Sodium-Glucose Cotransporter-2 Inhibitor Therapy and Longitudinal Changes in Kidney Function among Veterans with Autosomal Dominant Polycystic Kidney Disease.

KEY POINTS

Sodium-glucose cotransporter-2 inhibitor (SGLT2i) use has not been extensively studied in autosomal dominant polycystic kidney disease (ADPKD). In veterans with ADPKD, kidney function declined at a greater rate during the first 3 months post-SGLT2i initiation but subsequently attenuated. In veterans with ADPKD, SGLT2i versus dipeptidyl peptidase-4 inhibitor use was associated with slower long-term kidney function decline.

BACKGROUND

Sodium-glucose cotransporter-2 inhibitors (SGLT2is) are a pillar of kidney disease therapy, but their efficacy remains unknown in autosomal dominant polycystic kidney disease (ADPKD). We evaluated effects of SGLT2i on kidney function in ADPKD.

METHODS

This retrospective cohort study within the Veterans Health Administration included adults with an ADPKD diagnosis code who initiated SGLT2i between January 2017 and May 2023. Repeated measures models were used to evaluate eGFR slope before and after SGLT2i initiation. Among patients with ADPKD and type 2 diabetes mellitus, a target trial emulation was used to compare the effects of SGLT2i versus dipeptidyl peptidase-4 inhibitor (DPP4i) on eGFR slope.

RESULTS

Among 348 eligible patients with ADPKD who started an SGLT2i, 93% were male, mean±SD age was 68±11, and median eGFR was 53 (interquartile range: 16–127) ml/min per 1.73 m. In adjusted analyses, the preinitiation eGFR slope was −0.79 (95% confidence interval, −1.26 to −0.33) ml/min per 1.73 m per 90 days. The eGFR slope steepened to −2.78 (−4.04 to −1.53) ml/min per 1.73 m during the first 3 months postinitiation and then stabilized to −0.07 (−0.72 to 0.58) ml/min per 1.73 m during months 3–12 postinitiation. The target trial emulation compared 217 SGLT2i users with 198 DPP4i users. In adjusted analyses, eGFR declined −4.03 (−6.45 to −1.60) ml/min per 1.73 m per 90 days faster in SGLT2i users versus DPP4i users during the first 3 months postinitiation; however, during the subsequent 3–12 months, the slope was more stable in SGLT2i initiators than DPP4i initiators, with a difference of 1.29 (0.16 to 2.41) ml/min per 1.73 m per 90 days.

CONCLUSIONS

In older patients with mild ADPKD and a high prevalence of diabetes and cardiovascular disease who initiated an SGLT2i, there was an initial 3-month decline in eGFR followed by stabilization during the remainder of the year-long follow-up. Compared with DPP4i use, SGLT2i use was associated with a slower eGFR decline between 3 and 12 months postinitiation in patients with concurrent type 2 diabetes mellitus. These findings suggest that SGLT2is are potentially beneficial in older individuals with ADPKD in whom comorbid disease may play a greater role in kidney function decline, but further studies are required.