Initial Combination of Empagliflozin and Linagliptin in Subjects With Type 2 Diabetes

Empagliflozin is a potent and selective sodium–glucose cotransporter 2 (SGLT2) inhibitor (1) approved for the treatment of type 2 diabetes. Empagliflozin reduces renal glucose reabsorption, thereby increasing urinary glucose excretion, leading to a reduction in plasma glucose levels in subjects with type 2 diabetes in an insulin-independent manner (2). In a phase 3 trial in subjects with type 2 diabetes, empagliflozin 10 mg and 25 mg once daily as monotherapy for 24 weeks reduced HbA_1c, fasting plasma glucose (FPG), weight, and blood pressure (BP) compared with placebo and were well tolerated, with a low risk of hypoglycemia (3). In a double-blind extension of this trial, the reductions in HbA_1c, FPG, and weight achieved with empagliflozin were sustained up to week 76 (4).

Linagliptin is a potent and selective dipeptidyl peptidase-4 (DPP-4) inhibitor (5) approved for the treatment of patients with type 2 diabetes. Linagliptin prevents the inactivation of incretin peptides such as glucagon-like peptide 1 (GLP-1), stimulates insulin release, and inhibits glucagon secretion (6). In a phase 3 trial in subjects with type 2 diabetes, linagliptin 5 mg given as monotherapy for 24 weeks improved glycemic control without weight gain and was well tolerated, with a low risk of hypoglycemia (7). In an open-label extension of this trial, reductions in HbA_1c with linagliptin were sustained up to week 102 (8).

Guidelines for the management of type 2 diabetes recommend metformin as first-line pharmacological treatment (9), but initial combination therapy with oral antidiabetes drugs with complementary modes of action may provide more robust and durable glucose-lowering efficacy with enhanced outcomes compared with the traditional stepwise approach (10). Initial combination therapies typically include metformin (9); however, contraindications to metformin, such as renal impairment, are common among patients with type 2 diabetes (11–13), and 5–10% of patients cannot tolerate metformin owing to gastrointestinal side effects (14). The complementary mechanisms of action of SGLT2 inhibitors and DPP-4 inhibitors suggest that the combination of empagliflozin and linagliptin may offer sustained treatment benefits. Therefore, this study was undertaken to evaluate the efficacy and safety of the initial combination of empagliflozin/linagliptin in subjects with type 2 diabetes.

This was a phase 3, randomized, double-blind, parallel-group study conducted from August 2011 to September 2013 in 197 centers in 22 countries. The clinical trial protocol was approved by the institutional review boards, independent ethics committees, and competent authorities of the participating centers and complied with the Declaration of Helsinki in accordance with the International Conference on Harmonisation Harmonised Tripartite Guideline for Good Clinical Practice. All subjects provided written informed consent.

This study enrolled subjects aged ≥18 years with type 2 diabetes with BMI ≤45 kg/m² and HbA_1c >7% to ≤10.5% (>53 to ≤91 mmol/mol) at screening despite a diet and exercise regimen who had not received treatment with oral antidiabetes therapy, GLP-1 analog, or insulin for ≥12 weeks prior to randomization.

Exclusion criteria included uncontrolled hyperglycemia (glucose level >240 mg/dL after an overnight fast during the placebo run-in, confirmed by a second measurement); estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m² (using the MDRD equation); acute coronary syndrome, stroke, or transient ischemic attack within 3 months prior to consent; bariatric surgery in the last 2 years; and treatment with antiobesity drugs within 3 months prior to consent.

After a 2-week placebo run-in period, subjects were randomized (1:1:1:1:1) to receive empagliflozin 25 mg/linagliptin 5 mg as a fixed-dose combination (FDC) tablet, empagliflozin 10 mg/linagliptin 5 mg as an FDC tablet, empagliflozin 25 mg, empagliflozin 10 mg, or linagliptin 5 mg for 52 weeks. All study drugs were taken once daily in the morning. Randomization was performed using a third-party interactive voice and web response system and was stratified by HbA_1c at screening (<8.5% [<69 mmol/mol], ≥8.5% [≥ 69 mmol/mol]), eGFR at screening (≥90 mL/min/1.73 m², 60–89 mL/min/1.73 m²), and region (Asia, Europe, North America, South America). Study visits were scheduled at screening, at the start of the placebo run-in, at baseline, and at weeks 6, 12, 18, 24, 32, 40, and 52 of treatment. A follow-up visit occurred 4 weeks after the last dose of study drug for subjects who completed the treatment period or within 7 days after the last administration of study drug for those who discontinued treatment before week 52.

Rescue medication was to be initiated if a subject had blood glucose >240 mg/dL after an overnight fast between weeks 1 and 12, blood glucose >200 mg/dL after an overnight fast between weeks 12 and 24, or blood glucose >180 mg/dL or HbA_1c >8% (>63.9 mmol/mol) after an overnight fast between weeks 24 and 52. The initiation, choice, and dosage of rescue medication were at the discretion of the investigator, according to local prescribing information, but the use of DPP-4 inhibitors, GLP-1 analogs, and SGLT2 inhibitors was not permitted. In cases of hypoglycemia, rescue medication was to be reduced in dose or discontinued. If hyper- or hypoglycemia could not be controlled, the subject was discontinued from the trial.

The primary end point was the change from baseline in HbA_1c at week 24. Key secondary end points were change from baseline in FPG at week 24, change from baseline in weight at week 24, and the proportion of subjects with baseline HbA_1c ≥7% (≥53 mmol/mol) who had HbA_1c <7% (<53 mmol/mol) at week 24. Exploratory end points were as follows: changes from baseline in HbA_1c at week 24 in subgroups of subjects with HbA_1c ≥8.5% (≥ 69 mmol/mol) and <8.5% (<69 mmol/mol) at baseline; changes from baseline in HbA_1c, FPG, weight, systolic BP (SBP), and diastolic BP (DBP) at week 52; and the proportion of subjects with baseline HbA_1c ≥7% (≥53 mmol/mol) who had HbA_1c <7% (<53 mmol/mol) at week 52.

Safety end points included vital signs, clinical laboratory parameters, and adverse events (AEs) (preferred terms coded according to the Medical Dictionary for Regulatory Activities [MedDRA], version 16.0). AEs included all events with an onset after the first dose and up to 7 days after the last dose of study medication. Confirmed hypoglycemic AEs were defined as AEs with plasma glucose ≤70 mg/dL and/or requiring assistance. Events consistent with urinary tract infection (UTI), events consistent with genital infection, and events consistent with volume depletion (identified from AEs reported spontaneously by the investigator using prospectively defined search categories based on 77, 89, and 8 preferred terms, respectively), hypersensitivity reactions (based on three standardized MedDRA queries), and pancreatitis (based on one standardized MedDRA query and one preferred term) were assessed.

Efficacy analyses were performed on the full analysis set (FAS), which included subjects treated with ≥1 dose of study drug who had a baseline and ≥1 on-treatment HbA_1c value. Safety was assessed in the treated set, which comprised subjects treated with ≥1 dose of study drug.

The primary end point was assessed using an ANCOVA model, with treatment, region, and eGFR at baseline as fixed-effects and baseline HbA_1c as a linear covariate. Values observed after a subject started rescue medication were set to missing. A last observation carried forward (LOCF) approach was used to impute missing continuous efficacy data. Continuous key secondary end points were analyzed using the ANCOVA model described for the primary end point with the baseline value for the end point in question as an additional linear covariate. Sensitivity analyses of the change from baseline in HbA_1c and FPG at week 24 were performed using restricted maximum likelihood–based mixed-model repeated measures (MMRM) in the FAS using observed cases and included treatment, region, visit, visit-by-treatment interaction, and eGFR as fixed-effects and baseline HbA_1c as a linear covariate. HbA_1c over 52 weeks was analyzed using the same MMRM analysis. Subgroup analyses at week 24 and changes from baseline at week 52 were analyzed using the ANCOVA model described for the primary end point. Categorical changes in HbA_1c at week 24 and week 52 were analyzed using logistic regression with noncompleters considered failure imputation.

Treatment differences in the primary and key secondary end points were tested hierarchically in the following order: HbA_1c, FPG, body weight (empagliflozin/linagliptin versus linagliptin only), and the percentage of subjects who reached HbA_1c <7% (<53 mmol/mol). For each end point, the superiority of empagliflozin 25 mg/linagliptin 5 mg versus the individual components was tested followed by the test of empagliflozin 10 mg/linagliptin 5 mg versus the individual components. Each test was at a significance level of 5% (two-sided). A test of superiority was confirmatory only if the previous tests were positive. With this procedure, the family-wise error rate was preserved at 5% (two-sided). Safety analyses were descriptive, except for changes in lipid parameters, which were analyzed using ANCOVA.

A sample size of 133 subjects per group was required to provide power of 89% to detect a 0.5% difference in change from baseline in HbA_1c between empagliflozin/linagliptin and the individual components, assuming a common SD of 1.05% and using a significance level of 2.5% (one-sided).

In total, 677 subjects were randomized and treated, of whom 667 comprised the FAS (Supplementary Fig. 1). Baseline characteristics of the FAS were balanced between treatment groups (Table 1).

At week 24, reductions from baseline in HbA_1c were significantly greater with empagliflozin/linagliptin compared with the individual components, except for empagliflozin 25 mg/linagliptin 5 mg versus empagliflozin 25 mg (Fig. 1A). In subjects with HbA_1c ≥8.5% (≥69 mmol/mol) at baseline (mean baseline 9.08–9.39% [76–79 mmol/mol]), reductions from baseline in HbA_1c at week 24 were significantly greater with empagliflozin/linagliptin compared with the individual components, except for empagliflozin 25 mg/linagliptin 5 mg versus empagliflozin 25 mg (Fig. 1B). In subjects with HbA_1c <8.5% (<69 mmol/mol) at baseline (mean baseline 7.41–7.63% [57–60 mmol/mol]), reductions from baseline in HbA_1c were significantly greater with empagliflozin 10 mg/linagliptin 5 mg compared with the individual components but were not significantly different with empagliflozin 25 mg/linagliptin 5 mg compared with the individual components (Supplementary Fig. 2). Significantly more subjects with baseline HbA_1c ≥7% (≥53 mmol/mol) reached HbA_1c <7% (<53 mmol/mol) at week 24 with empagliflozin/linagliptin compared with the individual components (Fig. 1C). Reductions from baseline in FPG at week 24 were significantly greater with empagliflozin/linagliptin compared with linagliptin but were not significantly different compared with the respective empagliflozin components (Fig. 1D). Sensitivity analyses of changes from baseline in HbA_1c and FPG at week 24 were consistent with the results of the primary analyses (Supplementary Table 1). Reductions from baseline in weight at week 24 were significantly greater with empagliflozin/linagliptin compared with linagliptin but not compared with the respective empagliflozin components (Fig. 1E).

Reductions in HbA_1c with empagliflozin/linagliptin were sustained at week 52 (Fig. 2Aand B). Significantly greater proportions of subjects with baseline HbA_1c ≥7% (≥53 mmol/mol) had HbA_1c <7% (<53 mmol/mol) at week 52 with empagliflozin/linagliptin compared with the individual components, except for empagliflozin 25 mg/linagliptin 5 mg compared with empagliflozin 25 mg (Fig. 2C). At week 52, reductions from baseline in FPG were significantly greater with empagliflozin/linagliptin compared with linagliptin but were not significantly different compared with the respective empagliflozin components (Supplementary Table 2).

Over 52 weeks, 6 (4.5%), 5 (3.7%), 6 (4.5%), 12 (9.1%), and 27 (20.3%) subjects on empagliflozin 25 mg/linagliptin 5 mg, empagliflozin 10 mg/linagliptin 5 mg, empagliflozin 25 mg, empagliflozin 10 mg, and linagliptin 5 mg, respectively, received rescue therapy. Most subjects who received rescue therapy received metformin or sulfonylurea.

Reductions from baseline in weight at week 52 were significantly greater with empagliflozin/linagliptin compared with linagliptin but were not significantly different compared with the respective empagliflozin components (Fig. 2D).

Changes from baseline in SBP and DBP at week 52 were not significantly different between empagliflozin/linagliptin and the individual components (Fig. 2E and F). Reductions in BP in the empagliflozin/linagliptin and empagliflozin monotherapy groups were not associated with relevant changes in pulse rate: mean ± SD changes from baseline at week 52 were 0.15 ± 8.83 bpm with empagliflozin 25 mg/linagliptin 5 mg, 0.24 ± 9.44 bpm with empagliflozin 10 mg/linagliptin 5 mg, −0.38 ± 8.52 bpm with empagliflozin 25 mg, −1.33 ± 8.16 bpm with empagliflozin 10 mg, and 1.28 ± 9.23 bpm with linagliptin 5 mg.

Similar proportions of subjects in every treatment group had one or more AEs (Table 2). Most events were mild or moderate in intensity; severe events were reported in 5.9% of subjects in each of the combination therapy groups, 7.4% in each of the empagliflozin groups, and 0.7% on linagliptin 5 mg. There was one death in the empagliflozin 10 mg/linagliptin 5 mg group (hemorrhagic stroke), two deaths in the empagliflozin 25 mg group (meningitis tuberculosis and hepatic mass), and one death in the empagliflozin 10 mg group (brain edema).

No subjects receiving empagliflozin/linagliptin had confirmed hypoglycemic AEs. Confirmed hypoglycemic AEs were reported in one subject each in the empagliflozin 25 mg and linagliptin 5 mg groups and in four subjects in the empagliflozin 10 mg group; no events required assistance. Events consistent with UTI were reported in 12.5% of subjects on empagliflozin 25 mg/linagliptin 5 mg, 15.4% on empagliflozin 10 mg/linagliptin 5 mg, 10.4% on empagliflozin 25 mg, 16.3% on empagliflozin 10 mg, and 10.4% on linagliptin 5 mg; these events were reported in a greater proportion of female than male subjects in every group (Table 2). Only one subject (receiving empagliflozin 25 mg) reported an event consistent with UTI of severe intensity; this event did not lead to discontinuation of study drug. Chronic pyelonephritis was reported in one subject (receiving empagliflozin 10 mg); this event was mild in intensity and was not considered to be related to study drug. Events consistent with genital infection were reported in 5.9% of subjects on empagliflozin 25 mg/linagliptin 5 mg, 2.9% on empagliflozin 10 mg/linagliptin 5 mg, 4.4% on empagliflozin 25 mg, 5.2% on empagliflozin 10 mg, and 3% on linagliptin 5 mg; these events were reported in a greater proportion of female than male subjects in all groups except empagliflozin 25 mg/linagliptin 5 mg (Table 2). No severe events consistent with genital infection were reported, but two subjects (one on empagliflozin 25 mg/linagliptin 5 mg and one on empagliflozin 10 mg) discontinued study drug due to these events. Hypersensitivity reactions were reported in two subjects on empagliflozin 25 mg/linagliptin 5 mg (asthmatic crisis and eyelid edema), one on empagliflozin 10 mg/linagliptin 5 mg (asthma), two on empagliflozin 25 mg (urticaria), two on empagliflozin 10 mg (asthma), and none on linagliptin 5 mg. Pancreatitis (pancreatitis acute) was reported in one subject (empagliflozin 25 mg/linagliptin 5 mg). No subjects experienced worsening of heart failure or were hospitalized due to heart failure.

Changes from baseline in laboratory measurements at week 52 are shown in Supplementary Table 3. Changes from baseline in eGFR observed at week 52 and follow-up were small in all groups. Mean changes from baseline in hematocrit were 4.2–4.8% in the empagliflozin/linagliptin and empagliflozin groups and 1.3% in the linagliptin group. Mean changes from baseline in uric acid were −56.5 to −71.4 µmol/L in the empagliflozin/linagliptin and empagliflozin groups and 3.0 µmol/L in the linagliptin group. There were no clinically meaningful changes in electrolytes. There were no significant differences in changes from baseline in total cholesterol, LDL cholesterol, and triglycerides between empagliflozin/linagliptin and the individual components, except for a greater increase in total cholesterol with empagliflozin 25 mg/linagliptin 5 mg compared with linagliptin 5 mg (Supplementary Table 3). There were no significant differences in changes from baseline in HDL cholesterol between empagliflozin/linagliptin and the empagliflozin components, but there was a greater increase in HDL cholesterol with empagliflozin/linagliptin than with linagliptin (Supplementary Table 3). There were no consistent patterns in the proportions of subjects on empagliflozin/linagliptin with shifts in albuminuria categories from baseline to the end of treatment compared with the respective components (Supplementary Table 4).

This is the first randomized controlled trial to evaluate the efficacy and safety of the initial combination of an SGLT2 inhibitor and a DPP-4 inhibitor in subjects with type 2 diabetes. Initial combination of empagliflozin/linagliptin led to clinically meaningful reductions from baseline in HbA_1c in subjects with type 2 diabetes, with >50% of subjects reaching HbA_1c <7% (<53 mmol/mol) at week 52. Greater reductions from baseline in HbA_1c (of up to 1.93% [21.1 mmol/mol]) were observed with empagliflozin/linagliptin in subjects with baseline HbA_1c ≥8.5% (≥69 mmol/mol) compared with those with baseline HbA_1c <8.5% (<69 mmol/mol) who had reductions of up to 0.96% (10.5 mmol/mol). Initial combination of empagliflozin/linagliptin may be an option in patients presenting with marked hyperglycemia, as achieving glycemic targets may be challenging with traditional stepwise treatment escalation, and treatment escalation in clinical practice is often delayed. In a retrospective analysis of >3,000 patients on sulfonylurea monotherapy and >500 patients on metformin monotherapy, patients spent an average of 20.5 months with HbA_1c >8% (>64 mmol/mol) on sulfonylurea monotherapy and 14.5 months with HbA_1c >8% (>64 mmol/mol) on metformin monotherapy before a new treatment was initiated (15).

Unexpectedly, while empagliflozin 10 mg/linagliptin 5 mg demonstrated greater efficacy compared with empagliflozin 10 mg and compared with linagliptin, empagliflozin 25 mg/linagliptin 5 mg showed greater efficacy compared with linagliptin 5 mg but not compared with empagliflozin 25 mg. There were no imbalances in baseline characteristics, dropouts, or outliers between treatment groups that explained this finding, and the reasons behind it are unclear. However, although subjects did not receive treatment for ≥12 weeks before randomization, we cannot rule out the possibility that subjects who had previously received treatment had different responses compared with “true” drug-naïve subjects.

Empagliflozin consistently leads to weight loss, likely due to loss of calories via increased urinary glucose excretion (16), while linagliptin treatment is weight neutral (7). As expected, the effect of empagliflozin on body weight was maintained when empagliflozin was used in combination with linagliptin, with both combinations significantly reducing weight compared with linagliptin monotherapy. Empagliflozin reduces SBP via mechanisms that may include osmotic diuretic effects, weight loss, reduced arterial stiffness, or direct vascular effects (17–19), while linagliptin has no effect on blood pressure (20). In this study, treatment with empagliflozin/linagliptin and empagliflozin alone resulted in a consistent, modest reduction in SBP from baseline, but reductions with empagliflozin/linagliptin were not statistically significant compared with linagliptin monotherapy.

Slightly higher percentages of subjects had AEs leading to discontinuation, severe AEs, and serious AEs with empagliflozin/linagliptin or empagliflozin compared with linagliptin, but all the treatments were well tolerated. The overall safety profiles of empagliflozin/linagliptin were similar to the known safety profiles of the individual components. In patients with type 2 diabetes, hypoglycemia is associated with increased risk of cardiovascular events, reduced quality of life, and deterioration in glucose control (21). Empagliflozin and linagliptin are associated with a low risk of hypoglycemia when given as monotherapy (3,8), and no confirmed hypoglycemic AEs were observed with empagliflozin/linagliptin. Guidelines recommend that the potential for hypoglycemic events should play a major role in the selection of first-line therapy if metformin is not used and in the selection of combination therapies (9). Therefore, the low risk of hypoglycemic AEs with empagliflozin/linagliptin in this study is a particularly important finding.

Strengths of this study include the 52-week trial duration to assess the sustainability of the effect of empagliflozin/linagliptin. Limitations include the lack of a placebo arm, which means that the additive efficacy of empagliflozin/linagliptin compared with the individual components cannot be conclusively assessed, and the fact that the patient population studied may not have been a homogenous population of drug-naïve patients.

In conclusion, initial combination of empagliflozin/linagliptin led to clinically meaningful reductions from baseline in HbA_1c in subjects with type 2 diabetes. Reductions with empagliflozin 10 mg/linagliptin 5 mg were significantly greater than with the individual components, and reductions with empagliflozin 25 mg/linagliptin 5 mg were significantly greater compared with linagliptin 5 mg but not compared with empagliflozin 25 mg. Empagliflozin/linagliptin was well tolerated, with a low risk of hypoglycemia. Initial combination of empagliflozin/linagliptin may provide a treatment option for patients with type 2 diabetes who are intolerant to metformin and/or have marked hyperglycemia, without the side effects of weight gain or increased risk of hypoglycemia.

Duality of Interest. This study was funded by Boehringer Ingelheimhttp://dx.doi.org/10.13039/100001003 and Eli Lilly and Company. R.A.D. has served on scientific advisory boards for Boehringer Ingelheim, AstraZeneca, Janssen, Novo Nordisk, Takeda, and Lexicon; participated in speakers’ bureaus for AstraZeneca and Novo Nordisk; and received grants/research support from AstraZeneca, Janssen, Bristol-Myers Squibb, Takeda, and Xeris. S.P., D.L., R.K., H.J.W., and U.C.B. are employees of Boehringer Ingelheim. Medical writing assistance, supported financially by Boehringer Ingelheim, was provided by Elizabeth Ng and Wendy Morris of Fleishman-Hillard Group, Ltd., during the preparation of the manuscript. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. A.L. contributed to the acquisition and interpretation of data and writing of the manuscript. R.A.D. contributed to the interpretation of data and writing of the manuscript. S.P., D.L., R.K., H.J.W., and U.C.B. contributed to the study design, interpretation of data, and writing of the manuscript. All authors had full access to the study data, were responsible for the final decision to submit the manuscript, and approved the final version of the manuscript. R.A.D. 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 in abstract form at the 74th Scientific Sessions of the American Diabetes Association, San Francisco, CA, 13–17 June 2014, and at the at the 50th Annual Meeting of the European Association for the Study of Diabetes, Vienna, Austria, 15–19 September 2014.