities in their
information sources (79).
In conclusion, SGLT2 inhibitors seem to be an effective treatment option for adults with type 2 diabetes. They
may improve some short-term outcomes, but further rewww.annals.org
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
–
search is necessary to clarify effects on long-term clinical
outcomes, diabetic complications, and safety.
From Aristotle University Thessaloniki, Thessaloniki, Greece, and University of Oxford, Oxford, United Kingdom.
Note: All authors had full access to all of the data in the study and bear
responsibility for the integrity of the data analysis.
Potential Conflicts of Interest: Dr. Matthews: Other: Novo Nordisk,
Boehringer Ingelheim, AstraZeneca, SB Communications, Merck,
Takeda Chemical Industries, Johnson & Johnson, GlaxoSmithKline,
Servier. Dr. Tsapas: Grant: Boehringer Ingelheim, Novartis, Novo Nordisk, Sanofi-Aventis; Other: Novo Nordisk. All other authors have no
disclosures. Disclosures can also be viewed at www.acponline.org/authors
/icmje/ConflictOfInterestForms.do?msNum⫽M13-0277.
Requests for Single Reprints: Apostolos Tsapas, MD, PhD, MSc, Sec-
ond Medical Department, Aristotle University Thessaloniki, 49 Konstantinoupoleos Street, 54642 Thessaloniki, Greece; e-mail, atsapas
@auth.gr.
Current author addresses and author contributions are available at
www.annals.org.
References
1. Ferrannini E, Solini A. SGLT2 inhibition in diabetes mellitus: rationale and
clinical prospects. Nat Rev Endocrinol. 2012;8:495-502. [PMID: 22310849]
2. 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;54:342734. [PMID: 16306358]
3. Inzucchi SE, Bergenstal RM, Buse JB, Diamant M, Ferrannini E, Nauck M,
et al; American Diabetes Association (ADA). Management of hyperglycemia in
20 August 2013 Annals of Internal Medicine Volume 159 • Number 4 271
Review
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
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;35:1364-79. [PMID: 22517736]
4. U.S. Food and Drug Administration. FDA Briefing Document. NDA
202293. Dapagliflozin Tablets, 5 and 10 mg. Rockville, MD: U.S. Food and
Drug Administration; 2011. Accessed at www.fda.gov/downloads/Advisory
Committees/CommitteesMeetingMaterials/drugs/EndocrinologicandMetabolic
DrugsAdvisoryCommittee/ucm262994.pdf on 1 April 2013.
5. Burki TK. FDA rejects novel diabetes drug over safety fears. Lancet. 2012;
379:507. [PMID: 22334883]
6. European Medicines Agency. Assessment Report: Forxiga (Dapagliflozin).
Procedure no. EMEA/H/C/002322. London: European Medicines Agency;
2012. Accessed at www.ema.europa.eu/docs/en_GB/document_library/EPAR
_-_Public_assessment_report/human/002322/WC500136024.pdf on 1 April
2013.
7. U.S. Food and Drug Administration. FDA Briefing Document. NDA
204042. Invokana (Canagliflozin) Tablets. Rockville, MD: U.S. Food and Drug
Administration; 2013. Accessed at www.fda.gov/downloads/AdvisoryCommittees
/CommitteesMeetingMaterials/Drugs/EndocrinologicandMetabolicDrugs
AdvisoryCommittee/UCM334550.pdf on 24 January 2013.
8. U.S. Food and Drug Administration. FDA approves Invokana to treat type 2
diabetes [news release]. Rockville, MD: U.S. Food and Drug Administration;
29 March 2013. Accessed at www.fda.gov/NewsEvents/Newsroom
/PressAnnouncements/ucm345848.htm on 1 April 2013.
9. Clar C, Gill JA, Court R, Waugh N. Systematic review of SGLT2 receptor
inhibitors in dual or triple therapy in type 2 diabetes. BMJ Open. 2012;2.
[PMID: 23087012]
10. Musso G, Gambino R, Cassader M, Pagano G. A novel approach to control
hyperglycemia in type 2 diabetes: sodium glucose co-transport (SGLT) inhibitors:
systematic review and meta-analysis of randomized trials. Ann Med. 2012;44:
375-93. [PMID: 21495788]
11. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP,
et al. The PRISMA statement for reporting systematic reviews and meta-analyses
of studies that evaluate health care interventions: explanation and elaboration.
Ann Intern Med. 2009;151:W65-94. [PMID: 19622512]
12. Scottish Intercollegiate Guidelines Network. Search Filters. Edinburgh,
United Kingdom: Scottish Intercollegiate Guidelines Network; 2013. Accessed at
www.sign.ac.uk/methodology/filters.html on 8 January 2013.
13. Lefebvre C, Manheimer E, Glanville J. Searching for studies. In: Higgins
JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions.
Version 5.1.0. The Cochrane Collaboration; 2011.
14. Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD,
et al; Cochrane Bias Methods Group. The Cochrane Collaboration’s tool for
assessing risk of bias in randomised trials. BMJ. 2011;343:d5928. [PMID:
22008217]
15. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis
detected by a simple, graphical test. BMJ. 1997;315:629-34. [PMID: 9310563]
16. Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, et al. GRADE
guidelines: 1. Introduction-GRADE evidence profiles and summary of findings
tables. J Clin Epidemiol. 2011;64:383-94. [PMID: 21195583]
17. Guyatt GH, Oxman AD, Kunz R, Woodcock J, Brozek J, Helfand M,
et al; GRADE Working Group. GRADE guidelines: 7. Rating the quality of
evidence—inconsistency. J Clin Epidemiol. 2011;64:1294-302. [PMID:
21803546]
18. Diamond GA, Bax L, Kaul S. Uncertain effects of rosiglitazone on the risk
for myocardial infarction and cardiovascular death. Ann Intern Med. 2007;147:
578-81. [PMID: 17679700]
19. Review Manager (RevMan) [computer program]. Version 5.2. Copenhagen,
Denmark: Nordic Cochrane Center, Cochrane Collaboration; 2012.
20. Bailey CJ, Iqbal N, T’joen C, List JF. Dapagliflozin monotherapy in drugnaı̈ve patients with diabetes: a randomized-controlled trial of low-dose range.
Diabetes Obes Metab. 2012;14:951-9. [PMID: 22776824]
21. 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;33:2217-24. [PMID: 20566676]
22. Woo V, Tang W, Salsali A, List FJ. Long-term efficacy of dapagliflozin
monotherapy in patients with type 2 diabetes mellitus. Presented at International
Diabetes Federation 21st World Diabetes Congress, Dubai, United Arab Emirates, 4 – 8 December 2011.
272 20 August 2013 Annals of Internal Medicine Volume 159 • Number 4
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
23. Ferrannini E, Seman L, Seewaldt-Becker E, Hantel S, Pinnetti S, Woerle
HJ. A phase IIb, randomized, placebo-controlled study of the SGLT2 inhibitor
empagliflozin in patients with type 2 diabetes. Diabetes Obes Metab. 2013;15:
721-8. [PMID: 23398530]
24. Woerle HJ, Ferrannini E, Berk A, Manun’ebo M, Pinnetti S, Broedl UC.
Safety and efficacy of empagliflozin as monotherapy or add-on to metformin in a
78-week open-label extension study in patients with type 2 diabetes [Abstract].
Diabetes. 2012;61(Suppl 1A):LB13.
25. Fonseca VA, Ferrannini E, Wilding JP, Wilpshaar W, Dhanjal P, Ball G,
et al. Active- and placebo-controlled dose-finding study to assess the efficacy,
safety, and tolerability of multiple doses of ipragliflozin in patients with type 2
diabetes mellitus. J Diabetes Complications. 2013;27:268-73. [PMID:
23276620]
26. Inagaki N, Kondo N, Iwasaki T, Maruyama N, Susuta Y, Sakai M, et al.
Canagliflozin, a novel inhibitor of sodium glucose co-transporter 2 (SGLT2)
improves glycemic control and reduces body weight in Japanese type 2 diabetes
mellitus (T2DM) [Abstract]. Diabetes. 2011;60(Suppl 1):A274.
27. Kaku K, Inoue S, Matsuoka O, Kiyosue A, Azuma H, Hayashi N, et al.
Efficacy and safety of dapagliflozin as a monotherapy for type 2 diabetes mellitus
in Japanese patients with inadequate glycaemic control: a phase II multicentre,
randomized, double-blind, placebo-controlled trial. Diabetes Obes Metab. 2013;
15:432-40. [PMID: 23194084]
28. Kashiwagi A, Utsuno A, Kazuta K. ASP1941, a novel, selective SGLT2
inhibitor, was effective and safe in Japanese healthy volunteers and patients with
type 2 diabetes mellitus [Abstract]. Diabetes. 2010;59(Suppl 1):75-OR.
29. Kashiwagi A, Takinami Y, Kazuta K, Yoshida S, Utsuno A, Nagase I.
Ipragliflozin improved glycaemic control with additional benefits of reductions of
body weight and blood pressure in Japanese patients with type 2 diabetes mellitus: BRIGHTEN Study [Abstract]. Diabetologia. 2011;54(Suppl 1):S68-9.
30. List JF, Woo V, Morales E, Tang W, Fiedorek FT. Sodium-glucose cotransport inhibition with dapagliflozin in type 2 diabetes. Diabetes Care. 2009;32:
650-7. [PMID: 19114612]
31. Schwartz SL, Akinlade B, Klasen S, Kowalski D, Zhang W, Wilpshaar W.
Safety, pharmacokinetic, and pharmacodynamic profiles of ipragliflozin
(ASP1941), a novel and selective inhibitor of sodium-dependent glucose cotransporter 2, in patients with type 2 diabetes mellitus. Diabetes Technol Ther.
2011;13:1219-27. [PMID: 21854192]
32. Seino Y, Sasaki T, Fukatsu A, Samukawa Y, Sakai S, Watanabe T. A novel
potent and highly selective renal sodium-glocose co-transporter 2 (SGLT2) inhibitor, TS-071, improves glycaemic control and lowers body weight in Japanese
patiens with type 2 diabetes mellitus [Abstract]. Diabetologia. 2011;54(Suppl
1):S68.
33. Seino Y, Sasaki T, Fukatsu A, Samukawa Y, Sakai S, Watanabe T. Luseogliflozin (TS-071), a selective SGLT2 inhibitor, improves glycemic control and
lowers body weight in Japanese patients with type 2 diabetes mellitus [Abstract].
Diabetes. 2012;61(Suppl 1):A266-7.
34. Stenlöf K, Cefalu WT, Kim KA, Alba M, Usiskin K, Tong C, et al. Efficacy
and safety of canagliflozin monotherapy in subjects with type 2 diabetes mellitus
inadequately controlled with diet and exercise. Diabetes Obes Metab. 2013;15:
372-82. [PMID: 23279307]
35. AstraZeneca. A 24-week randomised, double-blind, parallel-group, multicentre, placebo-controlled phase III trial to evaluate the efficacy and safety of
dapagliflozin as monotherapy in Japanese subjects with type 2 diabetes who have
inadequate glycemic control with diet and exercise. Clinical Study Report Synopsis. Wilmington, DE: AstraZeneca; 2012. Accessed at www.astrazenecaclinicaltrials
.com/_mshost800325/content/clinical-trials/resources/pdf/D1692C00006 on 1
April 2013.
36. 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:
2223-33. [PMID: 20609968]
37. Bailey CJ, Gross JL, Hennicken D, Iqbal N, Mansfield TA, List JF. Dapagliflozin add-on to metformin in type 2 diabetes inadequately controlled with
metformin: a randomized, double-blind, placebo-controlled 102-week trial.
BMC Med. 2013;11:43. [PMID: 23425012]
38. Bode B, Stenlöf K, Sullivan D, Fung A, Usiskin K, Meininger G. Efficacy
and safety of canagliflozin (CANA), a sodium glucose co-transporter 2 inhibitor
(SGLT2), in older subjects with type 2 diabetes mellitus [Abstract]. Diabetologia.
2012;55(Suppl 1):S315.
www.annals.org
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
39. Bolinder J, Ljunggren Ö, Kullberg J, Johansson L, Wilding J, Langkilde
AM, 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;97:102031. [PMID: 22238392]
40. Ljunggren Ö, Bolinder J, Johansson L, Wilding J, Langkilde AM, Sjöström
CD, et al. Dapagliflozin has no effect on markers of bone formation and resorption or bone mineral density in patients with inadequately controlled type 2
diabetes mellitus on metformin. Diabetes Obes Metab. 2012;14:990-9. [PMID:
22651373]
41. Bolinder J, Ljunggren O, Johansson L. Dapagliflozin produces long-term
reductions in body weight, waist circumference and total fat mass in patients with
type 2 diabetes inadequately controlled on metformin [Abstract]. Diabetologia.
2012;55(Suppl 1):S308.
42. Cefalu WT, Leiter LA, Debruin TW, Gausenilsson I, Sugg J, Parikh SJ.
Dapagliflozin treatment for type 2 diabetes mellitus patients with comorbid cardiovascular disease and hypertension [Abstract]. Diabetes. 2012;61(Suppl 1):
A271.
43. Cefalu TW, Leiter LA, Niskanen L, Xie J, Millington D, Canovatchel W,
et al. Efficacy and safety of canagliflozin, a sodium glucose co-transporter 2 inhibitor, compared with glimepiride in patients with type 2 diabetes on background metformin [Abstract]. Diabetes. 2012;61(Suppl 1A):LB10.
44. Devineni D, Morrow L, Hompesch M, Skee D, Vandebosch A, Murphy J,
et al. Canagliflozin improves glycaemic control over 28 days in subjects with type
2 diabetes not optimally controlled on insulin. Diabetes Obes Metab. 2012;14:
539-45. [PMID: 22226086]
45. Dobbins RL, O’Connor-Semmes R, Kapur A, Kapitza C, Golor G, Mikoshiba I, et al. Remogliflozin etabonate, a selective inhibitor of the sodiumdependent transporter 2 reduces serum glucose in type 2 diabetes mellitus patients. Diabetes Obes Metab. 2012;14:15-22. [PMID: 21733056]
46. Goto K, Kashiwagi A, Kazuta K, Yoshida S, Ueyama E, Utsuno A. Ipragliflozin reduces A1C and body weight in type 2 diabetes patients who have inadequate glycemic control on metformin alone: ILLUMINATE study [Abstract].
Diabetes. 2012;61(Suppl 1):A269.
47. Heise T, Seewaldt-Becker E, Macha S, Hantel S, Pinnetti S, Seman L, et al.
Safety, tolerability, pharmacokinetics and pharmacodynamics following 4 weeks’
treatment with empagliflozin once daily in patients with type 2 diabetes. Diabetes
Obes Metab. 2013;15:613-21. [PMID: 23356556]
48. 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;66:446-56.
[PMID: 22413962]
49. Jabbour S, Hardy E, Sugg J, Parikh S. Dapagliflozin as add-on therapy to
sitagliptin with or without metformin: a randomized, double-blind, placebocontrolled study [Abstract]. Diabetes. 2012;61(Suppl 1):A275-6.
50. Kadowaki T, Ikeda S, Takano Y, Cynshi O, Christ AD, Boerlin V, et al.
Tofogliflozin, a novel and selective SGLT2 inhibitor improves glycemic control
and lowers body weight in patients with type 2 diabetes mellitus inadequately
controlled on stable metformin or diet and exercise alone [Abstract]. Diabetes.
2012;61(Suppl 1):A22.
51. Kashiwagi A, Shiga T, Akiyama N. Ipragliflozin reduced HbA1c and body
weight in Japanese type 2 diabetes patients who have inadequate glyceamic control on sulfonylurea or pioglitazone alone [Abstract]. Diabetologia. 2012;
55(Suppl 1):S302-3.
52. Kohan DE, Fioretto P, List J, Tang W. Efficacy and safety of dapagliflozin
in patients with type 2 diabetes and moderate renal impairment. J Am Soc
Nephrol. 2011;22(Suppl):232A-3A.
53. 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;85:513-9. [PMID: 19129749]
54. Leiter LA, Cefalu WT, Debruin TW, Gausenilsson I, Sugg J, Parikh SJ.
Efficacy and safety of dapagliflozin for type 2 diabetes mellitus patients with a
history of cardiovascular disease [Abstract]. Diabetes. 2012;61(Suppl 1):A287.
55. Matthews DR, Fulcher G, Perkovic V, DeZeeuw D, Mahaffey KW, Rosenstock J, et al. Efficacy and safety of canagliflozin (CANA), an inhibitor of sodium
glucose co-transporter 2 (SGLT2), added-on to insulin therapy ⫹/⫺ oral agents
in type 2 diabetes [Abstract]. Diabetologia. 2012;55(Suppl 1):S314-5.
56. Mudaliar S, Henry RR, Boden G, Smith S, Chalamandaris AG, Iqbal N,
et al. Changes in insulin sensitivity as measured by glucose disposal rate and acute
www.annals.org
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
Review
insulin secretion with the sodium glucose co-transporter 2 inhibitor dapagliflozin
[Abstract]. Diabetologia. 2011;54(Suppl 1):S349.
57. Nauck MA, Del Prato S, Meier JJ, Durán-Garcı́a S, Rohwedder K, Elze M,
et al. Dapagliflozin versus glipizide as add-on therapy in patients with type 2
diabetes who have inadequate glycemic control with metformin: a randomized,
52-week, double-blind, active-controlled noninferiority trial. Diabetes Care.
2011;34:2015-22. [PMID: 21816980]
58. Del Prato S, Nauck MA, Rohwedder K, Theuerkauf A, Langkilde A,
Parikh S. Long-term efficacy and safety of add-on dapagliflozin vs add-on glipizidein patients with type 2 diabetes mellitus inadequately controlled with metformin: 2-year results [Abstract]. Diabetologia. 2011;54(Suppl 1):S348.
59. Nucci G, Amin NB, Wang X, Lee DS, Rusnak JM. The sodium glucose
co-transporter-2 (SGLT2) inhibitor, PF04971729, provides multi-faceted improvements in diabetic patients inadequately controlled on metformin [Abstract].
Diabetologia. 2011;54(Suppl 1):S347.
60. Rosenstock J, Jelaska A, Seman LJ, Pinnetti S, Hantel S, Woerle HJ.
Efficacy and safety of BI 10773, a new sodium glucose cotransporter-2 (SGLT-2)
inhibitor, in type 2 diabetes patients inadequately controlled on metformin
[Abstract]. Diabetes. 2011;60(Suppl 1):A271.
61. 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;35:1473-8. [PMID: 22446170]
62. Rosenstock J, Aggarwal N, Polidori D, Zhao Y, Arbit D, Usiskin K, et al;
Canagliflozin DIA 2001 Study Group. Dose-ranging effects of canagliflozin, a
sodium-glucose cotransporter 2 inhibitor, as add-on to metformin in subjects
with type 2 diabetes. Diabetes Care. 2012;35:1232-8. [PMID: 22492586]
63. Schernthaner G, Gross JL, Rosenstock J, Guarisco M, Fu M, Yee J, et al.
Canagliflozin compared with sitagliptin for patients with type 2 diabetes who do
not have adequate glycemic control with metformin plus sulfonylurea: a 52-week
randomized trial. Diabetes Care. 2013. [PMID: 23564919]
64. Strojek K, Yoon KH, Hruba V, Elze M, Langkilde AM, Parikh S. Effect of
dapagliflozin in patients with type 2 diabetes who have inadequate glycaemic
control with glimepiride: a randomized, 24-week, double-blind, placebocontrolled trial. Diabetes Obes Metab. 2011;13:928-38. [PMID: 21672123]
65. Strojek K, Hruba V, Elze M, Langkilde A, Parikh S. Efficacy and safety of
dapagliflozin as an add-on to glimepiride in T2DM inadequately controlled with
glimepiride alone over 48 weeks. Presented at International Diabetes Federation
21st World Diabetes Congress, Dubai, United Arab Emirates, 4 – 8 December
2011.
66. Wilding JP, Norwood P, T’joen C, Bastien A, List JF, Fiedorek FT. A study
of dapagliflozin in patients with type 2 diabetes receiving high doses of insulin
plus insulin sensitizers: applicability of a novel insulin-independent treatment.
Diabetes Care. 2009;32:1656-62. [PMID: 19528367]
67. Wilding JP, Woo V, Soler NG, Pahor A, Sugg J, Rohwedder K, et al;
Dapagliflozin 006 Study Group. Long-term efficacy of dapagliflozin in patients
with type 2 diabetes mellitus receiving high doses of insulin: a randomized trial.
Ann Intern Med. 2012;156:405-15. [PMID: 22431673]
68. Wilding JP, Woo VC, Rohwedder K, Sugg JE, Parikh SJ. Long-term effectiveness of dapagliflozin over 104 weeks in patients with type 2 diabetes poorly
controlled with insulin [Abstract]. Diabetes. 2012;61(Suppl 1):A267-8.
69. Wilding J, Mathieu C, Vercruysse F, Usiskin K, Deng L, Canovatchel W.
Canagliflozin (CANA), a sodium glucose co-transporter 2 inhibitor, improves
glycemic control and reduces body weight in subjects with type 2 diabetes (T2D)
inadequately controlled with metformin (MET) and sulfonylurea (SU)
[Abstract]. Diabetes. 2012;61(Suppl 1):A262.
70. Wilding JP, Ferrannini E, Fonseca VA, Wilpshaar W, Dhanjal P, Houzer
A. Efficacy and safety of ipragliflozin in patients with type 2 diabetes inadequately
controlled on metformin: a dose-finding study. Diabetes Obes Metab. 2013;15:
403-9. [PMID: 23163880]
71. Yale JF, Bakris G, Cariou B, Yue D, David-Neto E, Xi L, et al. Efficacy and
safety of canagliflozin in subjects with type 2 diabetes and chronic kidney disease.
Diabetes Obes Metab. 2013;15:463-73. [PMID: 23464594]
72. AstraZeneca. A 16-week, multicentre, randomised, double-blind, placebocontrolled phase III study to evaluate the safety and efficacy of dapagliflozin
2.5 mg bid, 5 mg bid and 10 mg qd versus placebo in patients with type 2
diabetes who are inadequately controlled on metformin-IR monotherapy. Clinical Study Report Synopsis. Wilmington, DE: AstraZeneca; 2012. Accessed at
www.astrazenecaclinicaltrials.com/_mshost800325/content/clinical-trials/resources
/pdf/D1691C00003 on 1 April 2013.
20 August 2013 Annals of Internal Medicine Volume 159 • Number 4 273
Review
Sodium–Glucose Cotransporter 2 Inhibitors for Type 2 Diabetes
73. Drawing inferences from incomplete data. In: Panel of Handling Missing
Data in Clinical Trials, National Research Council, eds. The Prevention and
Treatment of Missing Data in Clinical Trials. Washington, DC: National Academies Pr; 2010:65-9.
74. Moher D, Hopewell S, Schulz KF, Montori V, Gøtzsche PC, Devereaux
PJ, et al; Consolidated Standards of Reporting Trials Group. CONSORT 2010
Explanation and Elaboration: Updated guidelines for reporting parallel group
randomised trials. J Clin Epidemiol. 2010;63:e1-37. [PMID: 20346624]
75. Lundh A, Sismondo S, Lexchin J, Busuioc OA, Bero L. Industry sponsor
VISITS
THE
ANNALS BOOTH
ship and research outcome. Cochrane Database Syst Rev. 2012;12:MR000033.
[PMID: 23235689]
76. Bourgeois FT, Murthy S, Mandl KD. Outcome reporting among drug trials
registered in ClinicalTrials.gov. Ann Intern Med. 2010;153:158-66. [PMID:
20679560]
77. Pub. L. No. 110-85, 121 Stat 823.
78. Missing trial data—briefing notes. AllTrials.net. Accessed at www.alltrials.net
/wp-content/uploads/2013/01/Missing-trials-briefing-note.pdf on 27 January 2013.
79. Godlee F. Goodbye PubMed, hello raw data [Letter]. BMJ. 2011;342:d212.
AT
SUBSPECIALTY MEETINGS
Annals staff will be at these upcoming meetings:
IDWeek, 3–5 October 2013, San Francisco
Chest 2013, 27–29 October 2013, Chicago
American College of Rheumatology, 27–29 October 2013, San Diego
Kidney Week, 7–9 November 2013, Atlanta
American Heart Association, 17–19 November 2013, Dallas
American Society of Hematology, 7–9 December 2013, New Orleans
Stop by the ACP/Annals booth and register to be a peer reviewer or
discuss your thoughts for submissions or topic coverage with Annals
staff.
274 20 August 2013 Annals of Internal Medicine Volume 159 • Number 4
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
www.annals.org
Annals of Internal Medicine
Current Author Addresses: Drs. Vasilakou, Karagiannis, Mainou,
Liakos, Bekiari, Sarigianni, and Tsapas and Ms. Athanasiadou: Second
Medical Department, Aristotle University Thessaloniki, Hippokratio
General Hospital, 49 Konstantinoupoleos Street, 54642 Thessaloniki,
Greece.
Dr. Matthews: Harris Manchester College, Mansfield Road, Oxford
OX1 3TD, United Kingdom.
Author Contributions: Conception and design: D. Vasilakou, T.
Karagiannis, E. Athanasiadou, E. Bekiari, M. Sarigianni, D.R. Matthews,
A. Tsapas.
Analysis and interpretation of the data: D. Vasilakou, T. Karagiannis,
E. Athanasiadou, M. Mainou, A. Liakos, E. Bekiari, M. Sarigianni,
D.R. Matthews, A. Tsapas.
Drafting of the article: D. Vasilakou, T. Karagiannis, M. Mainou,
A. Tsapas.
www.annals.org
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
Critical revision of the article for important intellectual content:
D. Vasilakou, T. Karagiannis, E. Athanasiadou, A. Liakos, E. Bekiari,
M. Sarigianni, D.R. Matthews, A. Tsapas.
Final approval of the article: D. Vasilakou, T. Karagiannis, E. Athanasiadou, A. Liakos, E. Bekiari, M. Sarigianni, A. Tsapas.
Provision of study materials or patients: D. Vasilakou, E. Athanasiadou,
A. Liakos, A. Tsapas.
Statistical expertise: D. Vasilakou, E. Athanasiadou, M. Mainou, A. Liakos, M. Sarigianni, A. Tsapas.
Administrative, technical, or logistic support: D. Vasilakou, A. Liakos.
Collection and assembly of data: D. Vasilakou, T. Karagiannis,
E. Athanasiadou, A. Liakos, A. Tsapas.
80. Roche Trials Database. A Study of RO4998452 in Patients With Type 2
Diabetes Mellitus. Accessed at http://roche-trials.com/studyResultGet.action
?studyResultNumber⫽BC25405 on 1 April 2013.
20 August 2013 Annals of Internal Medicine Volume 159 • Number 4
Appendix Figure. Summary of evidence search and selection.
Records identified from other sources (n = 52)
Conference abstracts: 25
ClinicalTrials.gov: 24
Pharmaceutical companies’ Web sites: 3
Records identified through database searches
(n = 743)
MEDLINE: 266
EMBASE: 393
Cochrane Library: 59
Regulatory databases (FDA and EMA): 25
Duplicate records removed (n = 158)
Studies already identified in
regulatory databases (n = 7)
Completed studies excluded
because of undisclosed results
(n = 25)
Records included (n = 20)
Primary studies: 15
Extensions: 7
Records screened (title and
abstract) (n = 585)
Records excluded (n = 484)
Full-text records assessed for
eligibility (n = 101)
Full-text records excluded (n = 66)
Editorial: 1
Translation of original article: 4
Trial did not assess any outcome
of interest: 7
Review: 50
Other publication type: 4
Records included (n = 35)
Primary studies: 34
Extensions: 2
Records included in systematic review (n = 55)
Primary studies: 49
Retrieved as conference abstract: 13
Retrieved both as conference abstract and from regulatory database: 7
Retrieved as journal article: 9
Retrieved both as journal article and from regulatory database: 16
Retrieved from pharmaceutical companies’ Web sites: 2
Retrieved from regulatory database: 2
Extensions: 9
Retrieved as conference abstract: 7
Retrieved as journal article: 1
Retrieved both as journal article and from regulatory database: 1
EMA ⫽ European Medicines Agency; FDA ⫽ U.S. Food and Drug Administration.
20 August 2013 Annals of Internal Medicine Volume 159 • Number 4
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
www.annals.org
Appendix Table. Characteristics of Studies and Participants Included in the Systematic Review
Study, Year (Reference)
Duration of
Intervention,
wk
Information Source
Study Groups
SGLT2 inhibitors vs. placebo or active comparator as monotherapy
Bailey et al, 2012 (20)
24
Journal article and RAR Dapagliflozin, 5 mg
Placebo
Primary study: Ferrannini et al, 24 and 78
Journal article and RAR Dapagliflozin, 10 mg
2010 (21)
(primary); abstract
Placebo
Extension study: Woo et al,
(extension)
2011 (22)*
Primary study: Ferrannini et al, 12 and 78
Journal article
Empagliflozin, 25 mg
2013 (23)
(primary); abstract
Placebo
Extension study: Woerle et al,
(extension)
Metformin
2012 (study 1) (24)†‡
Fonseca et al, 2013 (25)
12
Journal article
Ipragliflozin, 50 mg
Placebo
Metformin
Inagaki et al, 2011 (26)
12
Abstract
Canagliflozin, 300 mg
Placebo
Kaku et al, 2013 (27)
12
Journal article and RAR Dapagliflozin, 10 mg
Placebo
Kashiwagi et al, 2010 (28)
12
Abstract
Ipragliflozin, 50 mg
Placebo
Kashiwagi et al, 2011 (29)
18
Abstract
Ipragliflozin, 50 mg
Placebo
List et al, 2009 (30)
12
Journal article and RAR Dapagliflozin, 10 mg
Placebo
Metformin
Schwartz et al, 2011 (31)§
4
Journal article
Ipragliflozin, 50 mg
Placebo
Seino et al, 2011 (32)
12
Abstract
Luseogliflozin, 5 mg
Placebo
Seino et al, 2012 (33)
12
Abstract
Luseogliflozin, 5 mg
Placebo
Stenlöf et al, 2013 (34)
26
Journal article and RAR Canagliflozin, 300 mg
Placebo
NCT01294423, 2012 (35)
24
Pharmaceutical
Dapagliflozin, 10 mg
company Web site
Placebo
SGLT2 inhibitors vs. placebo or active comparator as add-on therapy
Primary study: Bailey et al,
24 and 78
Journal article and RAR Dapagliflozin, 10 mg, plus
2010 (36)
(primary); journal
metformin
Extension study: Bailey et al,
article (extension)
Placebo plus metformin
2013 (37)
Bode et al, 2012 (38)
26
Abstract and RAR
Canagliflozin, 300 mg, plus OAD
Placebo plus OAD
Primary study: Bolinder et al,
24, 26, and 52 Journal article and RAR Dapagliflozin, 10 mg, plus
2012 (39)
(primary); journal
metformin
Extension studies: Ljunggren
article and RAR
Placebo plus metformin
et al, 2012 (40); Bolinder
(Ljunggren et al);
et al, 2012 (41)
abstract (Bolinder
et al)
Cefalu et al, 2012 (42)
24
Abstract and RAR
Dapagliflozin, 10 mg, with or
without insulin or OAD
Placebo with or without insulin
or OAD
Cefalu et al, 2012 (43)
52
Abstract and RAR
Canagliflozin, 300 mg, plus
metformin
Glimepiride plus metformin
Devineni et al, 2012 (44)§
4
Journal article
Canagliflozin, 300 mg, plus
insulin with or without OAD
Placebo plus insulin with or
without OAD
Patients
Randomly
Assigned, n
Mean
HbA1c
Level at
Baseline
(SD), %
68
68
141
72
7.9 (1.0)
7.8 (1.1)
8.0 (1.0)
7.8 (0.9)
82
82
80
Mean
Duration
of
Diabetes
(SD), y
1.4 (3.2)
1.1 (2.0)
0.4
0.5
Mean Body
Weight at
Baseline
(SD), kg
85.4 (19.4)
90.0 (18.0)
93.1 (20.5)
88.8 (19.0)
7.8 (0.8) NR
7.8 (0.8) NR
8.1 (0.9) NR
NR
NR
NR
67
69
69
75
75
52
54
72
69
62
67
47
54
56
12
13
61
52
54
57
197
192
88
87
8.1 (0.8)
7.8 (0.8)
8.0 (0.9)
NR
NR
8.2 (0.7)
8.1 (0.7)
NR
NR
NR
NR
8.0 (0.8)
7.9 (0.9)
7.6 (0.8)
NR
NR
8.2 (1.0)
7.9 (0.7)
7.9 (0.7)
7.9 (0.9)
8.0 (1.0)
8.0 (1.0)
NR
NR
90.7 (20.8)
81.8 (17.6)
84.1 (21.8)
NR
NR
70.4 (17.5)
68.9 (14.9)
NR
NR
NR
NR
86.0 (17.0)
89.0 (18.0)
88.0 (20.0)
85.1 (12.7)
89.4 (13.5)
66.3 (12.4)
68.4 (13.4)
72.6 (13.9)
67.6 (13.1)
86.9 (20.5)
87.5 (19.5)
NR
NR
135
7.9 (0.8)
6.1 (5.4) 86.3 (17.5)
137
8.1 (1.0)
5.8 (5.1) 87.7 (19.2)
236
237
91
7.7 (0.8) 11.3 (7.2) 88.8
7.8 (0.8) 11.4 (7.3) 91.3
7.2 (0.4)
6.0 (4.5) 92.1 (14.1)
91
7.2 (0.5)
4.6 (4.9)
4.6 (5.9)
4.1 (4.7)
NR
NR
4.7 (4.7)
4.7 (3.8)
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
4.3 (4.7)
4.2 (4.1)
NR
NR
5.5 (5.3) 90.9 (13.7)
455
8.2 (0.8) 12.6 (8.7) 92.6 (20.6)
459
8.1 (0.8) 12.3 (8.2) 93.6 (19.5)
485
7.8 (0.9)
482
10
7.8 (0.9)
6.6 (5.0) 86.6
NR
NR
94.1 (16.2)
9
NR
6.7 (5.5) 86.6
NR
95.1 (14.0)
Continued on following page
www.annals.org
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
20 August 2013 Annals of Internal Medicine Volume 159 • Number 4
Appendix Table—Continued
Study, Year (Reference)
Dobbins et al, 2012 (45)§
Goto et al, 2012 (46)
Heise et al, 2013 (47)§
Henry et al, 2012 (study
1) (48)‡㛳
Henry et al, 2012 (study
2) (48)‡㛳
Jabbour et al, 2012 (49)
Duration of
Intervention,
wk
Information Source
12 d
Journal article
24
4
24
24
48
Kadowaki et al, 2012 (50)
12
Kashiwagi et al, 2012 (study
1) (51)‡
24
Kashiwagi et al, 2012 (study
2) (51)‡
24
Kohan et al, 2011 (52)
24
Komoroski et al, 2009 (53)§
Leiter et al, 2012 (54)
Matthews et al, 2012 (55)
Mudaliar et al, 2011 (56)
Primary study: Nauck et al,
2011 (57)
Extension study: Del Prato
et al, 2011 (58)
Nucci et al, 2011 (59)
2
24
18
12
52 and 52
12
Study Groups
Remogliflozin, 1000 mg, with or
without metformin
Placebo with or without
metformin
Abstract
Ipragliflozin, 50 mg, plus
metformin
Placebo plus metformin
Journal article
Empagliflozin, 25 mg, with or
without OAD
Placebo with or without OAD
Journal article and RAR Dapagliflozin, 5 mg, plus
metformin
Placebo plus metformin
Dapagliflozin, 5 mg, plus placebo
Journal article and RAR Dapagliflozin, 10 mg, plus
metformin
Placebo plus metformin
Dapagliflozin, 10 mg, plus
placebo
Abstract
Dapagliflozin, 10 mg, plus
sitagliptin with or without
metformin
Placebo plus sitagliptin with or
without metformin
Abstract
Tofogliflozin, 40 mg, plus
metformin
Placebo plus metformin
Abstract
Ipragliflozin, 50 mg, plus
sulfonylurea
Placebo plus sulfonylurea
Abstract
Ipragliflozin, 50 mg, plus
pioglitazone
Placebo plus pioglitazone
Abstract and RAR
Dapagliflozin, 10 mg, plus OAD
Placebo plus OAD
Journal article
Dapagliflozin, 5 mg, with or
without metformin
Placebo with or without
metformin
Abstract and RAR
Dapagliflozin, 10 mg, with or
without insulin or OAD
Placebo with or without insulin
or OAD
Abstract and RAR
Canagliflozin, 300 mg, plus OAD
or insulin with or without
sulfonylurea
Placebo plus OAD or insulin with
or without sulfonylurea
Abstract
Dapagliflozin, 5 mg, plus
metformin plus insulin
secretagogue
Placebo plus metformin plus
insulin secretagogue
Journal article and RAR Dapagliflozin, 2.5 to 10 mg, plus
(primary); abstract
metformin
(extension)
Glipizide plus metformin
Abstract
Ertugliflozin, 25 mg, plus
metformin
Placebo plus metformin
Sitagliptin plus metformin
Patients
Randomly
Assigned, n
Mean
HbA1c
Level at
Baseline
(SD), %
Mean
Duration
of
Diabetes
(SD), y
Mean Body
Weight at
Baseline
(SD), kg
9
8.0 (0.6) NR
91.9 (15.6)
9
8.0 (0.6) NR
86.5 (20.8)
112
8.3 (0.7)
7.5 (5.7) 68.7 (14.0)
56
16
8.4 (0.7)
7.5 (0.8)
8.1 (5.2) 67.6 (11.2)
5.8 (3.3) NR
16
194
6.9 (0.9)
9.2 (1.3)
6.9 (6.3) NR
1.6 (2.4) 84.1 (19.5)
201
203
211
9.2 (1.3)
9.1 (1.4)
9.1 (1.3)
1.6 (2.6) 85.6 (20.0)
1.6 (3.1) 86.2 (21.1)
2.2 (3.3) 88.4 (19.7)
208
219
9.1 (1.3)
9.1 (1.3)
1.9 (4.0) 87.2 (19.4)
2.1 (3.8) 88.5 (19.3)
223
7.9 (0.8)
5.7 (4.9) 91.0 (21.6)
224
8.0 (0.8)
5.6 (5.4) 82.0 (20.9)
7.9
6.4
66
81.6
65
165
7.9
6.0
84.0
8.4 (0.6) 10.3 (7.1) 68.8 (12.4)
75
97
8.3 (0.7) 10.8 (6.2) 63.9 (11.4)
8.2 (0.7)
6.3 (4.7) 73.2 (13.4)
54
85
84
11
8.4 (0.6)
7.7 (5.3)
8.2 (1.0) NR
8.5 (1.3) NR
NR
NR
73.0 (15.7)
NR
NR
NR
NR
NR
8
NR
480
8.0 (0.8) 13.5 (8.2) 94.5 (17.8)
482
8.1 (0.8) 13.0 (8.4) 93.2 (16.8)
627
8.3 (1.0) NR
NR
610
8.2 (1.0) NR
NR
23
7.5 (0.8) NR
99.8 (22.6)
21
7.6 (0.7) NR
99.0 (15.3)
406
7.7 (0.9)
6.0 (5.0) 88.4
408
7.7 (0.9)
7.0 (6.0) 87.6
55
8.3 (1.2)
6.0 (4.0) 81.8 (17.3)
54
55
8.1 (1.0)
8.2 (1.1)
6.4 (5.5) 83.8 (17.4)
6.3 (4.5) 85.5 (19.4)
Continued on following page
20 August 2013 Annals of Internal Medicine Volume 159 • Number 4
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
www.annals.org
Appendix Table—Continued
Study, Year (Reference)
Duration of
Intervention,
wk
Information Source
Study Groups
Patients
Randomly
Assigned, n
Mean
HbA1c
Level at
Baseline
(SD), %
12 and 78
Abstract (primary and
extension)
Empagliflozin, 25 mg, plus
metformin
Sitagliptin plus metformin
166
7.9 (0.8) NR
89.5 (16.2)
56
8.0 (0.9) NR
88.6 (14.9)
140
8.4 (1.0)
5.8 (6.4)
84.8 (22.2)
139
64
8.3 (1.0)
7.7 (1.0)
5.1 (5.1)
5.9 (5.2)
86.4 (21.3)
87.3 (15.9)
65
65
377
7.8 (0.8)
7.6 (1.0)
8.1 (0.9)
6.4 (5.0)
5.6 (4.7)
9.4 (6.1)
85.9 (19.5)
87.2 (18.0)
87.4 (23.2)
378
8.1 (0.9)
9.7 (6.3)
89.1 (23.2)
151
8.1 (0.8)
7.2 (5.5)
80.6
145
8.2 (0.7)
7.4 (5.7)
80.9
Primary study: Rosenstock
et al, 2011 (60)¶
Extension study: Woerle et al,
2012 (study 2) (24)†‡
Rosenstock et al, 2012 (61)
48
Rosenstock et al, 2012 (62)
12
Schernthaner et al, 2013 (63)
Primary study: Strojek et al,
2011 (64)
Extension study: Strojek et al,
2011 (65)
Wilding et al, 2009 (66)
Primary study: Wilding et al,
2012 (67)
Extension study: Wilding et al,
2012 (68)
Wilding et al, 2012 (69)
52
24 and 24
12
48 and 56
26
Wilding et al, 2013 (70)
12
Yale et al, 2013 (71)
26
NCT01106677, 2013 (7)
26
NCT01106690, 2013 (7)
NCT01217892, 2012 (72)
26
16
Journal article and RAR Dapagliflozin, 10 mg, plus
pioglitazone
Placebo plus pioglitazone
Journal article
Canagliflozin, 300 mg, plus
metformin
Placebo plus metformin
Sitagliptin plus metformin
Journal article and RAR Canagliflozin, 300 mg, plus
metformin plus sulfonylurea
Sitagliptin plus metformin plus
sulfonylurea
Journal article and RAR Dapagliflozin, 10 mg, plus
(primary); abstract
glimepiride
(extension)
Placebo plus glimepiride
Mean
Duration
of
Diabetes
(SD), y
Mean Body
Weight at
Baseline
(SD), kg
Journal article and RAR Dapagliflozin, 10 mg, plus insulin
24
plus OAD
Placebo plus insulin plus OAD
23
Journal article and RAR Dapagliflozin, 10 mg, plus insulin 194
(primary); abstract
plus OAD
(extension)
Placebo plus insulin plus OAD
193
8.5 (0.8) 13.5 (7.3)
94.5 (19.8)
Abstract and RAR
156
8.1 (1.0)
9.4 (6.4)
93.5
156
8.1 (0.9) 10.3 (6.7)
90.8
7.8 (0.7)
86.7 (13.7)
Canagliflozin, 300 mg, plus
metformin plus sulfonylurea
Placebo plus metformin plus
sulfonylurea
Journal article
Ipragliflozin, 50 mg, plus
metformin
Placebo plus metformin
Journal article and RAR Canagliflozin, 300 mg, plus OAD
Placebo plus OAD
RAR
Canagliflozin, 300 mg, plus
metformin
Placebo plus metformin
Sitagliptin plus metformin
RAR
Canagliflozin, 300 mg, plus
metformin plus pioglitazone
Placebo plus metformin plus
pioglitazone
Pharmaceutical
Dapagliflozin, 10 mg, plus
company Web site
metformin
Placebo plus metformin
68
8.4 (0.7) 11.8 (5.8) 103.4 (10.2)
8.4 (0.9) 13.8 (7.3) 101.8 (16.5)
8.6 (0.8) 14.2 (7.3)
94.5 (16.8)
6.0 (5.3)
66
89
90
367
7.7 (0.6)
5.7 (3.2)
89.0 (14.5)
8.0 (0.8) 17.0 (7.8)
90.2 (18.1)
8.0 (0.9) 16.4 (10.1) 92.8 (17.4)
8.0
NR
NR
183
366
114
8.0
NR
8.0
NR
NR
NR
NR
NR
NR
115
8.0
NR
NR
99
NR
NR
NR
101
NR
NR
NR
HbA1c ⫽ hemoglobin A1c; NR ⫽ not reported; OAD ⫽ oral antidiabetic drug; RAR ⫽ regulatory authorities’ report (references 4, 6, and 7); SGLT2 ⫽ sodium– glucose
cotransporter 2.
* Patients in the placebo group were switched to metformin in the extension study.
† Patients in the placebo group were switched to empagliflozin in the extension study.
‡ Report includes 2 separate randomized trials.
§ Studies ⬍12 wk in duration were included only in the analyses of adverse events.
㛳 Data from these trials were included both in the analysis versus placebo as add-on therapy and in the analysis versus active comparator as monotherapy.
¶ Data from the extension study are presented because baseline characteristics were undisclosed in the primary study.
www.annals.org
Downloaded From: http://annals.org/ by a Sheng LI Rd User on 08/21/2013
20 August 2013 Annals of Internal Medicine Volume 159 • Number 4