Renal functional reserve predicts GFR response to empagliflozin in RENALIS and RACELINES clinical trials

Abstract Background and hypothesis Glomerular hyperfiltration is common in type 2 diabetes (T2D) and may reflect reduced nephron number and/or perturbed intrarenal hemodynamics. Renal functional reserve (RFR), the capacity to increase GFR with physiological stimuli (e.g. a meal), may help reveal single-nephron hyperfiltration in patients with preserved baseline whole-kidney GFR. We hypothesized that reduced postprandial RFR predicts the acute hemodynamic GFR-response in T2D to the sodium-glucose co-transporter-2 inhibitor empagliflozin, but not to the dipeptidyl-peptidase-4 inhibitor linagliptin or a sulfonylurea. Methods This analysis pooled data from two 8-week randomized, double-blind, parallel-group mechanistic trials, encompassing 71 T2D patients with preserved whole-kidney GFR (mean ± SD age 65 ± 7 yrs, BMI 30.4 ± 3.9 kg/m2, HbA1c 7.8 ± 1.0% 61.6 ± 11.0 mmol/mol, GFR 86.5 ± 17.6 mL/min/1.73m2). Patients received empagliflozin (10 mg; N = 20), linagliptin (5 mg; N = 27) or sulfonylurea (glimepiride 1 mg, or gliclazide 30 mg; N = 24), in addition to metformin. Measured (m)GFR and effective renal plasma flow (ERPF) were determined by inulin/iohexol and PAH-clearance, respectively, based on timed urine-sampling in fasting and post-protein-rich meal conditions. Intrarenal-hemodynamics were calculated using Gomez-equations; fractional sodium-excretion (FENa) and systemic hemodynamics were evaluated. Results The meal increased mGFR (+7.3 ± 1.7 mL/min/1.73m2; P 0.001) and ERPF (+44.3 ± 14.9 mL/min/1.73m2; P = 0.005), with concomitant decrease in renal vascular resistance (RVR; −0.02 ± 0.01 mmHg/L/min; P 0.001), driven by reduced afferent arteriolar resistance (−1068 ± 241dyne/sec/cm−5; P 0.001) and lower FENa (−0.21 ± 0.05; P 0.001). Postprandial mGFR-changes did not correlate with baseline mGFR, but did correlate with postprandial RVR-change (r − 0.57; P 0.001). After 8 weeks, mGFR tended to decrease with sulfonylurea (P = 0.054) and decreased with empagliflozin (−9.1 ± 3.2 mL/min/1.73m2; P = 0.016), but not with linagliptin. Baseline meal-induced mGFR-changes correlated with 8-week treatment-induced mGFR-changes with empagliflozin (r:0.88; P 0.001), but not with linagliptin or sulfonylurea. Conclusion Baseline postprandial-RFR predicts GFR-dipping with empagliflozin after 8-weeks, but not with GFR-changes following linagliptin or sulfonylurea. As initial GFR-dipping is associated with long-term kidney benefit, RFR warrants evaluation as a potential additional tool to personalize SGLT2i-therapy.