Effect of empagliflozin on ectopic fat stores and myocardial energetics in type 2 diabetes: the EMPACEF study

Abstract

Background: Empagliflozin is a sodium-glucose cotransporter 2 (SGLT2) inhibitor that has demonstrated cardiovascular and renal protection in patients with type 2 diabetes (T2D). We hypothesized that empaglifozin (EMPA) could modulate ectopic fat stores and myocardial energetics in high-fat-high-sucrose (HFHS) diet mice and in type 2 diabetics (T2D).

Methods: C57BL/6 HFHS mice (n = 24) and T2D subjects (n = 56) were randomly assigned to 12 weeks of treatment with EMPA (30 mg/kg in mice, 10 mg/day in humans) or with placebo. A 4.7 T or 3 T MRI with ¹H-MRS evaluation-myocardial fat (primary endpoint) and liver fat content (LFC)-were performed at baseline and at 12 weeks. In humans, standard cardiac MRI was coupled with myocardial energetics (PCr/ATP) measured with ³¹P-MRS. Subcutaneous (SAT) abdominal, visceral (VAT), epicardial and pancreatic fat were also evaluated. The primary efficacy endpoint was the change in epicardial fat volume between EMPA and placebo from baseline to 12 weeks. Secondary endpoints were the differences in PCr/ATP ratio, myocardial, liver and pancreatic fat content, SAT and VAT between groups at 12 weeks.

Results: In mice fed HFHS, EMPA significantly improved glucose tolerance and increased blood ketone bodies (KB) and β-hydroxybutyrate levels (p < 0.05) compared to placebo. Mice fed HFHS had increased myocardial and liver fat content compared to standard diet mice. EMPA significantly attenuated liver fat content by 55%, (p < 0.001) but had no effect on myocardial fat. In the human study, all the 56 patients had normal LV function with mean LVEF = 63.4 ± 7.9%. Compared to placebo, T2D patients treated with EMPA significantly lost weight (- 2.6 kg [- 1.2; - 3.7]) and improved their HbA1c by 0.88 ± 0.74%. Hematocrit and EPO levels were significantly increased in the EMPA group compared to placebo (p < 0.0001, p = 0.041). EMPA significantly increased glycosuria and plasma KB levels compared to placebo (p < 0.0001, p = 0.012, respectively), and significantly reduced liver fat content (- 27 ± 23 vs. - 2 ± 24%, p = 0.0005) and visceral fat (- 7.8% [- 15.3; - 5.6] vs. - 0.1% [- 1.1;6.5], p = 0.043), but had no effect on myocardial or epicardial fat. At 12 weeks, no significant change was observed in the myocardial PCr/ATP (p = 0.57 between groups).

Conclusions: EMPA effectively reduced liver fat in mice and humans without changing epicardial, myocardial fat or myocardial energetics, rebutting the thrifty substrate hypothesis for cardiovascular protection of SGLT2 inhibitors. Trial registration NCT, NCT03118336. Registered 18 April 2017, https://clinicaltrials.gov/ct2/show/NCT03118336.

Keywords: 31P-MRS; Ectopic fat; Epicardial adipose tissue; MRI; Myocardial energetics; Pcr/atp; SGLT2 inhibitors.

Fig. 1

Study design and flow chart. a Mice design. b Human design. IPGTT, intraperitoneal glucose tolerance test; MRI, magnetic resonance imaging; MRS, magnetic resonance spectroscopy

Fig. 2

Changes in metabolic profile and ectopic fat stores at 12 weeks of treatment in mice. a Metabolic profile: Weight of mice during 12 weeks of standard diet (SD), High Fat High Sucrose diet (HFHS) or HFHS + empagliflozin (EMPA); glycosuria at 4 and 12 weeks in SD, HFHS and EMPA mice; mean daily water intake (mL/day) in SD, HFHS and EMPA mice measured every 2 days for 12 weeks; intraperitoneal glucose tolerance test (IPGTT) in SD, HFHS and EMPA mice measured at week 0, 4 and 12 of empagliflozin treatment; HOMA-IR index and insulin/glucagon ratio of SD, HFHS and EMPA mice measured at 12 weeks. b Ectopic fat stores: Evolution of typical ¹H-MRS spectra (triglycerides peak at 1.3 ppm) acquired in the heart or liver of SD, HFHS and EMPA mice at week 0, 4 and 12; Representative Oil-Red-O stain images (n = 10 images per mouse) of lipid droplets in heart and liver from SD, HFHS or EMPA mice; scale bars, 100 μm for heart and 50 µm for liver. c Ketone pathways: Ketone bodies and fasting β-hydroxybutyrate of SD, HFHS and EMPA mice measured at 4 weeks; RT-qPCR analyses of genes coding for enzymes involved in β-oxidation pathway (Bdh1, Bdh2, Hmgcs2, Oxct1) of SD, HFHS and EMPA mice after 12 weeks. SD (n = 10), HFHS (n = 12), EMPA (n = 12). Data represent mean values ± SEM. Differences between SD, HFHS and EMPA were determined by nonparametric unpaired Mann–Whitney comparative tests or ANOVA with Tukey's post hoc test; *p < 0.05, **p < 0.01, ***p < 0.001 (from post hoc test)

Fig. 3

Changes in myocardial energetics, metabolic profile and ectopic fat stores at 12 weeks of treatment in humans. a Changes in glycosuria, blood ketone body levels, hematocrit and PCr/ATP ratio obtained by ³¹P-magnetic resonance spectroscopy in placebo and EMPA group. b Liver and pancreatic fat content assessed with ¹H-magnetic resonance spectroscopy change between baseline and 12 weeks in placebo and EMPA group. c Visceral and subcutaneous abdominal fat change measured at L4-L5 intervertebral disk level; Epicardial fat volume assessed on cine MRI short axis slices and myocardial fat assessed with ¹H-magnetic resonance spectroscopy at the interventricular septum with a single-voxel PRESS sequence. Placebo (n = 16), EMPA (n = 18). Data represent mean values ± SEM. Differences between baseline and 12 weeks were determined by paired t-test/Wilcoxon test and differences between groups with ANOVA; *p < 0.05, **p < 0.01, ***p < 0.001