Abstract
BACKGROUND: Autosomal dominant polycystic kidney disease (ADPKD) is the fourth leading cause of kidney failure in Canada and internationally. To date, patients with ADPKD have been excluded from trials of sodium-glucose cotransporter type 2 inhibitors (SGLT2i), which have been demonstrated to positively influence a wide range of kidney outcomes across the spectrum of chronic kidney disease (CKD). This exclusion was primarily due to theoretic safety concerns, particularly hastening disease progression due to vasopressin stimulation. As a result, there is a paucity of data on SGLT2i use among patients with ADPKD. OBJECTIVES: To estimate the risk of kidney dysfunction with SGLT2i treatment among patients with ADPKD. DESIGN: Single-arm retrospective cohort study. SETTING AND PATIENTS: Adult patients (≥18 years old) with CKD with a primary diagnosis of ADPKD in British Columbia, Canada who had been exposed to any drug formulation containing empagliflozin, dapagliflozin or canagliflozin. METHODS AND MEASUREMENTS: We retrieved existing data from the province wide registry of patients with kidney disease and performed manual chart reviews on patients with ADPKD who were prescribed an SGLT2i from January 1, 2014, to December 31, 2024. The primary outcome was acute kidney injury (AKI). Secondary outcomes included eGFR slope before and after SGLT2i initiation, magnitude of "eGFR dip" after starting SGLT2i as well as the incidence of genitourinary (GU) infections requiring hospital admission, emergency room visit and/or outpatient diagnosis and treatment. RESULTS: We included 17 patients on SGLT2i in our retrospective chart review with a median exposure of 20.89 months. While on an SGLT2i, one (6%) patient met criteria for AKI. Three patients (18%) had an eGFR dip of greater than 10% after starting an SGLT2i. Before SGLT2i initiation, the estimated eGFR slope was -0.2571 mL/min/1.73 m(2). After initiation, the slope was -0.1435 mL/min/1.73 m(2) (P = .48). Two patients (12%) had documentation of a urinary tract infection, neither of whom required hospitalization, or an emergency department visit. LIMITATIONS: The main limitation was the lack of a comparator group, thereby making it difficult to determine the true risk of AKI in our cohort of patients with ADPKD on SGLT2i. Other limitations include our retrospective study design and small sample size, which limits the generalizability of these results. The median exposure time of our cohort to SGLT2i was only 20.89 months and we had limited eGFR data beyond 2 years post-SGLT2i initiation. We did not have data on total kidney volume of these patients. CONCLUSIONS: In this cohort of 17 patients with ADPKD on SGLT2i, we did not observe any signs of adverse kidney outcomes and only two instances of GU infections occurred, neither requiring emergency visits or hospitalization. More high-quality evidence is needed to determine the safety and efficacy of SGLT2i in this population.