Recent large trials on sodium–glucose cotransporter 2 (SGLT2) inhibitors showed that these agents improved cardiovascular outcomes in patients with type 2 diabetes (T2D) with a high risk of cardiovascular events (1). However, the underlying mechanisms of these clinical benefits and vascular effect of SGLT2 inhibitors in that patient group remain uncertain. Endothelial dysfunction is exacerbated by metabolic disorders, such as diabetes, and involved in the pathophysiology of diabetes-related cardiovascular complications (2,3). We therefore investigated whether empagliflozin added to standard therapy, compared with placebo, affected endothelial function in patients with T2D and established cardiovascular disease (CVD). This was a prospective, multicenter, randomized, placebo-controlled, double-blind clinical trial undertaken in 16 centers in Japan (clinical trial reg. no. UMIN000024502, www.umin.ac.jp/ctr/index.htm) (4). A total of 117 adults with T2D and established CVD were randomized (1:1) to receive either empagliflozin 10 mg daily or placebo for 24 weeks. Randomization was performed by using the web-based minimization dynamic allocation method stratified according to HbA1c, age, systolic blood pressure (BP), and smoking status. The placebo was indistinguishable from empagliflozin, and both study drugs were sequentially numbered and concealed. The participants and investigators remained masked to group assignments until the database lock (an action taken to prevent further changes to a clinical trial database). The effects of treatment on endothelial function were assessed by the reactive hyperemia peripheral arterial tonometry index (RHI) (5), with the primary efficacy end point being the change in RHI from baseline to 24 weeks. Ethics approval and informed patient consent were obtained.
Of the 117 patients, 105 were included in the full analysis set. Ten patients dropped out prior to initiation of the study drug, and two patients in the placebo group were excluded due to a serious protocol deviation (lack of data necessary for eligibility assessment). The baseline characteristics were as follows: mean (SD) age 64.9 (10.4) years, BMI 26.4 (5.3) kg/m2, BP 133.2/75.7 (15.0/10.5) mmHg, diabetes duration 13.2 (10.9) years, and HbA1c 7.2% (0.8%) (55 [9.0] mmol/mol); proportion of the patients who had previously taken dipeptidyl peptidase 4 inhibitors 70%, metformin 51%, thiazolidinedione 24%, and insulin 10%. All patients had at least one CVD (myocardial infarction 24%, angina 31%, stroke 20%, and heart failure 40%), and 77% of patients had been receiving treatment for hypertension (66% on ACE inhibitors or angiotensin receptor blockers, 49% on calcium channel blockers, and 36% on β-blockers) or dyslipidemia (75% on statins).
Twenty-four weeks of empagliflozin treatment significantly improved glycemic control (mean [SD] change volume of HbA1c −0.3% [0.5%] [−2.7 (5.3) mmol/mol] vs. 0.1% [0.7%] [0.7 (7.8) mmol/mol], P = 0.011, and of fasting blood glucose −17.9 (22.0) mg/dL vs. −0.8 (37.6) mg/dL, P = 0.007). Reductions in BMI in the empagliflozin group were significantly greater than those in the placebo group (mean [SD] −0.8 [1.0] vs. −0.2 [0.8] kg/m2, P = 0.002). There were borderline differences in the changes in BP between the treatment groups (systolic −7.6 [16.5] vs. −2.1 [12.1] mmHg, P = 0.063; diastolic −3.7 [8.7] vs. −0.2 [9.9] mmHg, P = 0.058).
RHI values at baseline and 24 weeks are shown in Fig. 1. Absolute changes in RHI from baseline to 24 weeks were −0.006 (SD 0.478) in the empagliflozin group and −0.025 (0.454) in the placebo group. The ANCOVA model adjusted for allocation factors demonstrated no significant intergroup difference in the changes in RHI (adjusted mean difference −0.020, 95% CI −0.199 to 0.158, P = 0.821). Likewise, in all the prespecified subgroups (e.g., age <65 or ≥65 years, HbA1c <7.0% [53 mmol/mol] or ≥7.0%, systolic BP <140 or ≥140 mmHg, eGFR <60 or ≥60 mL/min/1.73 m2, history of heart failure, and median RHI at baseline <1.77 or ≥1.77), no significant difference in changes in RHI was observed between the treatment groups.
To our knowledge, this is the first multicenter, randomized, placebo-controlled, double-blind trial to evaluate the effect of empagliflozin on endothelial function in patients with T2D and established CVD as a possible surrogate marker of the cardiovascular benefits from empagliflozin. This observation suggests that 24 weeks of empagliflozin treatment is not associated with an improvement in endothelial function in that population and that cardiovascular benefits observed in large trials (1) may have been attributable, at least in the early phase of treatment, to mechanisms other than amelioration of endothelial dysfunction. One limitation is that participants in our study had relatively well-controlled clinical conditions, such as BP, BMI, and HbA1c, partly differing from those in the large trials. Additionally, we had no measurement of flow-mediated vasodilation, which has been validated against hard outcomes in several studies. Further studies are needed to test whether SGLT2 inhibitors affect endothelial function in other T2D populations.
Acknowledgments. The authors thank the investigators, coordinators, and patients who participated in the trial. They also thank Takeshi Morimoto (Hyogo College of Medicine, Nishinomiya, Japan), Mitsuyoshi Urashima (Jikei University School of Medicine, Tokyo, Japan), and Takashi Nomiyama (Fukuoka University, Fukuoka, Japan), who served as members of the independent data and safety monitoring committee for the trial.
Duality of Interest. This study was funded by Boehringer Ingelheim and Eli Lilly and Company. A.T. received modest honoraria from Astellas, AstraZeneca, Boehringer Ingelheim, Daiichi Sankyo, Kowa, Merck, Mitsubishi Tanabe, Novo Nordisk, Taisho Toyama, Takeda, and Teijin. M.S. received an honorarium and endowed chair from Boehringer Ingelheim. H.T. received lecture fees from Bayer, Boehringer Ingelheim, Daichi Sankyo, Kowa, Takeda, Mitsubishi Tanabe, and Sanwa Kagaku Kenkyusho. Y.O. received lecture fees from Astellas, AstraZeneca, Merck Sharp & Dohme (MSD), Ono, Mitsubishi Tanabe, Bayer, Novo Nordisk, Eli Lilly and Company, Boehringer Ingelheim, Daiichi Sankyo, Kissei, Novartis, Kowa, and Sanwa Kagaku Kenkyusho and research funds from Kowa and Mitsubishi Tanabe. T.T. received honoraria from MSD, Astellas, AstraZeneca, Mitsubishi Tanabe, Boehringer Ingelheim, Novo Nordisk, and Taisho Toyama; research funding from Kowa; and scholarships from Novartis, AstraZeneca, Astellas, and Novo Nordisk. M.Y-T. received honoraria from Bayer, Mitsubishi Tanabe, Itamar, MSD, Nippon Shinyaku, Boehringer Ingelheim, and Daiichi Sankyo. S.U. received research grants from Bristol-Myers Squibb and Kowa; nonpurpose research grants from Bristol-Myers Squibb, Chugai, MSD, Pfizer, and Takeda; and lecture fees from Boehringer Ingelheim and MSD. Y.H. received consulting fees from Mitsubishi Tanabe related to this study, as well as honoraria and grants from Teijin, Boehringer Ingelheim, MSD, Sanofi, AstraZeneca, Kyowa Hakko Kirin, Takeda, Astellas, Daiichi Sankyo, Mochida, Nihon Kohden, Shionogi, Nippon Sigmax, Sanwa Kagaku Kenkyusho, Unex, and Kao and honoraria from Radiometer, Omron, Sumitomo Dainippon, Otsuka, Torii, Kowa, Fujiyakuhin, Amgen, Nippon Shinyaku, Itamar, Bayer, Eli Lilly and Company, and Ono. K.N. received honoraria from Eli Lilly and Company, Astellas, Ono, Takeda, Daiichi Sankyo, Boehringer Ingelheim, MSD, Mitsubishi Tanabe, and AstraZeneca; research grants from Amgen, Teijin, Terumo, Mitsubishi Tanabe, Asahi Kasei, Astellas, Boehringer Ingelheim, and Bayer; and scholarships from Bayer, Daiichi Sankyo, Teijin, Astellas, Takeda, and Bristol-Myers Squibb. No other potential conflicts of interest relevant to this article were reported.
Author Contributions. A.T. participated in the study design, operations, and analysis and interpretation of data and drafted the manuscript. M.S. helped to steer the study, collected and analyzed the data, and reviewed and edited the manuscript. N.M., H.T., Y.O., K.R.S., T.T, I.T., I.H., S.T., Y.M., H.T., M.Y.-T., H.Y., Y.S., Y.I., S.U., Y.H., and K.N. contributed to the study concept and design, study operations, data collection, and analysis and interpretation of data and critically reviewed and edited the manuscript. H.Y. contributed to the generation of the scheme for randomization. Y.S. contributed to the statistical analysis. S.U. and Y.H. helped to steer the study as research advisors and reviewed and edited the draft manuscript. K.N. was a principal investigator of the EMBLEM trial. K.N. 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 55th Annual Meeting of the European Association for the Study of Diabetes, Barcelona, Spain, 16–20 September 2019.
- Received June 14, 2019.
- Accepted July 2, 2019.
- © 2019 by the American Diabetes Association.