The Role of Glucagon in the Acute Therapeutic Effects of SGLT2 Inhibition

Abstract

Sodium-glucose cotransporter 2 inhibitors (SGLT2i) effectively lower plasma glucose (PG) concentration in patients with type 2 diabetes, but studies have suggested that circulating glucagon concentrations and endogenous glucose production (EGP) are increased by SGLT2i, possibly compromising their glucose-lowering ability. To tease out whether and how glucagon may influence the glucose-lowering effect of SGLT2 inhibition, we subjected 12 patients with type 2 diabetes to a randomized, placebo-controlled, double-blinded, crossover, double-dummy study comprising, on 4 separate days, a liquid mixed-meal test preceded by single-dose administration of either 1) placebo, 2) the SGLT2i empagliflozin (25 mg), 3) the glucagon receptor antagonist LY2409021 (300 mg), or 4) the combination empagliflozin + LY2409021. Empagliflozin and LY2409021 individually lowered fasting PG compared with placebo, and the combination further decreased fasting PG. Previous findings of increased glucagon concentrations and EGP during acute administration of SGLT2i were not replicated in this study. Empagliflozin reduced postprandial PG through increased urinary glucose excretion. LY2409021 reduced EGP significantly but gave rise to a paradoxical increase in postprandial PG excursion, which was annulled by empagliflozin during their combination (empagliflozin + LY2409021). In conclusion, our findings do not support that an SGLT2i-induced glucagonotropic effect is of importance for the glucose-lowering property of SGLT2 inhibition.

Trial registration: ClinicalTrials.gov NCT02792400.

© 2020 by the American Diabetes Association.

Figures

Figure 1

Fasting and postprandial glucose concentrations. PG concentrations during liquid MMT (A), fasting PG concentrations (B), AUC (C), and bsAUC (D) on days with placebo, the SGLT2i empagliflozin (25 mg), the GRA (300 mg LY2409021), and the combination (GRA + SGLT2i) in patients with type 2 diabetes (N = 12). Data are mean ± SEM (symbols ± error bars) (A) and mean ± SEM (bars ± error bars) with individual values (symbols) (B–D). Statistical comparisons were made with a linear mixed model, and P values illustrated are raw with P values adjusted for multiple comparisons by false discovery rate in parentheses. ns, not significant.

Figure 2

Glucose kinetics and urinary glucose excretion. Ra of total glucose (A), oral glucose (C), EGP (E), and Rd of glucose (G) with summarized mmol of glucose appearing or disappearing during the 240-min MMT (B, D, F, and H); urine volume (I); and urine glucose excretion (J) in patients with type 2 diabetes (N = 12). Data are mean ± SEM (symbols ± error bars) (A, C, E, and G) and mean ± SEM (bars ± error bars) with individual values (symbols) (B, D, F, and H–J). Statistical comparisons were made with a linear mixed model, and P values illustrated are raw with P values adjusted for multiple comparisons by false discovery rate in parentheses. ns, not significant.

Figure 3

Glucagon, C-peptide, and acetaminophen. Concentrations of glucagon (A), C-peptide (C), and acetaminophen (E) before and during the MMT in patients with type 2 diabetes (N = 12); data are mean ± SEM (symbols ± error bars). Summarized results as AUCs; data are mean ± SEM (bars ± error bars) with individual values (symbols) (B, D, and F). Statistical comparisons were made with a linear mixed model and P values illustrated are raw with P values adjusted for multiple comparisons by false discovery rate in parentheses. ns, not significant.

Figure 4

Glycerol concentrations and kinetics. Plasma glycerol concentrations during liquid MMT (A), Ra of glycerol (B), and Rd of glycerol (C) in patients with type 2 diabetes (N = 12). Data are mean ± SEM.