See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/270163841

Adverse effects and safety of SGLT-2 inhibitors

Article  in  Diabetes & Metabolism · December 2014

DOI: 10.1016/S1262-3636(14)72693-X

CITATIONS READS

57 544

2 authors:

Serge Halimi Bruno Vergès

University of Grenoble University of Burgundy
639 PUBLICATIONS   6,741 CITATIONS    631 PUBLICATIONS   7,685 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Continuous Medical Education View project

REHADIAB Study View project

All content following this page was uploaded by Bruno Vergès on 05 January 2018.

The user has requested enhancement of the downloaded file.

 Diabetes & Metabolism 40 (2014) S28-S34

Adverse effects and safety of SGLT-2 inhibitors

S. Halimia,*, B. Vergèsb
a
Scientific University Joseph-Fourier, and Diabetology Department Pavillon les Écrins, BP 217X, University Hospital Grenoble,
38043 Grenoble Cedex, France
b
Department of Endocrinology and Diabetology, University Hospital Dijon, and INSERM CRI 866, Dijon, France

Abstract
In type 2 diabetes (T2DM), glycaemic control delays the development and slows the progression of complications. Although there are
numerous glucose-lowering agents in clinical use, only approximately half of T2DM patients achieve glycaemic control, while undesirable
side-effects, such as hypoglycaemia and body weight gain, often impede treatment in those taking these medications. Thus, there is a
need for novel agents and treatment options. Sodium – glucose cotransporter-2 inhibitors (SGLT-2-i) have recently been developed for the
treatment of T2DM. The available data suggest a good tolerability profile for the three available drugs – canagliflozin, dapagliflozin and
empagliflozin – approved by the US Food and Drug Administration (FDA) for the American market as well as in other countries. The most
frequently reported adverse events with SGLT-2-i are female genital mycotic infections, urinary tract infections and increased urination. The
pharmacodynamic response to SGLT-2-i declines with increasing severity of renal impairment, requiring dosage adjustments or restrictions
with moderate-to-severe renal dysfunction. Most patients treated with SGLT-2-i also have a modest reduction in blood pressure and modest
effects on serum lipid profiles, some of which are beneficial (increased high-density lipoprotein cholesterol and decreased triglycerides) and
others which are not (increased low-density lipoprotein cholesterol, LDL-C). A number of large-scale and longer-term cardiovascular trials
are now ongoing. In patients treated with dapagliflozin, a non-significant excess number of breast and bladder cancers has been reported;
considered as due to a bias, this is nevertheless being followed in the ongoing trials. No other significant safety issues have been reported
so far. Although there is some benefit for several cardiovascular risk factors such as HbA1c, high blood pressure, obesity and increases in
LDL-C, adequately powered trials are still required to determine the effects of SGLT-2-i on macrovascular outcomes.
© 2014 Elsevier Masson SAS. All rights reserved.

Keywords: Diabetes; Type 2 diabetes; SGLT-2 inhibitors; Adverse events; Genital infection; Blood pressure; Cardiovascular endpoints

1. Introduction directly or indirectly dependent on the level of insulin secretion,

with the main exceptions being metformin, thiazolidinediones
Type 2 diabetes mellitus (T2DM) is an epidemic disease and α-glucosidase inhibitors. On the other hand, these newer
that affects more than 360 million people worldwide [1]. antidiabetic drugs have no direct effects on pancreatic β-cell
Today we have a greater choice of treatments for T2DM dysfunction.
patients, including several oral and injectable agents. In addi- A new class of oral antidiabetic drugs (OADs) – namely,
tion to lifestyle measures, which are always necessary, these the sodium-glucose cotransporter-2 inhibitors (SGLT-2-i), or
drugs have demonstrated an ability to improve glycaemic “gliflozins” – promotes urinary glucose excretion by inhibit-
control [2,3]. In recent years, several antidiabetic drugs such ing glucose reabsorption in the kidney [8]. In T2DM, many
as thiazolidinediones and incretins, with novel mechanisms studies have confirmed that SGLT-2-i have very significant
of action, have become available, extending the number beneficial effects on glycaemic control and induce body weight
of treatment options [2]. However, some patients remain loss as well as reduce blood pressure, all without causing
uncontrolled despite the current range of old and new drugs [4]. hypoglycaemia [8]. In short, this new antidiabetic class of
Indeed, for every one of these agents, there are good and poor drugs has shown promising results in both preclinical and
responders [5-7]. Therefore, unmet needs persist in many clinical studies of T2DM [9]. In contrast to other antidiabetic
patients. Many classes of antihyperglycaemic agents are agents, SGLT-2-i are equally effective for glycaemic control,

*Corresponding author.
E-mail address: [email protected] (S. Halimi).

© 2014 Elsevier Masson SAS. All rights reserved.

© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 31/12/2014 par Vergès Bruno (26176)
 S. Halimi and B. Vergès / Diabetes & Metabolism 40 (2014) S1-S34 S29

whatever the mechanisms causing hyperglycaemia, while and FDA [16-18], while the third one, empagliflozin, has only
being independent of insulin secretion and insulin action, just been approved for use in T2DM by the EMA in May and
the magnitude of hyperglycaemia, and initial body weight FDA in August of this year [19].
and duration of diabetes, with a consequently lower risk of
hypoglycaemia. This explains why these drugs may also be
3. Is familial renal glycosuria a model?
considered a treatment for type 1 diabetes mellitus (T1DM)
patients [10]. There is a natural experimental condition known as “familial
renal glycosuria”. This disorder, also called “renal diabetes”,
represents a benign autosomal-dominant syndrome, the result of
2. Rationale to ensure the safety and tolerability of
a decrease in the number and activity of SGLT-2 transporters. It
new antidiabetic agents
is an asymptomatic condition that does not lead to renal failure
Today, the safety of new drugs, and the frequency and or other complications [20]. However, although individuals with
severity of adverse events, represent an issue of major and familial renal glycosuria do not have diabetes (hyperglycaemia),
legitimate concern as well as a source of heated contro- this does not preclude the need for long-term safety data of OADs
versy [11]. The US Food and Drug Administration (FDA), in T2DM patients. Indeed, during the therapeutic use of OADs,
followed by the European Medicine Agency (EMA), have massive glycosuria may be drug-induced in hyperglycaemic
issued new guidelines to the drugs industry to assess the patients at high cardiovascular risk and with an increased risk
cardiovascular risk of new OADs and to promote studies to of other pathological states as infections and cancer.
monitor other potentially serious adverse events, including the
increased risk of various cancers [12]. Controversies regard-
4. Hypoglycaemia
ing the effects of rosiglitazone therapy on cardiovascular
complications, pioglitazone on bladder cancer, and incretins This adverse event is seen with some glucose-lowering
on exocrine pancreatic risk, pancreatitis and pancreatic therapies and is occasionally a limitation factor in achieving
carcinoma, and even thyroid carcinoma, have raised concerns good glycaemic control. Hypoglycaemia is also associated
about the use of those drugs [11,13-15]. Beyond such major with negative effects such as undesirable symptoms like
risks, the tolerability of all drugs is another important issue weight gain (through increased appetite), poor adherence to
for patients who need to take them every day for years, as therapy, uncontrolled glycaemia because of the fear of new
in T2DM. Potential adverse events such as gastrointestinal hypoglycaemic episodes, poor quality of life and, in some
disturbances [with metformin, α-glucosidase inhibitors patients, accidents such as falls and cognitive disorders,
and glucagon-like peptide (GLP)-1 agonists], hypogly- mainly in elderly and frail patients [21-23]. A link between
caemia (with insulin, sulphonylureas and meglitinides), severe hypoglycaemia and increased cardiovascular risk and
weight gain (with insulin, sulphonylureas, meglitinides and total mortality has recently been described [21,22]. Because
thiazolidinediones) and risk of cardiovascular disease (with of their mechanism of action – which is not dependent on
thiazolidinediones) limit their use. insulin secretion and includes lowering the renal threshold for
This is why today the tolerability and safety of new drugs glucose from 200mg/dl to around 100mg/dl while maintaining
are major concerns, and why all new antidiabetic agents are glucose levels above the hypoglycaemic range – the SGLT-2-i
under the surveillance of regulatory agencies based on their are less likely to cause hypoglycaemia by themselves. This
data before marketing, from phase-II and -III studies, and means that as a monotherapy add-on to metformin (Met),
post-marketing adverse drug reactions (AEs). Also, although there is no real (or only marginal) increase in hypoglycaemia
the drug regulatory process is designed to detect adverse drug compared with a placebo or when added to other drugs that
reactions before the drug receives marketing authorization, can cause hypoglycaemia, such as sulphonylurea (SU),
for obvious reasons the premarketing detection of all potential Met + SU, pioglitazone (Pio) with or without SU and especially
adverse reactions associated with a drug may not be possible. insulin [17-19,24-28]. Thus, the incidence of hypoglycaemia
In such cases, the regulatory authorities must also react to is low except in patients taking SUs or insulin as either their
and manage adverse reactions identified at the post-marketing allocated treatment or in addition. This represents a major
stage [12]. strength of the SGLT-2-i, as it allows the use of new drug
The present report includes an analysis of the data for this combinations without increasing hypoglycaemia.
new class of drugs, the SGLT-2-i, with the knowledge that some
medium- and long-term adverse effects cannot be discussed
5. Genital infection
because this family of drugs has only been recently introduced
into the market and then in only a few countries. This review By their very nature, the most common major safety
focuses on the first three representatives of this class already concern of SGLT-2 inhibition is that the drugs cause glucose
on the market: canagliflozin, dapagliflozin and empagliflozin. levels to rise in urine, which can lead to urinary tract and
These three compounds have all reached phase-III clinical genital infections, increased urinary frequency and electrolyte
trials: two of them (canagliflozin and dapagliflozin) have imbalances. An increase in urinary glucose excretion caused
previously been approved for use in T2DM by both the EMA by SGLT-2-i treatment also has the potential to increase

© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 31/12/2014 par Vergès Bruno (26176)
 S30 S. Halimi and B. Vergès / Diabetes & Metabolism 40 (2014) S1-S34

fungal growth in the perineum and genitourinary tract. Non- gender and a history of recurrent UTIs, but not the baseline
sexually transmitted perineal and genitourinary tract mycotic level of HbA1c, were predictors of UTIs. This analysis also
infections are considered adverse events of particular interest. failed to demonstrate a definitive dose relationship between
In fact, use of the three drugs (canagliflozin, dapagliflozin glycosuria and UTI.
and empagliflozin) is accompanied by an increase in genital Comparable data have been reported for empagli-
infections compared with a placebo and affects more women flozin [19,28] and canagliflozin [16,34-35]. In FDA briefing
than men (by four to five times), mostly as vulvitis [16- data for the latter drug, the reported incidence of UTIs was
19,24-30]. However, the vast majority of cases are diagnosed similar in subjects receiving 100mg or 300mg or a placebo
with typical symptoms and require only a single standard – about 3-6% – with the majority of cases in women (87%).
antibiotic treatment [29,30]. In women, the diagnosis is mostly This incidence of UTIs was slightly higher in patients with
mycotic vulvovaginitis and, in men, mycotic balanitis. Rates moderate renal impairment: 6.2% with canagliflozin 100mg
of genital infection in men are several-fold lower than rates and 7.4% with canagliflozin 300mg vs 6% in the placebo
of vulvovaginitis in women. Genital infections are more group [29]. This incidence was increased in patients who were
frequent in premenopausal women and much more common slightly older and with a longer mean duration of diabetes
in those with a history of genital infection and/or obesity, and (12 years). Severe and upper UTIs were extremely rare. Again,
not influenced by baseline HbA1c levels. In clinical trials, the the time lag before the first UTI was highest during the first
incidence of genital infections with the maximum drug dosage 18 weeks and then declined: for canagliflozin, the median
is between 5% and 15% [24,25,27,28], and is not proportional duration was 44 and 80 days (100mg and 300mg, respectively)
to the amount of glycosuria and, thus, not related to SGLT-2-i compared with a placebo at 120 days [16]. Other commonly
doses [24,25,28]. The time interval before the first genital reported adverse events were increased urination, vulvovaginal
mycotic infection is the same for men and women: rates are pruritus, thirst, constipation and nausea.
highest in the first few months of treatment followed by an
attenuation in frequency.
7. Blood pressure
To summarize, genital mycoses are the main side-effects of
this new OAD class, affecting 5-10% of users, especially pre- Chronic osmotic diuresis caused by glycosuria would be
menopausal women. However, the infections are mostly symp- anticipated to reduce blood pressure, and dose-related increases
tomatic and respond well to standard therapy [16-19,24-37]. in 24-h urinary volumes of between 100ml and 470ml have
been reported [16-19,32-36]. Reductions in systolic blood
pressure (SBP) of up to 5mmHg have been described in
6. Urinary tract infections
trials of dapagliflozin, whether used as an add-on therapy
Urinary tract infections (UTIs) are common in patients or on its own. Canagliflozin has similarly been shown to
with T2DM. Johnsson et al. [31] pooled the safety data of significantly reduce SBP. It is well known that SGLT-2-i
12 randomized placebo-controlled trials with dapagliflozin have diuretic-like effects, lowering SBP by 3-5mmHg, which
to evaluate the relationship between glycosuria and UTIs in could benefit the majority of patients with T2DM. The precise
patients with inadequately controlled diabetes (HbA1c > 6.5- mechanism behind the BP-lowering action of SGLT-2-i,
12%). Patients were treated with dapagliflozin (2.5mg, 5mg however, is still unclear and does not appear to be based on
or 10mg) or placebo once daily either as monotherapy or natriuretic effects. Indeed, although these agents have mildly
as an add-on to Met, insulin, SU or thiazolidinedione for natriuretic effects, they are nothing like diuretics. Part of
12-24 weeks. The incidences of UTIs, which were either their BP-lowering effect is presumed to be due to osmotic
clinically diagnosed or suggested by events, were quantified diuresis [38]. However, they might also reduce BP too much
in 3 152 patients, who received once-daily dapagliflozin at in some patients, with consequences including hypotension
2.5mg (n = 814), 5mg (n = 1 145) or 10mg (n = 1 193) as (mainly postural), dizziness and dehydration [16-19]. In the
monotherapy or as add-on treatment, and 1 393 placebo- FDA briefing data for canagliflozin, these symptoms were
treated patients. For dapagliflozin at 2.5mg, 5mg, 10mg and two to three times more frequent in the treated vs placebo
placebo, the reported rates of diagnosed infections were 3.6%, groups, depending on the dose, and also much greater in older
5.7%, 4.3% and 3.7%, respectively. Although urinary glucose patients aged ≥ 75 years [34,35]. These symptoms were also
levels increased progressively in relation to dapagliflozin more frequent in patients using loop diuretics and those with
dosages, the incidence of UTIs remained stable. Most of the a glomerular filtration rate (GFR) < 60ml/min.
identified infections were considered typical for patients with
diabetes. Two-thirds of the UTIs were diagnosed from the
8. Renal effects
usual symptoms, and most of the diagnosed infections were
mild to moderate and responded to standard treatment, the Chronic kidney disease (CKD) is a major health problem
vast majority resolved by a single course of antibiotics. Drug in patients with T2DM. Stage 3-5 CKD (GFR < 60ml/min)
discontinuation due to UTIs was extremely rare, reported in affects around 25% of such patients and represents an under-
eight (0.3%) dapagliflozin-treated patients and one (0.1%) recognized problem in clinical practice. Most OADs have
placebo-treated patient [30]. Having an age > 65 years, female limitations in cases of renal impairment because they require

© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 31/12/2014 par Vergès Bruno (26176)
 S. Halimi and B. Vergès / Diabetes & Metabolism 40 (2014) S1-S34 S31

dose adjustments or are contraindicated for safety reasons. far, however, the long-term effects of SGLT-2-i-induced lipid
In fact, the activity of SGLT-2-i depends on the number of changes remain unknown, although the results of the ongoing
nephrons, which means that the first consequence of renal CV outcome trials are expected to clarify this point.
impairment on SGLT-2-i action is reduced efficacy. For this
reason, the use of SGLT-2-i is neither recommended nor allowed
10. Cardiovascular safety
in those with an estimated GFR (eGFR) < 40ml/min because
of the lack of effectiveness rather than risks. In patients with Currently, the available information on outcomes such as
reduced baseline eGFR (40-60ml/min), mild transient changes stroke, heart attack and other CV complications is limited.
in eGFR (–4 to 6ml/min), albumin-to-creatinine ratio and blood The data for dapagliflozin submitted to the US FDA are the
urea nitrogen were observed in the early phase of a study in most comprehensive so far, and include a hazard ratio of 0.67
stage-3 CKD patients. However, 26-week SGLT-2-i treatment [95% confidence interval (CI): 0.42-1.08] for a composite
resulted in a return of these parameters to baseline levels, endpoint comprising CV-related death, non-fatal stroke, non-
along with an increase in serum potassium and magnesium fatal myocardial infarction and hospitalized angina [18].
in such patients [18-19]. In canagliflozin-treated patients, the Yet, a short-term (24-week) study in T2DM patients with
24-h glucosuric effect decreased progressively in those with preexisting CV disease treated with dapagliflozin failed to
moderate-to-severe renal impairment [16,39]. In diabetic patients find any excess CV risk [42]. For this reason, several large and
with glomerular hyperfiltration, SGLT-2-i may improve GFR longer-term CV outcomes studies with SGLT-2-i are currently
and might possibly protect long-term renal function [40-41]. ongoing [44-47]. However, looking at the preliminary data
from phase-II/-III studies of dapagliflozin, and the phase-
II/-III studies [18] and intermediate data from CANVAS
9. Lipoproteins
(Canagliflozin Cardiovascular Assessment Study), there were
The cardiovascular (CV) safety of antidiabetic agents no signs of CV harm during the first 18 months [45]. In all
represents a major issue because of the well-established high phase-II/-III studies, only one specific indicator was found
CV risk in T2DM subjects. The FDA has imposed strict rules for fatal and non-fatal stroke with canagliflozin, but it was not
for the potentially increased CV risk of every new antidiabetic significant [relative risk (RR): 1.47 (95% CI: 0.83-2.59)]. Is
agent based on premarketing, phase-II and phase-III trials, the transient hypotension observed with the drug a possible
and post-marketing study data [12]. Yet, all SGLT-2-i raise explanation for this? This was not confirmed by the time course
low-density lipoprotein cholesterol (LDL-C), thereby raising of the events. Some data suggest an early increase (within the
doubt concerning the CV risk induced by such increases. first 30 days) for CV events, while MACE-plus analysis in
Thus, in T2DM patients with a baseline LDL-C between 90 CANVAS vs controls revealed 13 vs one events (two-to-one
and 110mg/dl, the increase is about 5% with dapagliflozin randomization, n = 2 886/1 441 patients).
10mg, and 2.4% and 3.1% with empagliflozin 10mg and These data represent a very small and non-significant
25mg, respectively [18]. The magnitude of the increase seems number of patients and, at this time, CV risk is not considered
slightly higher with canagliflozin, with mean increases in a serious hazard with these drugs [45]. Nevertheless, only
LDL-C of 4-5% with 100mg/day [16,42]. long-term trials specifically designed to answer the question
On the other hand, a small increase in high-density will give clear information in future. Thus, larger studies with
lipoprotein cholesterol (HDL-C) is consistently reported CV endpoints are currently ongoing and expected to provide
with all SGLT-2-i [16-19]. With canagliflozin monotherapy, data from 2017 onwards. So far, CANVAS I (NCT01032629)
HDL-C was increased by 6.8% and 6.1% in the 100mg/day has recruited > 4 000 T2DM patients who are at increased CVD
and 300mg/day treatment groups, respectively, compared risk [45], while the Empagliflozin Cardiovascular Outcome
with a placebo (p < 0.001 and p < 0.01, respectively) [42]. Event Trial (EMPA-REG-OUTCOMES, NCT01131676) has
With dapagliflozin 10mg, a mean 6.3% HDL-C increase has recruited an estimated 7 000 patients to date [46] and the
been reported [43]. It has sometimes been postulated that the Dapagliflozin Effect on Cardiovascular Events trial (DECLARE,
detrimental effects of the LDL-C increase induced by SGLT-2-i NCT01730534) has only just begun recruiting participants [47].
might be counterbalanced by the parallel HDL-C increase,
leading to little or no change in the LDL-C/HDL-C ratio in
11. Drug interactions
most clinical trials. However, these findings have to be viewed
with caution, as HDL-C particles are dysfunctional in T2DM UDP-glucuronosyltransferase (UGT) inducers (such as
patients – there is a marked reduction in their antiatherogenic rifampin, phenytoin, phenobarbital and ritonavir) increase
effects – and it is still not clear whether the slight HDL-C the metabolism of canagliflozin, thereby decreasing active
increase observed with SGLT-2-i is truly cardioprotective. canagliflozin levels in the blood. Thus, doses of canagliflozin
Serum triglyceride levels have been mildly reduced need to be increased from 100mg to 300mg in such patients.
in several studies with SGLT-2-i [16–19]. A mean 5.2% On the other hand, canagliflozin increases the area under
decrease in plasma triglycerides was reported with dapagli- the curve (AUC) for digoxin and, thus, patients receiving
flozin 10mg [43]. With canagliflozin 300mg, mean plasma digoxin treatment should be monitored [16]. On the other hand,
triglyceride levels decreased by 7.5% to 16% [18,35]. Thus there were no clinically meaningful drug – drug interactions

© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 31/12/2014 par Vergès Bruno (26176)
 S32 S. Halimi and B. Vergès / Diabetes & Metabolism 40 (2014) S1-S34

15. Bone effects and fractures

for dapagliflozin observed with several medications used
concomitantly by T2DM patients, including Met, sitagliptin,
digoxin, simvastatin and warfarin [48,49]. Owing to its mechanism of action, SGLT-2-i could poten-
tially affect renal tubular transport of bone minerals [55].
For this reason, bone turnover has been studied for several
12. Uric acid levels
SGLT-2-i. Dapagliflozin had no effect on markers of bone
SGLT-2-i have been reported to lower serum uric acid formation and resorption or on bone mineral density (BMD)
(SUA) levels [50]. This is related to the subsequent increased after 50 weeks of treatment in both male and postmenopausal
urinary excretion of SUA most probably due to glycosuria. female T2DM patients, who were inadequately controlled by
Met alone [55]. Bone formation was measured by procol-
lagen type 1 N-terminal propeptide (P1NP) and resorption
13. Liver effects
by C-terminal cross-linking telopeptides of type 1 collagen
Mean exposure to dapagliflozin in patients with moderate (CTX), while BMD was assessed by standardized dual-energy
and severe hepatic impairment was 36% and 67% higher, X-ray absorptiometry (DXA) measurements. In a study where
respectively, compared with healthy subjects with no liver BMD was followed by DXA in older T2DM patients receiving
issues. However, no pharmacokinetic interactions were found canagliflozin treatment for 52 weeks (DIA3010), a very small,
between dapagliflozin and simvastatin, valsartan, warfarin non-clinically relevant BMD decrease was measured and was
or digoxin [48,49]. The possible hepatic adverse effects of probably due to weight loss [55].
canagliflozin, described in a report submitted to the FDA at As for fractures, the number seen with dapagliflozin was
the end of 2012 [51], have not been confirmed so far. extremely small and similar to that reported for placebo.
When large sets of data were pooled for patients receiving
canagliflozin (n = 6177), there was a small increase in adju-
14. SGLT-2 inhibitors and malignancy
dicated fractures, with an incidence rate/1 000 person-years
The potential relationship between SGLT-2-i and cancer is of exposure of 12.28 ± 1.24 for canagliflozin vs 9.44 ± 1.55
also under investigation. Cancers are more frequently seen in in the non-canagliflozin group (FDA Broad Dataset, July
diabetic patients, and some antidiabetic agents, such as Pio, and 2012). This discrepancy between normal formation/resorption
GLP-1 drugs for pancreatitis have been suspected of increasing markers and BMD evolution, and the fact that fractures were
their incidence, although this has not been confirmed by either recorded early after the initiation of treatment has yet to be
studies [13-15] or regulatory agencies [52]. The overall propor- clarified, which means that, so far, no definitive conclusions
tion of patients with cancers or undetermined tumours is similar can be drawn [55,56]. The potential excess number of fractures
in TD2M patients treated with dapagliflozin compared with a was mostly limited to the upper extremities, making it highly
placebo or comparator agent (1.43% vs 1.30%, respectively). unusual for a bone-fragility issue and suggesting that falls
However, breast and bladder cancers are somewhat more may be a factor, although again the information submitted so
commonly seen with dapagliflozin. The FDA submission far is not sufficient for drawing any conclusions [56]. Further
file reported nine breast cancers in 4 287 patients receiving data on BMD and fractures are forthcoming and are needed
dapagliflozin vs no cases in 1 941 patients taking a placebo to confirm whether there is a safety issue related to fractures
or comparator, and seven bladder cancers in 4 310 patients and a putative link with falls.
taking dapagliflozin compared with no cases in 1 962 patients
taking a placebo or comparator [53]. However, it should be
16. In summary
noted that haematuria was documented before exposure to
dapagliflozin in four of the seven patients later found to have Based on the currently available data, the overall safety
bladder cancer, and the women with breast cancer had been of the three SGLT-2-i discussed in this review is relatively
taking dapagliflozin for < 1 year (two of the nine cases were good. Monami et al. [57] recently published a meta-analysis
diagnosed within 6 weeks of starting dapagliflozin treatment). of 31 trials (22 trials vs placebo, five vs metformin, one
The incidence of breast or bladder tumour events was low vs glipizide and three vs sitaglitpin), involving a total of
for canagliflozin and had a similar rate vs non-canagliflozin- 7 524 patients taking SGLT-2-i and 3 628 taking a placebo
treated patients [16]. A pooled analysis of nine trials with or comparator agent, for efficacy and adverse events (AEs).
approximately 8 000 person-years of exposure showed no In their review as in our present study, these new antidiabetic
differences in the incidence of bladder cancer between the agents were mostly found to induce a significant increase
canagliflozin (five out of 6 648 patients) and control (four in genital mycotic infections [Mantel-Haenszel odds ratio
out of 3 640 patients) groups [34]. No data for cancer cases (MH-OR): 3.34-6.04] that was more frequent in women and
with other SGLT-2-i studies are currently available. The US especially those who were premenopausal, with only a slight
FDA in its review of dapagliflozin noted a non-significant increase in men. The risk of UTIs was consistent but very
excess number of breast and bladder cancers, with the latter small; found in all trials vs comparators or a placebo, it was
potentially due to ascertainment bias related to the treatment only statistically significant in the single study vs glipizide
of drug-related urinary symptoms [54]. (p < 0.025). Predictors of these two infectious syndromes

© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 31/12/2014 par Vergès Bruno (26176)
 S. Halimi and B. Vergès / Diabetes & Metabolism 40 (2014) S1-S34 S33

are a history of such recurrent infections, female gender American Diabetes Association (ADA) and the European Association
and age > 65 years, but not baseline HbA1c levels, amount for the Study of Diabetes (EASD). Diabetes Care. 2012;35:1364-79.
[3] Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA. 10-year
of glycosuria or SGLT-2-i dose. For all serious AEs, the follow-up of intensive glucose control in type 2 diabetes. N Engl J
MH-OR was 0.74-1.07 (no risk) and, for total hypoglycaemia, Med 2008;359: 577-89.
the MH-OR was 0.74 vs Met (not significant, NS), 0.73 [4] Ong KL, Cheung BM, Wong LY, Wat NM, Tan KC, Lam KS. Prevalence,
vs sitagliptin (NS) and 0.05 vs glipizide (p < 0.01). For treatment, and control of diagnosed diabetes in the U.S. National
Health and Nutrition Examination Survey 1999-2004. Ann Epidemiol
hypoglycaemia vs placebo, the MH-OR was 1.34 (p < 0.01), 2008;18:222-9.
which is usually found with a placebo (as with Met and [5] Hsieh MC, Lin KD, Tien KJ, Tu ST, Hsiao JY, Chang SJ, et al.
sitagliptin) and not clinically meaningful. Common polymorphisms of the peroxisome proliferator-activated
receptor-gamma (Pro12Ala) and peroxisome proliferator-activated
receptor-gamma coactivator-1 (Gly482Ser) and the response to
17. Conclusion pioglitazone in Chinese patients with type 2 diabetes mellitus.
Metabolism. 2010;59:1139-44.
Apart from genital mycotic infections and urinary infec- [6] Oh TJ, Jung HS, Bae JH, Kim YG, Park KS, Cho YM, et al. Clinical
tions that are somewhat frequent, but usually mild and rarely characteristics of the responders to dipeptidyl peptidase-4 inhibitors in
Korean subjects with type 2 diabetes. J Korean Med Sci 2013;28:881-7.
leading to drug discontinuation, SGLT-2-i appear to be well [7] Anderson SL, Trujillo JM, McDermott M, Saseen JJ. Determining
tolerated. However, at this time, they are not recommended predictors of response to exenatide in type 2 diabetes. J Am Pharm
for patients with a history of recurrent urinary and/or genital Assoc (2003) 2012;52:466-71.
infections. The long-term CV tolerability is being monitored [8] Bays H. Sodium Glucose Co-transporter Type 2 (SGLT-2) Inhibitors:
Targeting the Kidney to Improve Glycemic Control in Diabetes Mellitus.
in several ongoing trials, as justified by the LDL-C elevation, Diabetes Ther 2013;4:195-220.
transient hypotension and volume depletion seen with all [9] Abdul-Ghani MA, Defronzo RA. Lowering Plasma Glucose
SGLT-2-i. On the other hand, potentially beneficial effects Concentration by Inhibiting Renal Sodium-Glucose Co-Transport.
are expected with favourable effects on BP, body weight, J Intern Med 2014 Apr 1. doi: 10.1111/joim.12244. [Epub ahead of
print] PubMed PMID: 24690096.
triglycerides and HDL-C. As for malignancies, the only risk [10] Perkins BA, Cherney DZ, Partridge H, Soleymanlou N, Tschirhart
found was with dapagliflozin for bladder cancer, which merits H, Zinman B, et al. Sodium-glucose cotransporter 2 inhibition and
follow-up, although it appears unlikely because of its early glycemic control in type 1 diabetes: results of an 8-week open-label
appearance after starting the drug. The need for adequate proof-of-concept trial. Diabetes Care 2014;37:1480-3.
[11] Hiatt WR, Kaul S, Smith RJ. The cardiovascular safety of diabetes
renal function is well documented, although the mode of drugs--insights from the rosiglitazone experience. N Engl J Med
action in SGLT-2 transporters in the proximal tubule should 2013;369:1285-7.
not aggravate any existing damage in the glomerulus, and [12] http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulato-
may even offer benefits through less glucotoxicity, lower ryInformation/Guidances/UCM071627.pdf
[13] Monami M, Dicembrini I, Mannucci E. Thiazolidinediones and cancer:
SBP and reduced proteinuria. Indeed, SGLT-2-i at all doses results of a meta-analysis of randomized clinical trials. Acta Diabetol
have shown no association with serious major AEs and, so 2014;51:91-101.
far, appear to be a safe new class of OADs for treating T2DM, [14] Monami M, Dicembrini I, Martelli D, Mannucci E. Safety of dipeptidyl
particularly because of significant improvements in HbA1c, peptidase-4 inhibitors: a meta-analysis of randomized clinical trials.
Curr Med Res Opin. 2011;27:57-64.
body weight and BP without hypoglycaemia, thus contributing [15] Alves C, Batel-Marques F, Macedo AF. A meta-analysis of serious
to the management of T2DM patients by offering new oral adverse events reported with exenatide and liraglutide: acute pancreatitis
drug combinations. and cancer. Diabetes Res Clin Pract 2012;98:271-84.
[16] FDA.http://www.fda.gov/downloads/AdvisoryCommittees/Com-
mitteesMeetingMaterials/Drugs/EndocrinologicandMetabolicDrugsAd-
Disclosure of interests visoryCommittee/UCM334550.pdf, 2013 Jan 23.
[17] FDA.http://www.fda.gov/downloads/AdvisoryCommittees/Com-
S. H. has received grants and honoraria for lectures at mitteesMeetingMaterials/Drugs/EndocrinologicandMetabolicDrug-
scientific meetings sponsored or arranged by AstraZeneca, sAdvisoryCommittee/UCM378076.pdf, 2013 Dec.
[18] Janssen Pharmaceuticals Inc. InvokanaTM (canagliflozin) tablets, for
Bristol-Myers Squibb, Boehringer Ingelheim, Merck oral use: US prescribing information. Available from: http://www.
Sharp & Dohme, Novartis, Novo Nordisk, Eli Lilly, Roche, janssenmd.com/pdf/ invokana/PI-INVOKANA.pdf
Johnson & Johnson, Sanofi-Aventis and Takeda. [19] FDA.http://www.drugs.com/newdrugs/fda-approves-jardiance-empa-
B. V. has received honoraria for expertise and lectures from gliflozin-type-2-diabetes-4064.html
[20] Ly JP, Onay T, Sison K, Sivaskandarajah G, Sabbisetti V, Li L, et al.
AstraZeneca/Bristol-Myers Squibb, Bayer, Janssen, Merck The Sweet Pee model for SGLT-2 mutation. J Am Soc Nephrol
Sharp & Dohme, Novo Nordisk, Novartis and Sanofi-Aventis. 2011;22;113-23.
[21] Gerstein HC, Miller ME, Byington RP, Goff DC Jr, Bigger JT, Buse
References JB, et al. Action to Control Cardiovascular Risk in Diabetes Study
Group. Effects of intensive glucose lowering in type 2 diabetes. N
[1] Whiting DR, Guariguata L, Weil C, Shaw J. IDF Diabetes Atlas: Global Engl J Med 2008;358:2545-59.
estimates of the prevalence of diabetes for 2011 and 2030. Diabetes [22] Duckworth W, Abraira C, Moritz T, Reda D, Emanuele N, Reaven PD,
Res Clin Pract 2011;94:311-21. et al. VADT Investigators. Glucose control and vascular complications
[2] Inzucchi SE, Bergenstal RM, Buse JB, Diamant M, Ferrannini E, Nauck in veterans with type 2 diabetes. N Engl J Med 2009;360:129-39.
M, et al. American Diabetes Association (ADA); European Association [23] Lundkvist J, Berne C, Bolinder B, Jonsson L. The economic
for the Study of Diabetes (EASD). Management of hyperglycemia in and quality of life impact of hypoglycemia. Eur J Health Econ
type 2 diabetes: a patient-centered approach: position statementof the 2005;6:197-202.

© 2014 Elsevier Masson SAS. Tous droits réservés. - Document téléchargé le 31/12/2014 par Vergès Bruno (26176)
 S34 S. Halimi and B. Vergès / Diabetes & Metabolism 40 (2014) S1-S34

[24] Kaku K, Inoue S, Matsuoka O, Kiyosue A, Azuma H, Hayashi N, tubular cells – renoprotection in diabetic nephropathy? PLoS One
et al: Efficacy and safety of dapagliflozin as a monotherapy for type 2 2013;8:e54442.
diabetes mellitus in Japanese patients with inadequate glycaemic control: [41] Stanton RC. Sodium glucose transport 2 (SGLT-2) inhibition decreases
a phase II multicentre, randomized, double-blind, placebo-controlled glomerular hyperfiltration: is there a role for SGLT-2 inhibitors in
trial. Diabetes Obes Metab 2013;15:432-40. diabetic kidney disease? Circulation 2014;129:542-4.
[25] Bhatia J, Gamad N, Bharti S, Arya DS. Canagliflozin-current status [42] Lamos EM, Younk LM, Davis SN. Canagliflozin, an inhibitor of
in the treatment of type 2 diabetes mellitus with focus on clinical trial sodium-glucose cotransporter 2, for the treatment of type 2 diabetes
data. World J Diabetes 2014;5:399-406. mellitus. Expert Opin Drug Metab Toxicol 2013;9:763-75.
[26] Kaushal S, Singh H, Thangaraju P, Singh J. Canagliflozin: A Novel [43] Dziuba J, Alperin P, Racketa J, Iloeje U, Goswami D, Hardy E,
SGLT-2 Inhibitor for Type 2 Diabetes Mellitus. N Am J Med Sci et al. Modeling effects of SGLT-2 inhibitor dapagliflozin treatment
2014;6:107-13. versus standard diabetes therapy on cardiovascular and microvascular
[27] Cefalu WT, Leiter LA, Yoon KH, Arias P, Niskanen L, Xie J, et al. outcomes. Diabetes Obes Metab 2014;16:628-35.
Efficacy and safety of canagliflozin versus glimepiride in patients [44] Leiter LA, Cefalu WT, de Bruin TW, Gause-Nilsson I, Sugg J, Parikh
with type  2 diabetes inadequately controlled with metformin SJ. Dapagliflozin added to usual care in individuals with type 2 diabetes
(CANTATA-SU): 52 week results from a randomised, double-blind, mellitus with preexisting cardiovascular disease: a 24-week, multicenter,
phase 3 non-inferiority trial. Lancet 2013;382:941-50. randomized, double-blind, placebo-controlled study with a 28-week
[28] Rosenstock J, Seman LJ, Jelaska A, Hantel S, Pinnetti S, Hach extension. J Am Geriatr Soc 2014;62:1252-62.
T, et al. Efficacy and safety of empagliflozin, a sodium glucose [45] Neal B, Perkovic V, de Zeeuw D, Mahaffey KW, Fulcher G, Stein P,
cotransporter 2 (SGLT-2) inhibitor, as add-on to metformin in et al. Rationale, design, and baseline characteristics of the Canagliflozin
type 2 diabetes with mild hyperglycaemia. Diabetes Obes Metab Cardiovascular Assessment Study (CANVAS)-a randomized placebo-
2013;15:1154-60. controlled trial. Am Heart J 2013;166:217-23.
[29] Johnsson KM, Ptaszynska A, Schmitz B, Sugg J, Parikh SJ, List JF. [46] Ridderstråle M, Svaerd R, Zeller C, Kim G, Woerle HJ, Broedl UC;
Vulvovaginitis and balanitis in patients with diabetes treated with EMPA-REG H2H-SU trial investigators. Rationale, design and baseline
dapagliflozin. J Diabetes Complications 2013;27:479-84. characteristics of a 4-year (208-week) phase III trial of empagliflozin,
[30] Nyirjesy P, Sobel JD, Fung A, Mayer C, Capuano G, Ways K, et al. an SGLT-2 inhibitor, versus glimepiride as add-on to metformin in
Genital mycotic infections with canagliflozin, a sodium glucose patients with type 2 diabetes mellitus with insufficient glycemic
co-transporter 2 inhibitor, in patients with type 2 diabetes mellitus: a control. Cardiovasc Diabetol 2013 5;12:129.
pooled analysis of clinical studies. Curr Med Res Opin 2014;30:1109-19. [47] DECLARE Trial. Dapaglifozin. http://clinicaltrials.gov/show/
[31] Johnsson KM, Ptaszynska A, Schmitz B, Sugg J, Parikh SJ, List NCT01730534
JF. Urinary tract infections in patients with diabetes treated with [48] Kasichayanula S, Chang M, Liu X, Shyu WC, Griffen SC, LaCreta FP,
dapagliflozin. J Diabetes Complications 013;27:473-8. et al. Lack of pharmacokinetic interactions between dapagliflozin and
[32] Rosenstock J, Vico M, Wei L, Salsali A, List JF. Effects of dapagliflozin, simvastatin, valsartan, warfarin, or digoxin. Adv Ther 2012;29:163-77.
an SGLT-2 inhibitor, on HbA(1c), body weight, and hypoglycemia risk [49] Food and Drug Administration. FDA Briefing Document NDA 202293.
in patients with type 2 diabetes inadequately controlled on pioglitazone Dapagliflozin tablets 5 and 10mg. In: Products DoMae (ed). Silver
monotherapy. Diabetes Care 2012;35:1473-8. Spring: Food and Drug Administration 2011.
[33] Kovacs CS, Seshiah V, Swallow R, Jones R, Rattunde H, Woerle [50] Chino Y, Samukawa Y, Sakai S, Nakai Y, Yamaguchi JI, Nakanishi T,
HJ, et al. On behalf of the EMPA-REG PIO™ trial investigators. et al. SGLT-2 inhibitor lowers serum uric acid through alteration of uric
Empagliflozin improves glycaemic and weight control as add-on acid transport activity in renal tubule by increased glycosuria. Biopharm
therapy to pioglitazone or pioglitazone plus metformin in patients Drug Dispos. 2014 Jul 7. doi: 10.1002/bdd.1909 [Epub ahead of print].
with type 2 diabetes: a 24-week, randomized, placebo-controlled [51] John Senior and Leonard Seeff to Mary Parks. Possible hepatic adverse
trial. Diabetes Obes Metab. 2013 Aug 1. doi:10.1111/dom. 12188. effects of canagliflozin. 28 November 2012. http://www.accessdata.
[Epub ahead of print] fda.gov/drugsatfda_ docs/nda/2013/204042Orig1s000OtherR.pdf
[34] Schernthaner G, Gross JL, Rosenstock J, Guarisco M, Fu M, Yee [52] Committee for Medicinal Products for Human Use (CHMP) of the
J, et al. Canagliflozin compared with sitagliptin for patients with European Medicines Agency (EMA). Investigation into GLP-1 based
type 2 diabetes who do not have adequate glycemic control with diabetes therapies concluded. No new concerns for GLP-1 therapies
metformin plus sulfonylurea: a 52-week randomized trial. Diabetes identified on the basis of available evidence. 26 July 2013. http://www.
Care 2013;36:2508-15 [Erratum in: Diabetes Care 2013;36:4172]. ema.europa.eu/ema/index
[35] Rosenstock J, Aggarwal N, Polidori D, Zhao Y, Arbit D, Usiskin K, [53] FDA.http://www.fda.gov/downloads/AdvisoryCommittees/Com-
et al. Canagliflozin DIA 2001 Study Group. Dose-ranging effects mitteesMeetingMaterials/Drugs/EndocrinologicandMetabolicDrug-
of canagliflozin, a sodium-glucose cotransporter 2 inhibitor, as sAdvisoryCommittee/UCM378079
add-on to metformin in subjects with type 2 diabetes. Diabetes Care [54] Bolinder J, Ljunggren O, Kullberg J, Johansson L, Wilding J, Langkilde
2012;35:1232-8. AM, et al. Effects of dapagliflozin on body weight, total fat mass, and
[36] List JF, Woo V, Morales E, Tang W, Fiedorek FT. Sodium-glucose regional adipose tissue distribution in patients with type 2 diabetes
cotransport inhibition with dapagliflozin in type 2 diabetes. Diabetes mellitus with inadequate glycemic control on metformin. J Clin
Care 2009;32:650-7. Endocrinol Metab 2012;97:1020-31.
[37] Nauck M, Prato SD, Rohwedder K, Elze M, Parikh S. Dapagliflozin [55] Ljunggren Ö, Bolinder J, Johansson L, Wilding J, Langkilde AM,
vs glipizide in patients with type 2 diabetes mellitus inadequately Sjöström CD, et al. Dapagliflozin has no effect on markers of bone
controlled on metformin: 52-weekresults of a double-blind, randomised formation and resorption or bone mineral density in patients with
controlled trial. Diabetologia 2010;53:S1-556. inadequately controlled type 2 diabetes mellitus on metformin. Diabetes
[38] Oliva RV, Bakris GL. Blood pressure effects of sodium-glucose Obes Metab 2012;14:990-9.
co-transport 2 (SGLT-2) inhibitors. J Am Soc Hypertens 2014;8:330-9. [56] FDA.http://www.fda.gov/downloads/AdvisoryCommittees/Com-
[39] Yale JF, Bakris G, Cariou B, Yue D, David-Neto E, Xi L, et al. Efficacy mitteesMeetingMaterials/Drugs/EndocrinologicandMetabolicDrug-
and safety of canagliflozin in subjects with type 2 diabetes and chronic sAdvisoryCommittee/UCM336236. pdf, 2013 Jan 23.
kidney disease. Diabetes Obes Metab 2013;15:463-73. [57] Monami M, Nardini C, Mannucci E. Efficacy and safety of sodium
[40] Panchapakesan U, Pegg K, Gross S, Komala MG, Mudaliar H, Forbes glucose co-transport-2 inhibitors in type 2 diabetes: a meta-analysis
J, et al. Effects of SGLT-2 inhibition in human kidney proximal of randomized clinical trials. Diabetes Obes Metab 2014;16:457-66.

© 2014 Elsevier Masson SAS.

View publication statsTous droits réservés. - Document téléchargé le 31/12/2014 par Vergès Bruno (26176)