Increase in endogenous glucose production with SGLT2 inhibition is attenuated in individuals who underwent kidney transplantation and bilateral native nephrectomy

AIMS/HYPOTHESIS: The glucosuria induced by sodium–glucose cotransporter 2 (SGLT2) inhibition stimulates endogenous (hepatic) glucose production (EGP), blunting the decline in HbA(1c). We hypothesised that, in response to glucosuria, a renal signal is generated and stimulates EGP. To examine the effect of acute administration of SGLT2 inhibitors on EGP, we studied non-diabetic individuals who had undergone renal transplant with and without removal of native kidneys. METHODS: This was a parallel, randomised, double-blind, placebo-controlled, single-centre study, designed to evaluate the effect of a single dose of dapagliflozin or placebo on EGP determined by stable-tracer technique. We recruited non-diabetic individuals who were 30–65 years old, with a BMI of 25–35 kg/m(2) and stable body weight (±2 kg) over the preceding 3 months, and HbA(1c) <42 mmol/mol (6.0%). Participants had undergone renal transplant with and without removal of native kidneys and were on a stable dose of immunosuppressive medications. Participants received a single dose of dapagliflozin 10 mg or placebo on two separate days, at a 5- to 14-day interval, according to randomisation performed by our hospital pharmacy, which provided dapagliflozin and matching placebo, packaged in bulk bottles that were sequentially numbered. Both participants and investigators were blinded to group assignment. RESULTS: Twenty non-diabetic renal transplant patients (ten with residual native kidneys, ten with bilateral nephrectomy) participated in the study. Dapagliflozin induced greater glucosuria in individuals with residual native kidneys vs nephrectomised individuals (8.6 ± 1.1 vs 5.5 ± 0.5 g/6 h; p = 0.02; data not shown). During the 6 h study period, plasma glucose decreased only slightly and similarly in both groups, with no difference compared with placebo (data not shown). Following administration of placebo, there was a progressive time-related decline in EGP that was similar in both nephrectomised individuals and individuals with residual native kidneys. Following dapagliflozin administration, EGP declined in both groups, but the differences between the decrement in EGP with dapagliflozin and placebo in the group with bilateral nephrectomy (Δ = 1.11 ± 0.72 μmol min(−1) kg(−1)) was significantly lower (p = 0.03) than in the residual native kidney group (Δ = 2.56 ± 0.33 μmol min(−1) kg(−1)). In the population treated with dapagliflozin, urinary glucose excretion was correlated with EGP (r = 0.34, p < 0.05). Plasma insulin, C-peptide, glucagon, prehepatic insulin:glucagon ratio, lactate, alanine and pyruvate concentrations were similar following placebo and dapagliflozin treatment. β-Hydroxybutyrate increased with dapagliflozin treatment in the residual native kidney group, while a small increase was observed only at 360 min in the nephrectomy group. Plasma adrenaline (epinephrine) did not change after dapagliflozin and placebo treatment in either group. Following dapagliflozin administration, plasma noradrenaline (norepinephrine) increased slightly in the residual native kidney group and decreased in the nephrectomy group. CONCLUSIONS/INTERPRETATION: In nephrectomised individuals, the hepatic compensatory response to acute SGLT2 inhibitor-induced glucosuria was attenuated, as compared with individuals with residual native kidneys, suggesting that SGLT2 inhibitor-mediated stimulation of hepatic glucose production via efferent renal nerves occurs in an attempt to compensate for the urinary glucose loss (i.e. a renal–hepatic axis). TRIAL REGISTRATION: ClinicalTrials.gov NCT03168295 FUNDING: This protocol was supported by Qatar National Research Fund (QNRF) Award No. NPRP 8-311-3-062 and NIH grant DK024092-38. Graphical abstract [Image: see text] ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1007/s00125-020-05254-w) contains supplementary material, which is available to authorized users.