Kidney outcomes in patients with diabetes mellitus did not differ between individual sodium-glucose cotransporter-2 inhibitors

Abstract

Data comparing kidney outcomes between individual sodium-glucose cotransporter-2 (SGLT2) inhibitors are limited. Here, we aimed to compare the subsequent risk of developing kidney outcomes between individual inhibitors. This would be the first study to compare kidney outcomes of patients with diabetes mellitus who were newly treated with individual SGLT2 inhibitors using a large-scale real-world dataset. To do this, we analyzed results from 12,100 patients with diabetes mellitus who were taking different SGLT2 inhibitors (2,573 with empagliflozin; 2,214 with dapagliflozin; 2,100 with canagliflozin; and 5,213 with other such inhibitors). The primary outcome was the rate of estimated glomerular filtration rate (eGFR) decline as assessed using a linear mixed-effects model with an unstructured covariance. The median age of the patients was 53 years, and 84.4% of the patients were men. The median fasting plasma glucose and HbA1c levels were 147 (interquartile range 126-178) mg/dL and 7.5 (6.9-8.4)%, respectively. The median eGFR was 78 mL/min/1.73 m² (interquartile range 68-90). The mean follow-up period was 773 days. The annual eGFR slopes of empagliflozin, dapagliflozin, canagliflozin, and other SGLT2 inhibitors were -1.15 (95% confidence interval, -1.33 to -0.96), -1.14 (-1.32 to -0.96), -1.24 (-1.44 to -1.04), and -1.06 (-1.18 to -0.94) ml/min/1.73 m², respectively. No significant interaction was detected between the SGLT2 inhibitors and time using a linear mixed-effects model. A multitude of sensitivity analyses confirmed the robustness of our primary results. Thus, we found that there was no significant difference in the annual eGFR decline slopes between patients taking different SGLT2 inhibitors.

Keywords: SGLT2 inhibitor; diabetes mellitus; estimated glomerular filtration rate decline.