MON-540 Prolonged Glucosuria and Acidosis Following SGLT2-Inhibitors Discontinuation: A Retrospective, Single-Center Study

Abstract Disclosure: P. Pile: None. F.S. Celi: None. Introduction: Sodium-glucose cotransporter-2 inhibitors (SGLT2-i) therapy is well-known for its beneficial effects on cardiovascular and renal disease beyond its glucose lowering properties. These effects have led to widening of its therapeutic indications to heart failure and chronic renal insufficiency, with a dramatic increase in the number of prescriptions. Euglycemic diabetic ketoacidosis (eDKA), particularly in the setting of hospitalization, represents one of the major adverse events associated with SGLT2-i therapy. The package inserts of the most prescribed SGLT2-i recommend stopping these medications three days before surgical procedures to minimize the risk of eDKA. However, some reports indicate prolonged glucosuria associated with eDKA following discontinuation of SGLT2-i therapy. Here we present a retrospective, single-center study directed to assess the prolonged effects of SGLT2-i following its discontinuation. Methods: After receiving exemption from the local Institutional Review Board, the electronic medical records of our institution were reviewed to identify patients with prior outpatient use of SGLT2-i who were admitted to the hospital (all causes) in the period November 2019 - November 2024 with at least two urinalysis (UA) and basic metabolic profile determinations; admission (baseline) and follow-up (repeat UA during hospitalization or following discharge) after at least three days following admission, while not receiving SGLT2-i. Results: Thirteen patients (7 females, 6 males) age 68±16 years were identified. Average follow up was 13±2 days (range 3-27 days). Baseline glucosuria was 815±305 mg/dL, while at follow up was 515±315 mg/dL. Baseline average glucose was 165±72 mg/dL, while at follow-up was 165±59 mg/dL. At baseline, severe glucosuria (>500 mg/dL) was observed in 12/13 patients, and on follow-up it was present in 9 patients. The remaining 4 patients had at least mild glucosuria (100-250 mg/dL). Glucosuria persisted as late as 27 days after discontinuation of SGLT2-i. At baseline, anion gap was 14.00±1.63 mmol/L (3-11 mmol/L), while at follow-up was 13.30±1.54 mmol/L. Conclusion: Our investigation revealed that the glucosuric effect of SGLT2-i is still present long after their discontinuation. Prolonged glucosuria in association with elevated anion gap raises the concern for acidosis which could be attributed to residual effects of SGLT2-i. This study provides insight regarding the recommendations of discontinuing SGLT2-i prior to surgical procedures or during hospitalization. In addition to the common practice of discontinuing this class of medications, providers should aggressively monitor patients for acidosis and have a low threshold to institute insulin therapy even in cases of euglycemic state or minimal hyperglycemia to prevent euglycemic ketoacidosis associated with SGLT2-i therapy. Presentation: Monday, July 14, 2025