Efficacy and safety of dapagliflozin, a sodium glucose cotransporter 2 (SGLT2) inhibitor, in diabetes mellitus

Dapagliflozin is indicated for patients aged 18 years and over with T2DM for use as
monotherapy to improve glycemic control in patients with inadequate glycemic control
who are intolerant to metformin, and as an adjunct to diet and exercise in combination
with other glucose-lowering medicinal agents in patients inadequately controlled on
existing antidiabetic medications, including insulin 36].

In patients with moderate renal impairment [creatinine clearance 60 mL/min or estimated
glomerular filtration rate (eGFR) 60 mL/min/1.73 m
²
] dapagliflozin has been shown to be less effective. Therefore, dapagliflozin is not
recommended for use in patients with moderate to severe renal impairment. Such limitations
have not been introduced for the presence of side effects, but in recognition of reduced
efficacy in this patient population.

The efficacy and safety of dapagliflozin as monotherapy and as add-on/combination
therapy with existing antidiabetic treatment in patients with diabetes mellitus has
been established in a series of phase II and III trials in the clinical development
program and is supported by five recently published comprehensive reviews or meta-analyses
37]–41]. It should be noted that, while different dosages of dapagliflozin were evaluated
in the above-mentioned trials, the only dosage currently approved in Europe is 10 mg
in oral administration once daily.

Efficacy as monotherapy

Dapagliflozin has been shown to be effective in several large (n  261), randomized, double-blind, placebo-controlled studies of 12 or 24 weeks’ duration
in treatment-naïve patients with T2DM 25], 42]–46]. Dapagliflozin was evaluated at doses ranging from 1 to 50 mg once daily when added
to background diet and exercise 42]–44]; compared with metformin monotherapy 45], 47]; or administered as monotherapy in both treatment-experienced and treatment-naïve
patients 25].

Dapagliflozin monotherapy demonstrated dose-related reductions from baseline in glycosylated
hemoglobin (HbA
_1c
) in all dapagliflozin groups, ranging from mean reductions of ?0.55 to ?1.45 % overall
25], 42]–46] (Table 1). Statistically significant reductions in fasting plasma glucose (FPG) and body weight
for dapagliflozin versus placebo or metformin were observed. Hypoglycemia was uncommon
in dapagliflozin-treated patients; genital and urinary tract infections were more
common in dapagliflozin groups than in controls.

Table 1. Glycemic efficacy of dapagliflozin when used as monotherapy in patients with type
2 diabetes mellitus

Dapagliflozin monotherapy was generally as effective as metformin monotherapy 47], 48], and the glycemic efficacy and reduction in weight of dapagliflozin monotherapy seen
in short-term studies is maintained in the long term (52–102 weeks) 33], 48].

In addition, the data from the studies in Asian patients with T2DM suggest that dapagliflozin
is effective in that population 25], 33], 48]. More recent data support similar efficacy of dapagliflozin in white, black and Hispanic
patients, suggesting its use may be beneficial irrespective of race 49].

Efficacy as add-on or combination therapy

A number of randomized, double-blind, placebo-controlled studies in patients with
T2DM have evaluated the efficacy of dapagliflozin as adjunctive therapy with metformin
28], 47], 50]–57] and insulin 58]–61].

Glycemic efficacy and body weight data from key studies of dapagliflozin add-on/combination
therapy with metformin or insulin are presented in Table 2. Dapagliflozin is associated with clinically and statistically significant improvements
in glycemic control and reductions in body weight, compared with placebo, when added
to concurrent metformin or insulin therapy. Mean adjusted changes from baseline HbA
_1c
ranged from ?0.39 to ?0.96 % in the primary studies, compared with +0.02 to ?0.39
for placebo (Table 2). Mean adjusted changes from baseline body weight with dapagliflozin ranged from
?0.90 to ?5.05 kg in the primary studies, compared with +0.4 to ?1.55 kg for placebo
(Table 2). Changes in body weight associated with dapagliflozin primarily result from changes
in fat mass (mediated by glycosuria), rather than fluid: according to Bolinder et
al., after 102 weeks reductions in total body fat mass were ?2.80 kg (95 % CI ?3.67,
?1.93) in the dapagliflozin group, compared with ?1.46 kg (95 % CI ?2.25, ?0.68) in
the placebo group 52].

Table 2. Glycemic efficacy of dapagliflozin when used in combination therapy in patients with
type 2 diabetes mellitus

In addition to placebo-controlled trials, dapagliflozin was compared with glipizide
as add-on therapy to concurrent metformin 53], 54], 56]. At 52 weeks, glycemic efficacy was similar with dapagliflozin (change from baseline
HbA
_1c
?0.52 %) and glipizide (?0.52 %), but dapagliflozin produced significant benefits
in body weight reduction (?3.2 kg versus +1.2 kg, respectively; P  0.0001), proportion
of patients achieving ?5 % body weight reduction (33.3 versus 2.5 %, respectively;
P  0.0001) and proportion of patients experiencing hypoglycemia (3.5 versus 40.8 %,
respectively; P  0.0001) 53]. Long-term data from the extension of this study to 104 weeks 54] and 208 weeks 56] showed that, compared with glipizide, dapagliflozin is associated with sustained
glycemic efficacy, greater reductions in body weight and systolic blood pressure,
and lower frequency of hypoglycemia (Table 2).

Dapagliflozin as an add-on or in combination with other antidiabetic agents also consistently
produced reductions in blood pressure 37], 39], 40].

The results of longer-term extension studies demonstrate that the glycemic and body
weight improvements with dapagliflozin are maintained after up to 4 years of follow-up.
Mean adjusted changes from baseline HbA
_1c
at the end of the extension studies ranged from ?0.30 to ?0.82 % with dapagliflozin,
compared with ?0.10 to ?0.43 for placebo (Table 2). Meanwhile, mean adjusted changes from baseline body weight ranged from ?1.0 to
?4.54 kg with dapagliflozin, compared with +1.8 to ?2.12 kg for placebo (Table 2).

The effects of dapagliflozin have also been evaluated in add-on or combination with
the dipeptidyl peptidase-4 inhibitor sitagliptin, the sulfonylurea glimepiride, and
the thiazolidinedione pioglitazone. In patients inadequately controlled on sitagliptin
with or without metformin, add-on therapy with dapagliflozin 10 mg provided additional
clinical benefit without increasing hypoglycemia events 62].

In patients with inadequate glycemic control with glimepiride, dapagliflozin (at the
5 or 10 mg dose) significantly improved HbA
_1c
and significantly reduced body weight, compared with glimepiride alone 63]. At 24 weeks, the mean adjusted changes from baseline HbA
_1c
were ?0.13 % for placebo versus ?0.63 % with dapagliflozin 5 mg and ?0.82 % with dapagliflozin
10 mg, respectively (both P  0.0001 vs. placebo). Corresponding mean adjusted changes
from baseline body weight were ?0.72, ?1.56 and ?2.26 kg, respectively, for placebo,
dapagliflozin 5 mg (P  0.001 vs. placebo), and 10 mg (P  0.0001 vs. placebo), respectively.
There was a higher incidence of hypoglycemia in the dapagliflozin group (7.1–7.9 %
vs. 4.8 %, respectively), as has been observed when dapagliflozin is added to sulfonylureas
in other studies, and no patient discontinued treatment because of hypoglycemia.

Finally, in patients not adequately controlled on pioglitazone, dapagliflozin further
lowered HbA
_1c
and lessened pioglitazone-associated weight gain 64]. At 24 weeks, the mean adjusted changes from baseline HbA
_1c
were ?0.42 % for placebo versus ?0.82 % with dapagliflozin 5 mg (P  0.001 vs. placebo)
and ?0.97 % with dapagliflozin 10 mg (P  0.0001 vs. placebo) respectively. Patients
in the pioglitazone alone group gained significantly more body weight than those in
the pioglitazone plus dapagliflozin groups. At 24 weeks, the mean adjusted changes
from baseline body weight were +1.64, +0.09 and ?0.14 kg, respectively, for placebo,
dapagliflozin 5 mg, and dapagliflozin 10 mg (both P  0.001 vs. placebo). By 48 weeks,
patients in the pioglitazone alone group had gained a mean of 2.99 kg from baseline,
compared with 1.35 kg for dapagliflozin 5 mg, and 0.69 kg for dapagliflozin 10 mg,
respectively. However, dapagliflozin is not recommended for use in patients concomitantly
treated with pioglitazone 65].

Longer-term extension phases of these trials demonstrate that reductions in HbA
_1c
, FPG, and bodyweight are maintained during follow-up periods of up to 4 years 50], 52], 53], 56], 58], 61]–64], 66].

A recent meta-analysis designed to evaluate whether dapagliflozin is synergistic with
other antidiabetic agents without affecting body weight concluded that dapagliflozin
in combination with conventional antidiabetic drugs (metformin, glimepiride, pioglitazone,
and metformin/sitagliptin) improved glycemic control (the overall effect size was
?0.52 %) and reduced weight gain in patients with T2DM (the effect size was ?2.10 kg)
39]. Twelve randomized controlled trials with a total of 3986 participants were included
in the glycemic control analysis (1996 dapagliflozin; 1990 controls), and 4008 in
the body weight analysis (2005 dapagliflozin; 2003 controls). Follow-up durations
ranged from 12 to 208 weeks 56].

Of interest, dapagliflozin 10 mg treatment significantly improved glycemic control
and reduced body weight both in 151 early-stage and in 58 late-stage patients with
T2DM, reflecting the usefulness of dapagliflozin as monotherapy in patients in the
early stage of T2DM, and as add-on or combination therapy in late-stage patients on
high doses of insulin plus oral insulin sensitizers 60].

To date, dapagliflozin in combination with glucagon-like peptide 1 (GLP-1) analogs
is still being studied (https://clinicaltrials.gov/ct2/show/NCT02229396).

Dapagliflozin and ?-cell function

Improvements in ?-cell function have been demonstrated for dapagliflozin using the
homeostasis model assessment of ?-cell function (HOMA ?-cell) 40], 65], possibly resulting from a reduction in glucose toxicity. This hypothesis has been
examined in a small study that demonstrated that lowering of plasma glucose concentration
by dapagliflozin-induced glycosuria improved ?-cell function and insulin resistance
in patients with T2DM 13]. Importantly, these results demonstrate that the glucotoxic effect of chronic hyperglycemia
on ?-cell function in T2DM is, at least in part, reversible 13]. The authors theorized that, as dapagliflozin has not been shown to act directly
on ?-cell function, improvement in ?-cell function was related to amelioration of
hyperglycemia, that is, by reversing glucose toxicity 13]. Further investigation is warranted.

Nephroprotection

In addition to improved glycemic control, reductions in serum uric acid levels and
tubular glucose toxicity and attenuation of diabetes-related hyperfiltration suggest
that SGLT2 inhibitors may be able to influence renal hemodynamics independently of
glucose reduction 27], 40], 67], 68]. However, the effect of dapagliflozin and other SGLT2 inhibitors in slowing the development
and progression of diabetic nephropathy is currently speculative, although preliminary
data on the effect of dapagliflozin in patients already treated with renin-angiotensin
system blockers have recently been presented, showing that dapagliflozin reduced albuminuria
without increasing renal adverse events 69], 70].

Safety and tolerability

The safety of dapagliflozin as monotherapy and as add-on therapy to existing antidiabetic
treatment in patients with T2DM has been evaluated in multiple randomized controlled
trials and has been assessed in five comprehensive reviews or meta-analyses 37]–40], 71].

A pooled safety analysis of 12 placebo-controlled phase II/III clinical trials of
up to 102 weeks duration that assessed dapagliflozin at doses of 2.5, 5, and 10 mg
once daily in a total of more than 4000 patients has recently been published 71]. The analysis was supplemented by data from an active comparator trial, from a trial
in patients with moderate renal impairment, and from five additional studies that
were ongoing at the time of analysis, giving a total of 19 studies. In general, the
analyses were performed for patients who received at least one dose of study medication
during the double blind phase of the study. The patient populations in the pooled
analyses were representative of the general population of patients with T2DM.

The majority (90 %) of treatment-emergent adverse events (AEs) were mild-to-moderate
in intensity, without a discernible dose relationship. One or more AEs were reported
by 61.7 % of dapagliflozin recipients and 56.9 % of placebo recipients, and AEs were
considered treatment-related in 17.3 % of dapagliflozin recipients and 13.3 % of the
placebo group.

Hypoglycemia, urinary tract infections, vulvovaginitis/balanitis and related genital
infections, back pain, polyuria, dysuria, and dyslipidemia were the most commonly
reported AEs in the pooled dapagliflozin group (Table 3) 71]. In none of the studies did hypoglycemia lead to withdrawal, and it predominantly
occurred when dapagliflozin was used with a sulfonylurea or insulin (Table 4). Similar proportions of dapagliflozin and placebo recipients reported serious AEs
(3.7 vs. 3.3 %, respectively) and AEs resulting in study discontinuation (2.8 vs.
2.5 %, respectively). The tolerability profile demonstrated in the short-term studies
was maintained consistently in patients receiving long-term dapagliflozin treatment.

Table 3. Incidence of adverse events (percent patients) reported in patients treated with dapagliflozin
5 mg, dapagliflozin 10 mg, or placebo

Table 4. Incidence of hypoglycemia (percent patients) stratified by monotherapy and add-on
therapies

No substantial adverse effects on serum electrolytes, liver function, or renal function
were reported 71]. In general, there was a transient decrease in eGFR during the first weeks of dapagliflozin
treatment, followed by a return to baseline levels or higher. Overall, there was no
evidence of new or worsening renal impairment, acute nephrotoxicity, or progression
of diabetic nephropathy in dapagliflozin-treated patients at up to 2 years 71]. Recent findings from routine clinical practice confirm that dapagliflozin therapy
is not associated with nephrotoxicity 72].

Volume-related events, i.e. hypotension, dehydration and hypovolemia, were infrequent
and none were serious, but occurred more often in the dapagliflozin group (0.8 %)
compared with placebo (0.4 %) 71]. There was a higher risk of volume depletion events for dapagliflozin compared with
placebo in patients also receiving loop diuretics (6.1 vs. 1.8 %, respectively). Numerically
higher rates of dyslipidemia with dapagliflozin versus placebo have been reported,
with small elevations in total cholesterol, low-density lipoprotein cholesterol and
high-density lipoprotein cholesterol, and reductions in triglycerides 71]. There are some reports about patients on SGLT2 inhibitor treatment developing diabetic
ketoacidosis 73]. Although there are no published reports of ketoacidosis specifically with dapagliflozin,
it can be assumed that this is a class effect due to inappropriate prescriptions in
patients with insufficient insulin (either endogenous or exogenous) and in which ketoacidosis
has been masked by concomitant euglycaemia 74]. The US Food and Drug Administration and the European Agency of Medicine are currently
investigating this issue through an extensive review of all available data and will
consider whether any changes are needed in the way these medicines are used in the
US and EU 75].

Although investigation into the cardiovascular effects of dapagliflozin and other
SGLT2 inhibitors is ongoing, the cardiovascular safety of dapagliflozin appears to
be favorable. An independently-adjudicated meta-analysis of data from 9000 patients
with T2DM suggests that dapagliflozin does not increase cardiovascular risk in terms
of MACE (major adverse cardiac events; a composite of cardiovascular death, non-fatal
stroke, and non-fatal myocardial infarction) versus placebo or active control 76]. No adverse impact on cardiovascular safety of dapagliflozin treatment, compared
with placebo, was recently reported even in high-risk patients with pre-existing cardiovascular
disease and a history of hypertension 77]. Of interest, a multicenter, double blind, randomized, parallel group trial is underway
(DECLARE-TIMI58 study; https://clinicaltrials.gov/show/NCT01730534) to evaluate the possibility that dapagliflozin may have a beneficial effect on the
incidence of cardiovascular events in patients with T2DM and established cardiovascular
disease or multiple risk factors.

Among some phase 2 and phase 3 trials in the dapagliflozin development program there
was a small excess in event rates for male bladder cancer and female breast cancer
over rates expected of the age-matched diabetic population 71], 78]. However, the imbalance was not statistically significant, and the diagnosis of all
breast and bladder cancers within 1 and 2 years, respectively, of starting dapagliflozin
and the wide biological heterogeneity of, in particular, the bladder cancers argue
against a single causative agent. Updated data gathered from an additional 21 trials
to November 2013 did not find any imbalance of malignancies in dapagliflozin-treated
patients 78]. Nevertheless, although there are no indications of a statistically significantly
increase in risk of cancer with dapagliflozin and it appears that early tumor diagnosis
may be attributable to detection bias rather than as a result of a causal relationship,
this issue, specifically regarding the risk of bladder cancer, will be further explored
in ongoing dapagliflozin trials 71], 78]. Moreover, there have been no carcinogenicity or mutagenicity signals in the preclinical
development program 65], and a recent study found that exposure of mice and rats to dapagliflozin for up
to 2 years at levels greater than 100-fold and up to 186-fold human clinical exposure,
respectively, did not increase bladder or mammary tumor rates, urinary bladder proliferative/preneoplastic
lesions, or enhance tumor growth in murine models of human bladder transitional cell
carcinoma, suggesting that dapagliflozin does not promote tumor growth 79].

In older patients (?65 years), the incidence and nature of AEs was similar to that
observed in the overall population and in patients younger than 65 years 71].

Although there was a higher incidence of fractures reported among dapagliflozin compared
with placebo recipients in a study of T2DM patients with moderate renal impairment
35], there was no evidence for an imbalance in fracture rate of patients in the dapagliflozin
and all control groups in the pooled safety data 71]. Furthermore, dapagliflozin did not affect markers of bone formation and resorption,
or bone mineral density, at week 50 of a placebo-controlled study in 182 female and
male patients with T2DM 80].

The generally favorable and predictable tolerability profile of dapagliflozin reported
in this section is supported by the various meta-analyses and comprehensive reviews
published to date 37]–40], 71].