Effect of Dapagliflozin on Measured vs. Panel-Estimated Glomerular Filtration Rate

Abstract

Sodium–glucose cotransporter 2 (SGLT2) inhibitors can cause a reversible decline in glomerular filtration rate (GFR), which may influence dosing recommendations for renally excreted medications. In practice, GFR is typically estimated by serum creatinine concentration, but creatinine may not be a reliable indicator of GFR decline in the setting of SGLT2 inhibitor use. Alternative filtration markers such as cystatin C, β-trace protein (BTP), and β2-microglobulin (B2M) may be more appropriate, but little is known about how these markers are affected by SGLT2 inhibitor use. Therefore, we determined creatinine, cystatin C, BTP, and B2M concentration in a crossover study of 35 people with type 2 diabetes receiving 12 weeks of dapagliflozin treatment or placebo. Estimated GFR (eGFR) based on creatinine (eGFRcre), cystatin C (eGFRcys), their combination (eGFRcomb), or a panel of all four markers (eGFRpanel) was compared with measured GFR (mGFR) based on plasma clearance of chromium-51 labeled ethylenediamine tetraacetic acid (⁵¹Cr-EDTA). Dapagliflozin treatment was associated with a significant decrease in mGFR (−9 mL/min/1.73 m², P < 0.001) but not a corresponding increase in concentration of any filtration marker. No eGFR equation accurately predicted change in mGFR between treatment periods, but eGFRcomb and eGFRpanel yielded the highest overall accuracy relative to mGFR across both treatment periods. These findings highlight the stability in performance gained by combining multiple filtration markers but suggest that eGFR in general is not an ideal metric for assessing short-term GFR decline in people initiating SGLT2 inhibitor therapy.