Dapagliflozin Improves Glycemic Control and Reduces Body Weight as Add-on Therapy to Metformin Plus Sulfonylurea: A 24-Week Randomized, Double-Blind Clinical Trial

Metformin is recommended as initial pharmacologic therapy for patients with type 2 diabetes in both the U.S. and the European Union because of its glycemic efficacy, weight neutrality, low risk of hypoglycemia, good tolerability, and relatively low cost (1). However, because of the progressive nature of the disease, there is often a loss of efficacy over time with metformin and a need for a second-line therapy (2,3).

Sulfonylureas are commonly used as add-on therapy for patients inadequately controlled with metformin (4). Dual combination therapy is recommended by the American Diabetes Association and the European Association for the Study of Diabetes should HbA_1c levels exceed 9% (75 mmol/mol) (1) and by the Canadian Diabetes Association should HbA_1c levels exceed 8.5% (69 mmol/mol) (5). Initially this approach is often effective, but there is a lack of durability in combination therapy with metformin and sulfonylurea. Despite an initial improvement in glycemic control following the addition of sulfonylurea to metformin, deterioration can resume as early as 6 months at a rate comparable to that observed with metformin alone (2,6).

If dual therapy fails to adequately control HbA_1c levels, the American Diabetes Association, the European Association for the Study of Diabetes, and the American Association of Clinical Endocrinologists recommend the addition of a third oral therapy as one potential therapeutic option (1,7). This could include the addition of a GLP-1 analog, a dipeptidyl peptidase-4 inhibitor, or basal insulin. Inhibitors of sodium–glucose cotransporter 2 (SGLT2) are another possibility as an add-on to dual therapy. SGLT2 inhibitors are not reliant on β-cell function and can be used across all stages of the disease (8).

Dapagliflozin is a highly selective inhibitor of SGLT2 that improves glycemic control through the elimination of glucose in the urine. The associated caloric loss leads to a reduction in body weight (or prevents weight gain). Dapagliflozin therapy is also accompanied by modest reductions in blood pressure (BP) that are consistent with mild diuresis (9–16).

The aim of this study was to evaluate the safety and efficacy of dapagliflozin as part of a triple combination therapy in patients with type 2 diabetes and inadequate glycemic control with metformin and sulfonylurea.

Overall, 46 centers in North America (Canada) and Europe (Czech Republic, Germany, Poland, Slovak Republic, and Spain) enrolled patients, and 45 centers randomized patients.

This was a 24-week, international, multicenter, randomized, double-blind, parallel-group, placebo-controlled, phase 3b study (NCT01392677, clinicaltrials.gov). The study consisted of a screening period of up to 3 weeks; an 8-week, single-blind, placebo lead-in period; and a 24-week, double-blind, placebo-controlled treatment period (Supplementary Fig. 1). During the double-blind treatment period, patients were evaluated by the investigators at weeks 4, 8, 16, and 24.

This study was performed in accordance with the ethical principles that have their origin in the Declaration of Helsinki and that are consistent with the International Conference on Harmonisation/Good Clinical Practice, applicable regulatory requirements, and the AstraZeneca policy on bioethics. All patients gave written informed consent.

Patients on metformin and sulfonylurea were randomized 1:1 via an interactive response system (web or voice based; Perceptive Services Limited) to receive dapagliflozin 10 mg once daily or matched placebo (identical in size, color, smell, taste, packaging, and labeling).

Men and women who were not of childbearing potential (defined as women who were postmenopausal or who were permanently or surgically sterilized) or women of childbearing potential using a highly effective method of birth control and a negative pregnancy test 72 h prior to the start of medication and at each visit were eligible to enter the study. Patients were at least 18 years of age, had a diagnosis of type 2 diabetes, were receiving a stable-dose combination therapy of metformin ≥1,500 mg/day and a maximum tolerated dose (at least half the maximum dose according to local use) of sulfonylurea for at least 8 weeks prior to enrollment, and had inadequate glycemic control (HbA_1c ≥7.0% [53 mmol/mol] to ≤10.5% [91 mmol/mol] at randomization).

Patients were excluded from the study if they had a diagnosis of type 1 diabetes, BMI ≥45.0 kg/m², measured serum creatinine value of ≥1.5 mg/dL (133 μmol/L) for men or ≥1.4 mg/dL (124 μmol/L) for women, unstable or rapidly progressing renal disease, cardiovascular events within 2 months prior to enrollment, congestive heart failure (New York Heart Association Class IV), systolic BP (SBP) ≥160 mmHg, or diastolic BP (DBP) ≥100 mmHg at randomization.

The primary end point was change from baseline to week 24 in HbA_1c levels. Prespecified subgroup analyses of the change in HbA_1c were performed in patients with baseline HbA_1c values <8% (<64 mmol/mol), ≥8% (≥64 mmol/mol) to <9% (<75 mmol/mol), and ≥9% (≥75 mmol/mol). Key secondary end points included change from baseline to week 24 in fasting plasma glucose (FPG) and total body weight, proportion of patients achieving a therapeutic glycemic response (defined as HbA_1c <7.0% [53 mmol/mol]) at week 24, and change from baseline to week 8 in seated SBP (average of three measurements taken from the same arm after the patient had been sitting and resting for at least 5 min).

During the placebo lead-in period, concomitant antihypertensive medications could be adjusted in patients with seated SBP ≥160 mmHg or seated DBP ≥100 mmHg, without adding new antihypertensive agents to the patient’s regimen. Only those patients with seated SBP <160 mmHg and DBP <100 mmHg at the end of the placebo lead-in period were randomized. There were no changes in antihypertensive medications permitted up to week 8 unless the patient had a confirmed SBP ≥160 mmHg or DBP ≥100 mmHg or had symptomatic hypotension or documented orthostatic hypotension during a study visit. After week 8, changes in antihypertensive medication were made as needed for appropriate BP management.

Other secondary end points included change from baseline to week 24 in fasting lipids (total cholesterol, LDL cholesterol, HDL cholesterol, LDL/HDL cholesterol ratio, and triglycerides) and C-peptide and proportion of patients who discontinued for lack of efficacy or were rescued for failing to maintain FPG below prespecified rescue criteria.

Rescue therapy was mandated by the study protocol to guard the safety of patients. Open-label rescue therapy was administered with the patients’ study medication if FPG exceeded 240 mg/dL (13.32 mmol/L) between weeks 4 and 16 or exceeded 200 mg/dL (11.10 mmol/L) between weeks 16 and 24.

Sulfonylurea could be down-titrated only once during the treatment period to mitigate the risk of recurrent hypoglycemic events at the discretion of the investigator.

The safety and tolerability of dapagliflozin versus placebo were assessed through the evaluation of adverse events, including cardiovascular events, laboratory values, electrocardiogram, vital signs, hypoglycemic events, calculated creatinine clearance, estimated glomerular filtration rate, and physical examination findings over 24 weeks.

Adverse events were defined as the development, at any time throughout the study, of an undesirable medical condition or the deterioration of a preexisting medical condition whether or not considered causally related to the treatment. Serious adverse events were defined as adverse events that were immediately life threatening, required in-patient hospitalization or prolongation of existing hospitalization, resulted in persistent or significant disability or incapacity, were congenital abnormalities or birth defects, were important medical events jeopardizing the patient or requiring medical intervention to prevent one of the outcomes listed above, or resulted in death. Events of urinary tract infection and genital infection were determined using a predefined list of Medical Dictionary for Regulatory Activities terms (MedDRA version 15.1). Patients reported symptoms of genital infection and urinary tract infection spontaneously throughout the study as well as in response to questions posed by the investigators at study visits.

Minor episodes of hypoglycemia were defined as either a symptomatic episode with a capillary or plasma glucose measurement <3.5 mmol/L (<63 mg/dL), regardless of need for external assistance, or an asymptomatic capillary or plasma glucose measurement <3.5 mmol/L (<63 mg/dL) that did not qualify as a major episode. A major episode was defined as a symptomatic episode requiring third-party assistance due to severe impairment in consciousness or behavior with a capillary or plasma glucose value <3 mmol/L (<54 mg/dL) and prompt recovery after glucose or glucagon administration.

Efficacy analyses were based on the full analysis set, which included all randomized patients who received at least one dose of study medication during the 24-week double-blind treatment period with a nonmissing baseline value and one or more postbaseline value for at least one efficacy variable analyzed at week 24. The intention-to-treat principle was preserved despite the exclusion of patients who took no study medication. The primary safety analyses included all data regardless of rescue.

Longitudinal repeated measures analysis was used to evaluate the primary end point using a direct likelihood method with fixed categorical effects of treatment, week, treatment-by-week interaction, as well as the continuous fixed covariates of baseline and baseline-by-week interaction. Data for scheduled time points up to week 24 prior to rescue were included in the longitudinal repeated measures analysis.

Variables other than the primary end point were analyzed using an ANCOVA with treatment group as a fixed effect and the baseline measurement as a covariate, and the last observation carried forward (LOCF) approach (17) was used for all other variables at 24 weeks. LOCF analyses from baseline to 24 weeks were based on measurements available at that time point; if no measurement was available, the last postbaseline measurement prior to 24 weeks was used. Unless otherwise specified, if a patient initiated rescue medication, the last value taken on or before the first rescue dose was used for LOCF analysis. All statistical testing performed was two-sided. A hierarchical closed testing procedure was used to control for type I error rate across the primary and key secondary objectives. Treatment comparisons were individually tested for the primary and key secondary end points at a two-sided significance level of 0.05. For all other variables, including changes from baseline in lipids and C-peptide, nominal P values were reported without significance testing. The analysis of percentage changes in lipids is conducted using the logarithms of the absolute lipid values. All analyses were performed with SAS (SAS Institute, Cary, NC) version 8.2 or higher.

Patient disposition is summarized in Supplementary Fig. 2. Briefly, 219 patients entered study randomization, one of whom died before randomization. Patients were randomized 1:1 to receive placebo (n = 109) or dapagliflozin 10 mg/day (n = 109) in addition to their metformin and sulfonylurea treatment. Approximately 93% of the patients in each treatment arm completed the 24-week double-blind treatment period. Adverse events leading to patient withdrawal included one event each of somnolence, skin reaction, and renal cell carcinoma in the placebo group and one event of chronic obstructive pulmonary disease in the dapagliflozin group. Three patients in the placebo group and two in the dapagliflozin group discontinued treatment due to reasons classed as “other” and included patient choice, lack of glycemic control determined by the investigator, and the patient moving out of the country in the placebo group and patient choice and patient travel in the dapagliflozin group. Two patients in the placebo group and three in the dapagliflozin group were discontinued from the study for not meeting the prespecified inclusion criteria or for meeting the exclusion criteria (incorrect enrollment). In addition, two patients in the dapagliflozin group withdrew their consent.

The treatment groups were generally balanced with respect to demographics and diabetes-related baseline characteristics, with a higher proportion of women in the dapagliflozin treatment arm (Table 1). Most patients were white, had a mean age of 61 years, had a prior history of cardiovascular disease, and were receiving antihypertensive medications (mostly angiotensin receptor blockers and/or ACE inhibitors).

HbA_1c levels were significantly reduced from baseline (dapagliflozin 8.08% [65 mmol/mol]; placebo 8.24% [67 mmol/mol]) over 24 weeks in patients treated with dapagliflozin compared with placebo (–0.86% [–9.4 mmol/mol] vs. –0.17% [–1.9 mmol/mol], respectively; P < 0.0001) (Table 2 and Fig. 1A).

In patients with higher baseline HbA_1c values (≥8% [≥64 mmol/mol] to <9% [<75 mmol/mol] and ≥9% [≥75 mmol/mol]), the mean placebo-subtracted decreases in HbA_1c seen with dapagliflozin were numerically greater (–0.64% [–7.0 mmol/mol] and –0.82% [–9.0 mmol/mol]) compared with –0.36% (–3.9 mmol/mol) in patients with lower baseline HbA_1c values (<8% [<64 mmol/mol]) (Supplementary Table 1). The impact of baseline HbA_1c on the reduction in HbA_1c from baseline was evaluated using a subgroup-by-treatment interaction analysis and was statistically significant (P = 0.0038).

A greater proportion of patients receiving dapagliflozin 10 mg/day compared with placebo achieved a therapeutic glycemic response (defined as HbA_1c <7.0% [<53 mmol/mol]) at week 24 (LOCF) (31.8 vs. 11.1%; P < 0.0001) (Table 2).

Greater mean reductions from baseline to week 24 (LOCF) in FPG were observed with dapagliflozin compared with placebo (–34.2 mg/dL [–1.90 mmol/L] vs. –0.8 mg/dL [–0.04 mmol/L], respectively; adjusted mean difference –33.5 mg/dL [–1.86 mmol/L]; P < 0.0001) (Table 2 and Fig. 1B).

Significantly greater mean reductions from baseline to week 24 (LOCF) in body weight were observed with dapagliflozin compared with placebo (–2.7 vs. –0.6 kg, respectively; adjusted mean difference –2.1 kg; P < 0.0001) (Table 2 and Fig. 1C).

Patients in the dapagliflozin group showed a significantly larger reduction in seated SBP from baseline to week 8 (LOCF) compared with placebo (–4.0 vs. –0.3 mmHg, respectively; adjusted mean difference –3.8 mmHg; P = 0.025).

No patients in the dapagliflozin group and 10 patients in the placebo group (9.3%) were rescued due to lack of glycemic control at week 24 (LOCF).

Patients receiving dapagliflozin showed a mean placebo-corrected adjusted increase in total cholesterol (P = 0.0091) (Table 2). Increases were also seen with LDL cholesterol (P = 0.0030) and HDL cholesterol (P = 0.0172). No significant change from baseline was observed in LDL/HDL cholesterol ratio (P = 0.2008) or triglyceride levels (P = 0.1755).

Mean serum levels of C-peptide did not show any change in either treatment group (placebo-subtracted change –0.11 ng/mL).

The frequency of adverse events was similar between placebo (51.4%) and dapagliflozin (48.6%) groups (Table 3). The majority of adverse events were mild or moderate and not considered related to treatment by the investigator. Significantly more patients with dapagliflozin compared with placebo experienced events of hypoglycemia (12.8 vs. 3.7%, respectively; P = 0.024; post hoc analyses). There were no major episodes of hypoglycemia, and no events of hypoglycemia led to discontinuation in either group.

Over 24 weeks, significantly more events of genital infection were experienced by patients receiving dapagliflozin (six [5.5%]; five women, one man) compared with placebo (0%; P = 0.029; post hoc analyses): two events of vulvovaginal candidiasis and one episode each of balanitis candida, genital infection, vaginal infection, and vulvovaginitis. All events of genital infection were single events and were treated with antibiotic agents. Three of the events occurred within the first 12 weeks of treatment. Two of these events required additional treatment. No patient was discontinued from study medication due to an event of genital infection. Events of urinary tract infection were reported by seven (6.4%) patients in each group (one man and six women receiving placebo and three men and four women receiving dapagliflozin). Three patients in the dapagliflozin group experienced 2 events of urinary tract infection for a total of 10 events. In the placebo group, five of the seven events of urinary tract infection were reported in the first half of the study as well as six of the seven events initially reported in the dapagliflozin group. Antimicrobial treatment was administered for 7 of the 10 events in the dapagliflozin group and 6 of the 7 events in the placebo group. One patient in the dapagliflozin group was given additional treatment owing to an inadequate response to the initial treatment. No patient was discontinued from study medication because of a urinary tract infection event. One event of chronic pyelonephritis was observed in a female patient in the dapagliflozin group that was identified by sonograph while the patient was being investigated for chronic hematuria. The patient experienced no further symptoms and did not receive any specific treatment for the event.

One event of volume depletion was noted over the 24-week treatment period. This was an episode of orthostatic hypotension that resolved without recurrence and occurred in a patient treated with dapagliflozin.

Marked abnormalities of liver function tests were rare (one episode of elevated total bilirubin in a patient receiving dapagliflozin). One nonserious unspecified bladder neoplasm was reported (diagnosis at 3 months) in a female patient in the dapagliflozin group with hematuria at randomization and after 2 months. Approximately 6 months after start of treatment, the patient was diagnosed with bladder cancer.

Little information is available involving the concomitant use of three oral antidiabetic drugs in patients with type 2 diabetes, although thiazolidinediones and dipeptidyl peptidase-4 inhibitors have been shown to be effective as add-on treatments in patients inadequately controlled with metformin and sulfonylureas (7,18,19).

Dapagliflozin has been shown to be well tolerated and effective when administered as monotherapy or in combination with other antihyperglycemic medications (9–12,15,16,20). The safety and efficacy of dapagliflozin as part of a triple therapy with metformin and sitagliptin have been evaluated previously (13). In addition, preliminary analyses were performed to evaluate dapagliflozin in combination with metformin plus sulfonylurea (prespecified analyses with no multiplicity testing) and with metformin plus insulin (post hoc analyses) (20). The results indicated that dapagliflozin was effective as part of a triple therapy in improving glycemic control and reducing body weight in type 2 diabetes. This study evaluated the safety and efficacy of dapagliflozin administered as part of a triple therapy for patients with type 2 diabetes and inadequate glycemic control with metformin and sulfonylurea. In patients with inadequate glycemic control taking metformin plus a sulfonylurea, dapagliflozin resulted in a significant improvement in glycemic control (reductions in HbA_1c and FPG) and reduced body weight when compared with placebo. Similar results have been reported with other SGLT2 inhibitors when administered in combination with metformin and sulfonylurea (21,22). Consistent with the reported diuretic effects of dapagliflozin, seated SBP was also reduced with dapagliflozin compared with placebo. Type 2 diabetes is associated with an increased risk of cardiovascular disease (23), and the weight loss and BP reduction associated with dapagliflozin treatment could provide a significant cardiovascular benefit to patients with type 2 diabetes inadequately controlled with their current antihyperglycemic regimens. However, it should be noted that the effects on lipid parameters observed with dapagliflozin in this study (increases in total cholesterol, LDL cholesterol, and HDL cholesterol but no significant change in LDL vs. HDL cholesterol ratio or triglyceride levels) are consistent with pooled data across the phase 2b/3 program (24).

Patients with type 2 diabetes have an increased susceptibility to asymptomatic bacteriuria, urinary tract infections, and genital infections (25). An increase in the growth of microbiological flora in the genital tract would not be unexpected under conditions of glucosuria such as induced through the inhibition of glucose reabsorption, and as is observed with all SGLT2 inhibitors, dapagliflozin was associated with a higher frequency of genital infection events. These were predominantly mild to moderate in intensity, responded to antibiotic treatment, and did not result in any study discontinuations. Urinary tract infections were reported at a similar frequency in both groups and were also primarily mild to moderate in intensity and responded to antimicrobial treatment. Clinicians prescribing dapagliflozin should advise their patients to report any symptoms suggestive of genital or urinary tract infections and assess suitability for treatment.

Dapagliflozin has a low intrinsic risk of hypoglycemia due to its insulin-independent mode of action. However, there may be a potential to increase this risk when added to therapies known to cause hypoglycemia such as sulfonylureas and insulin. In this trial, hypoglycemia was observed more frequently in patients receiving dapagliflozin compared with placebo when added to preexisting metformin and sulfonylurea treatment. An increase in the frequency of hypoglycemic events was also observed with other SGLT2 inhibitors when administered in combination with metformin and sulfonylurea (21,22). In clinical practice, when using dapagliflozin in patients already on background sulfonylurea therapy, discontinuation or down-titration of sulfonylurea should be considered in patients at risk (26).

The results presented here indicate that triple combination therapy that includes dapagliflozin 10 mg/day with metformin and sulfonylurea is a suitable treatment option for patients with inadequate glycemic control on metformin and a sulfonylurea, with the added benefits of weight loss and BP reduction. Consistent with other SGLT2 inhibitors, dapagliflozin was associated with an increase in frequency of genital infections and increases in total, LDL, and HDL cholesterol that may necessitate additional lipid management. Further studies are underway to explore the effects of longer-term exposure to dapagliflozin.

Duality of Interest. This study was sponsored by Bristol-Myers Squibbhttp://dx.doi.org/10.13039/100002491 and AstraZenecahttp://dx.doi.org/10.13039/100004325. Medical writing assistance was provided by Karen Pemberton of PPSI (a PAREXEL company) and was funded by AstraZeneca. S.M. holds advisory board membership at Abbott, AstraZeneca, Bristol-Myers Squibb (BMS), Boehringer Ingelheim, Janssen, Merck Sharp & Dohme, Novo Nordisk, and Roche and is part of the speakers’ bureau at Abbott, AstraZeneca, Bayer, BMS, Boehringer Ingelheim, Janssen, Lilly, Merck Sharp & Dohme, Novartis, Novo Nordisk, Roche, and Sanofi. K.B. holds advisory board membership at AstraZeneca, Boehringer Ingelheim, BMS, Eli Lilly, Merck, Novo Nordisk, Pfizer, Johnson & Johnson, and Takeda; is part of the speakers’ bureau at AstraZeneca, Boehringer Ingelheim, BMS, Eli Lilly, Merck, and Novo Nordisk; and has performed clinical research for AstraZeneca, Boehringer Ingelheim, Merck, Novartis, Novo Nordisk, and Eli Lilly. K.R. is a full-time employee of AstraZeneca. A.G. is a full-time employee of Aptiv Solutions. S.P. is a full-time employee and stock holder of AstraZeneca. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. S.M. reviewed/edited the manuscript and contributed to the discussion. K.B., K.R., A.G., and S.P. researched data, reviewed/edited the manuscript, and contributed to the discussion. S.M. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented at the 49th Annual Meeting of the European Association for the Study of Diabetes, Barcelona, Spain, 23–27 September 2013.