Effects of empagliflozin on acute kidney outcomes not modified by acute eGFR dips

ERA 2025 – The treatment effect of empagliflozin on acute kidney outcomes were not modified by the predicted acute eGFR dip, in an individual participant-level meta-analysis among patients with T2D, HF and CKD.

This summary is based on the presentation of Natalie Staplin, PhD (Oxford, UK) at the ERA Congress 2025 – Impact of acute eGFR dips and markers of disease severity on effects of empagliflozin on acute kidney outcomes.

Introduction and methods

Previous studies have shown that SGLT2 inhibitors reduce the risk of acute kidney injury (AKI). Still, AKI is listed as an adverse effect on the label of SGLT2 inhibitors. The aim of the present meta-analysis was to investigate whether the size of acute eGFR dips impacts treatment adherence to SGLT2 inhibitors or modifies the treatment effects of SGLT2 inhibitors.

The current analysis included individual-level data from 23,340 participants of 4 large empagliflozin RCTs with systematic eGFR measurements during follow-up. Patients from EMPA-REG OUTCOMES (patients with T2D; n=7020), EMPEROR Program (2 trials with patients with HF; n=9711) and EMPA-KIDNEY (patients with CKD; n=6609) were included. An inverse variance weighted approach was used to determine the effect modification on acute kidney outcomes using heterogeneity or trend tests for key subgroup analyses.

Firstly, a predictive model for post-randomization acute eGFR dips that utilized only baseline characteristics was developed using data from patients randomized to empagliflozin. The model showed good discrimination and calibration.

Subsequently, this model was used to predict acute eGFR dips for all participants. In the overall study population, the mean predicted acute dip in eGFR was -4.9% ± 2.4. The largest group of participants (51%) had a predicted eGFR dip of ≥-5%, 28% of participants had a predicted dip of ≥-7% to <-5%, 9% of participants had a predicted dip of ≥-8% to <-7%, and 9% of participants had a predicted dip of <-8%. Data was missing for 4% of the participants.

Two acute kidney outcomes were included in this meta-analysis: (1) ≥50% increase in serum creatinine in consecutive follow-up samples (local or central, and taken <1 year previously), and (2) AKI, as defined by the MedDRA Preferred Term.

Main results

Predictors of acute eGFR dip

• After adjusting for other variables in the model, the independent predictors of a larger acute eGFR dip included older age, higher eGFR, higher UCAR, diuretic use, RAASi use, higher SBP, and a higher BMI.

Treatment adherence

• In the overall population, the proportion of participants who had stopped study treatment at 12 weeks was comparable between the empagliflozin group and the placebo group (both 1.8%).
• The differences between the empagliflozin and placebo groups were small when stratified by the predicted eGFR dip subgroups (1.5% vs. 1.6% in the smallest dip group; 2.4% vs. 1.8% in the ≥-7% to <-5% group; 1.8% vs. 2.3% in the ≥-8% to <-7% group; and 2.1% vs. 1.8% in the largest dip group, respectively).

Treatment effect

• Compared with placebo, empagliflozin reduced the risk of ≥50% increase in serum creatinine in consecutive follow-up samples in the overall study population (HR: 0.80; 95%CI: 0.72–0.88).
• The treatment effect of empagliflozin on the increase in serum creatinine was not modified by eGFR, predicted acute eGFR dip, cause of kidney disease, diabetes, or UCAR (P for heterogeneity or trend test all >0.05).
• Empagliflozin versus placebo reduced the risk of AKI adverse events by 27% (HR: 0.73; 95%CI: 0.63–0.85).
• The treatment effect of empagliflozin on AKI events was not modified by any of the key subgroups.

Conclusion

This individual participant-level meta-analysis of 4 RCTs showed that participants with a larger acute eGFR dip were only slightly more likely to discontinue study drug treatment. Treatment with empagliflozin reduced the risk of a ≥50% increase in serum creatinine in consecutive follow-up samples and acute kidney injury adverse events, compared with placebo. The treatment effect of empagliflozin was not modified by any subgroups, including by the predicted eGFR dip.

- Our reporting is based on the information provided at the ERA Congress 2025 -