Empagliflozin Reduces Blood Pressure in Patients With Type 2 Diabetes and Hypertension

The lifetime risk of cardiovascular disease in patients with diabetes is 67–78% (1). Management of patients with type 2 diabetes should not only aim to control glycemia but also include the modification of cardiovascular risk factors (2,3). Hypertension affects approximately two-thirds of patients with diabetes (4,5) and is a significant contributing factor to cardiovascular complications (3,6). Lowering blood pressure (BP) has been shown to reduce cardiovascular events in patients with diabetes and to exert a renoprotective effect (3,7,8). The Joint National Committee (JNC) 2003 guidelines recommended targets of systolic BP (SBP) <130 mmHg and diastolic BP (DBP) <80 mmHg in patients with hypertension and diabetes (9). The guidelines were recently updated to recommend a target BP <140/90 mmHg in these patients (10). As BP control is difficult to achieve in patients with diabetes (11,12), combination therapy is often required (13).

Empagliflozin, a potent and selective sodium–glucose cotransporter 2 (SGLT2) inhibitor (14), is in development for the treatment of type 2 diabetes. Inhibition of SGLT2, located in the proximal tubule of the kidney, leads to increased urinary glucose excretion and improvements in hyperglycemia in patients with type 2 diabetes via a mechanism of action independent of β-cell function and insulin resistance (15). Treatment with empagliflozin for up to 90 weeks led to sustained improvements in glycemic control and reductions in body weight, SBP, and DBP in patients with type 2 diabetes (16‒20). Empagliflozin was well tolerated and associated with a low risk of hypoglycemia (16‒20).

The primary objective of this phase 3 study (EMPA-REG BP) was to determine the effect of 10 mg and 25 mg empagliflozin o.d. for 12 weeks on BP, glycemic control and to assess safety and tolerability compared with placebo in patients with type 2 diabetes and hypertension.

Patients with type 2 diabetes and hypertension (mean seated office SBP 130–159 mmHg and DBP 80–99 mmHg) aged ≥18 years with a BMI ≤45 kg/m² and HbA_1c ≥7.0 to ≤10.0% (≥53 to ≤86 mmol/mol) at screening were eligible for inclusion. Patients were required to receive up to two antihypertensive medications at a stable dose for ≥4 weeks at screening and throughout a 2-week, open-label, placebo run-in period. For the treatment of type 2 diabetes, patients were required to be on a diet and exercise regimen and be drug naïve (no oral antidiabetes therapy, GLP-1 analog, or insulin for ≥12 weeks [or ≥16 weeks for pioglitazone] prior to randomization) or pretreated with any oral antidiabetes therapy, GLP-1 analog, or insulin for ≥12 weeks prior to randomization. Antidiabetes therapy doses were to have remained unchanged for ≥12 weeks (or ≥16 weeks for pioglitazone) prior to randomization or, for insulin, the dose was not to have been changed within 12 weeks prior to randomization by >10% from the dose at randomization.

Key exclusion criteria included uncontrolled hyperglycemia (plasma glucose >13.3 mmol/L after an overnight fast during the run-in period, confirmed by a second measurement), mean seated SBP ≥160 mmHg and/or mean seated DBP ≥100 mmHg during the run-in period (confirmed by a second measurement), known/suspected secondary hypertension, history/evidence of hypertensive retinopathy or hypertensive encephalopathy, renal impairment (estimated glomerular filtration rate [eGFR] <60 mL/min/1.73 m², using the MDRD equation), or indication of liver disease (serum alanine aminotransferase, aspartate aminotransferase, or alkaline phosphatase more than three times the upper limit of normal) during screening/run-in period; acute coronary syndrome, stroke, or transient ischemic attack within 3 months of consent; use of antiobesity drugs within 3 months of consent or bariatric surgery within 2 years; and any uncontrolled endocrine disorder except type 2 diabetes. Screening took place ≤21 days prior to the first intake of the study drug. All patients provided written informed consent prior to study participation.

Patients were required to have a successful ambulatory BP monitoring (ABPM) measurement prior to randomization. After a 2-week, open-label, placebo run-in, patients were randomized (1:1:1) to receive 10 mg empagliflozin o.d., 25 mg empagliflozin o.d., or placebo double blind for 12 weeks. Patients were followed up for 2 weeks after termination of the trial medication. Patients continued their antihypertensive and antidiabetes background therapy throughout the trial at an unchanged dose and regimen, if possible. Randomization was undertaken using a computer-generated, pseudo-random sequence and an interactive voice and web response system and was stratified by HbA_1c (<8.5%, ≥8.5% [<69 mmol/mol, ≥69 mmol/mol]), previous antihypertensive medications (0, 1, 2), renal function (eGFR ≥90 or 60–89 mL/min/1.73 m²), and region (Europe and Middle East, North America).

Patients underwent 24-h ABPM ≤7 days prior to randomization and at week 12. At both times, patients’ daytime and nighttime activities during the 24 h were to be similar. ABPM at week 12 was conducted after administration of study medication. The ABPM devices were programmed to measure BP and pulse every 20 min.

Changes in antihypertensive medication could be initiated if a patient had a mean SBP ≥160 mmHg and/or a mean DBP ≥100 mmHg at a clinic visit. Rescue medication for hyperglycemia could be initiated at the discretion of the investigator if, after an overnight fast, a patient had plasma glucose >13.3 mmol/L during the first 12 weeks of treatment or >11.1 mmol/L during the follow-up period.

The study protocol was approved by the appropriate institutional review boards and independent ethics committees. The study was undertaken in compliance with the principles of the Declaration of Helsinki in accordance with the International Conference on Harmonisation Harmonised Tripartite Guideline for Good Clinical Practice.

The primary efficacy end point was change from baseline in HbA_1c at week 12. The coprimary end point was change from baseline in mean 24-h SBP at week 12, and the key secondary end point was change from baseline in mean 24-h DBP at week 12, both measured via ABPM (SpaceLabs 90207). Other secondary efficacy end points measured at week 12 were changes from baseline in fasting plasma glucose (FPG), body weight, daytime (6:00 a.m.–9.59 p.m.) and nighttime (10:00 p.m.–5:59 a.m.) mean SBP and DBP (ABPM), and mean seated office SBP and DBP. An additional end point was the use of rescue medication (end point defined as the use of additional antidiabetes medication or an increase in the dose of antidiabetes background medication [for insulin: >10% increase in dose if baseline was ≥40 units/day or >4 units/day increase if baseline was <40 units/day] for ≥7 days or until treatment discontinuation). Safety was assessed based on adverse events (AEs), clinical laboratory tests, and vital signs. AEs of special interest included confirmed hypoglycemic AEs (plasma glucose ≤3.9 mmol/L and/or assistance required), AEs consistent with urinary tract infection (UTI) (based on a prospectively defined search of 70 preferred terms), AEs consistent with genital infection (based on a prospectively defined search of 89 preferred terms), and AEs consistent with volume depletion (based on 8 preferred terms). The change from baseline in the proportion of patients with a positive orthostatic BP test at week 12 was analyzed (defined as positive if the difference between supine and standing readings met any of the following criteria: decrease in SBP ≥20 mmHg, decrease in DBP ≥10 mmHg, increase in pulse rate ≥20 beats per min [bpm]).

A sample size of 272 patients per treatment group would provide a power of ≥90% to detect a treatment difference in change from baseline in 24-h SBP at a significance level of 5% (two-sided), assuming an estimated difference of 4 mmHg (SD 14) and a 5% dropout rate. This sample size would have >95% power to detect a 0.5% difference in HbA_1c, assuming an SD of 1.1%, and 90% power to detect a 2-mmHg difference in mean 24-h DBP assuming an SD of 7 mmHg.

Efficacy analysis was performed in the full analysis set (FAS), which included patients treated with ≥1 dose of study drug who had baseline HbA_1c and baseline mean 24-h SBP values. Values observed after a patient started rescue medication were set to missing; in the context of this imputation, rescue medication was defined as the use of additional antidiabetes medication or a change in the dose of antidiabetes background medication (for insulin: >10% change in dose if baseline was ≥40 units/day or >4 units/day change if baseline was <40 units/day) for ≥7 days. The last observation carried forward (LOCF) approach was used to impute missing continuous efficacy data. For BP results, data after a change in antihypertensive medication were also set to missing before imputing (LOCF-H). Analyses were performed using SAS (version 9.2; SAS Institute, Cary, NC).

The primary end point was assessed using an ANCOVA model with treatment, eGFR, number of antihypertensive medications, and region as fixed effects and baseline HbA_1c as a linear covariate. The coprimary, key secondary, and additional continuous secondary end points were analyzed using the same model, with the baseline value for the end point in question as an additional linear covariate. A post hoc subgroup analysis in patients with uncontrolled BP at baseline (mean 24-h SBP ≥130 mmHg and/or DBP ≥80 mmHg) and controlled BP at baseline was undertaken using the ANCOVA model. Mean 24-h SBP and 24-h DBP were based on the means of hourly means of successful, valid, and usable SBP or DBP measurements (ABPM). Measurements were defined as successful if the start time was between 6:30 and 11:30 a.m., duration was ≥23.5 h, ≥70% of readings were valid (defined as SBP ≥50 to ≤250 mmHg, DBP ≥20 to ≤130 mmHg, pulse pressure ≥15 to ≤150 mmHg, pulse rate ≥20 to ≤200 bpm, and measurement being plausible compared with surrounding values), there were ≥18 h in total with one or more valid reading, and there were ≤6 h in total and ≤2 consecutive hours with no valid readings. Readings within 24 h of the beginning of ABPM were defined as usable. ABPM measurements at baseline or week 12 could be repeated once if unsuccessful.

The impact of methods for handling missing data was assessed through sensitivity analyses including restricted maximum likelihood–based mixed-model repeated-measures analyses of the primary end point.

Safety was assessed in the treated set (patients treated with ≥1 dose of study drug) except for some laboratory parameters that were analyzed in patients from the FAS who had a follow-up visit. LOCF imputation without setting values after rescue therapy to missing was used for analysis of lipid parameters.

Between June 2011 and July 2012, 825 patients with type 2 diabetes and hypertension were randomized and 824 received study medication. Of these, 823 patients were included in the FAS (Supplementary Fig. 1). Patient demographics and baseline characteristics were balanced across treatment groups (Table 1).

Hourly mean SBP and DBP at week 12 are shown in Fig. 1A and B. Reductions from baseline in mean 24-h SBP and mean 24-h DBP were significantly greater with 10 and 25 mg empagliflozin compared with placebo at week 12 (Fig. 1C and D). Patients with 24-h mean ABPM ≥130/80 mmHg at baseline had greater decreases in mean 24-h SBP and mean 24-h DBP compared with placebo at week 12 than those with BP <130/80 mmHg at baseline (Table 2). Reductions from baseline in mean seated office SBP and DBP were significantly greater with 10 and 25 mg empagliflozin compared with placebo at week 12 (Fig. 1E and F), consistent with the ABPM results. Empagliflozin (10 and 25 mg) significantly reduced daytime mean SBP and DBP from baseline compared with placebo at week 12 (Table 2). Reductions from baseline in nighttime mean SBP were significantly greater with 10 and 25 mg empagliflozin compared with placebo at week 12; changes in nighttime DBP were only significant for 25 mg empagliflozin compared with placebo (Table 2). Reductions in nighttime BP were less pronounced than daytime effects. Reductions in BP were not associated with increases in pulse rate; at week 12, mean (SD) changes from baseline in mean 24-h pulse rate (ABPM) were −0.27 (6.10) bpm, −0.17 (7.70) bpm, and −0.74 (6.16) bpm with placebo, 10 mg empagliflozin, and 25 mg empagliflozin, respectively, in the treated set. The proportions of patients who changed antihypertensive medication during the study were 1.1%, 1.1%, and 2.2% in the placebo and 10 and 25 mg empagliflozin groups.

Results from sensitivity analyses confirmed the robustness of the results of the primary and key secondary analyses (Supplementary Table 1).

Reductions from baseline in HbA_1c and FPG were significantly greater with 10 and 25 mg empagliflozin compared with placebo at week 12 (Fig. 2A and Supplementary Table 2). The proportions of patients receiving rescue medication for hyperglycemia were 2.6%, 1.4%, and 2.9% of patients on placebo, 10 mg empagliflozin, and 25 mg empagliflozin, respectively. Reductions in body weight were significantly greater with empagliflozin compared with placebo at week 12 (Fig. 2B).

Results from sensitivity analyses confirmed the robustness of the results of the primary analysis (Supplementary Table 3).

The number of patients with AEs is summarized in Table 3. Most patients (97%) with one or more AEs reported only events that were mild or moderate in intensity. Serious AEs were reported in a higher proportion of patients receiving placebo than empagliflozin (Table 3); one was considered drug-related by the investigator (sudden death in a patient receiving 10 mg empagliflozin). The most frequently reported drug-related AEs (preferred terms) were pollakiuria, hypoglycemia, polyuria, and thirst.

Events consistent with volume depletion were reported by two patients: one receiving placebo (hypotension and orthostatic hypotension) and one receiving 10 mg empagliflozin (syncope [not considered drug related]). The number of patients with a positive orthostatic BP test increased in all groups, with a higher proportion of patients in the empagliflozin groups than the placebo group at week 12 (Table 3). No patients in the empagliflozin groups with a positive orthostatic BP test had an AE that was potentially related to hypotension on the day of the test. Confirmed hypoglycemic AEs were reported in more patients receiving empagliflozin than placebo (Table 3), none of which required assistance or was severe. The percentage of patients with events consistent with UTI was similar in the empagliflozin and placebo groups. All events consistent with UTI were mild or moderate in intensity, and none led to premature discontinuation. The percentage of patients with events consistent with genital infection was higher with empagliflozin than placebo. Events consistent with genital infection were of mild or moderate intensity in all cases and led to premature discontinuation in two patients (both receiving 25 mg empagliflozin).

Small increases in hematocrit were observed in the empagliflozin groups, which returned to near baseline at follow-up (Supplementary Table 4). Small decreases in eGFR were seen in the 25 mg empagliflozin group, which returned to baseline at follow-up (Supplementary Table 4). There were no changes in electrolytes. There were small increases from baseline in total cholesterol and LDL cholesterol with 25 mg empagliflozin compared with placebo. No major differences in mean changes from baseline in HDL cholesterol or triglycerides were noted between placebo and either dose of empagliflozin or in total cholesterol and LDL cholesterol between 10 mg empagliflozin and placebo (Supplementary Table 5).

Hypertension is a common comorbidity in patients with type 2 diabetes and increases the risk of cardiovascular complications (5). This study was undertaken to establish the effect of 10 and 25 mg empagliflozin o.d. for 12 weeks on BP and metabolic control and to assess the safety and tolerability of empagliflozin in patients with type 2 diabetes and hypertension. We conducted a careful assessment of changes in BP using both ABPM and office BP measurements, evaluating data obtained on stable antihypertensive therapy. We assessed 24-h ambulatory BP as the key efficacy outcome, as ABPM avoids the “white coat” effect seen with office BP measurements (13), is more closely correlated with markers of organ damage, and is a more sensitive predictor of cardiovascular outcomes (13,21).

Treatment with empagliflozin for 12 weeks led to significant and clinically meaningful improvements in 24-h SBP and DBP compared with placebo, supported by reductions in daytime and nighttime BP, hourly mean ABPM, and office BP. The risk of cardiovascular disease doubles for each increment of 20 mmHg in SBP or 10 mmHg in DBP across the BP range from 115/75 to 185/115 mmHg (22). Randomized controlled trials have shown reductions in major cardiovascular events when SBP was lowered to <150 mmHg and DBP was lowered to ≤ 85 mmHg in patients with diabetes (23,24). In patients with type 2 diabetes and hypertension, a treatment approach that included control of BP and glycemia significantly reduced the risk of cardiovascular complications and mortality (25).

BP follows a circadian rhythm, with the lowest levels observed at night and peaks during the day (26). A blunted nocturnal BP decline in patients with hypertension is more frequent in patients with obesity, diabetes, and overt cardiovascular or renal disease and associated with a higher cardiovascular risk (27). Empagliflozin maintains the circadian BP rhythm, with higher reductions in daytime versus nighttime BP. Greater reductions in BP were observed in patients with a BP ≥130/80 mmHg (ABPM) compared with those with <130/80 mmHg at baseline (the target recommended by the JNC at the time this study was designed [9]), suggesting that the risk of hypotension in normotensive patients treated with empagliflozin may be low.

The mechanism by which empagliflozin reduces BP has yet to be fully elucidated but may be related to improved glucose control, weight loss, volume contraction due to osmotic diuresis, and improved arterial stiffness (28). Reductions in SBP and DBP have also been observed with other SGLT2 inhibitors (29,30). An ongoing cardiovascular outcome trial (EMPA-REG OUTCOME [clinical trial reg. no. NCT01131676]) will determine the cardiovascular safety of empagliflozin and provide insights into the potential benefits of empagliflozin on cardiovascular and microvascular outcomes in patients with type 2 diabetes at high cardiovascular risk. In conclusion, 10 and 25 mg empagliflozin o.d. were well tolerated and led to clinically meaningful improvements in SBP and DBP in patients with type 2 diabetes and hypertension, in addition to significant reductions in HbA_1c and body weight.

Duality of Interest. This study was funded by Boehringer Ingelheimhttp://dx.doi.org/10.13039/100001003 and Eli Lillyhttp://dx.doi.org/10.13039/100004312. I.T. has received consulting fees/payments for lectures and support for travel to meetings from Boehringer Ingelheim. K.N., C.Z., A.S., U.C.B., and H.J.W. are employees of Boehringer Ingelheim. A.G. works on behalf of Boehringer Ingelheim. Medical writing assistance, supported financially by Boehringer Ingelheim, was provided by Elizabeth Ng 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. I.T. contributed to the acquisition and interpretation of data, drafted the manuscript, was fully responsible for all content and editorial decisions, was involved at all stages of manuscript development, and approved the final version. K.N., U.C.B., and H.J.W. contributed to the study design and interpretation of data, reviewed and edited the manuscript, were fully responsible for all content and editorial decisions, were involved at all stages of manuscript development, and approved the final version. C.Z., A.G., and A.S. contributed to the interpretation of data, reviewed and edited the manuscript, were fully responsible for all content and editorial decisions, were involved at all stages of manuscript development, and approved the final version. I.T. 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. Some of the results of this trial were presented at the 49th Annual Meeting of the European Association for the Study of Diabetes, 23–27 September 2013, Barcelona, Spain, and the American Heart Association Scientific Sessions, 16–20 November 2013, Dallas, TX, and were published in abstract form in the following publications: 1) Tikkanen I, Narko K, Zeller C, Green A, Salsali A, Broedl UB, Woerle HJ. Empagliflozin improves blood pressure in patients with type 2 diabetes (T2DM) and hypertension. Diabetologia 2013;56(Suppl. 1): S377 and 2) Tikkanen I, Narko K, Zeller C, Green A, Salsali A, Broedl UB, Woerle HJ. Empagliflozin improves 24-hour blood pressure profiles in patients with type 2 diabetes and hypertension. Circulation 2013;128:A14501.