Euglycemic DKA Associated with Dapagliflozin Requiring Prolonged Insulin Infusion

Presentation Number: SAT 603
Date of Presentation: April 1st, 2017

Raquel Villavicencio* and Amale A Lteif
Richard L Roudebush Veterans Administration Medical Center, Indianapolis, IN


Background: In May 2015 the FDA issued a warning that sodium glucose cotransporter 2 inhibitors (SGLT2i) may lead to DKA, however differences in treatment of these cases is not well defined. We present the case of a man requiring prolonged insulin infusion with persistent glucosuria 

Clinical case: A 41-year-old well-appearing man with history of Addison’s disease, hypothyroidism and five-year history of diabetes mellitus type 2 on dapagliflozin 10 mg daily for the past 18 months along with glipizide, metformin, sitagliptin, and pioglitazone, presented for initial visit to endocrinology clinic where he was noted to have a fruity odor to his breath. Labs showed an anion gap (AG) of 20 mmol/L (5-15), blood glucose of 110 mg/dL, serum bicarb of 20 mmol/L (21-32), β-hydroxybutyrate (β-OH) level of 3.53 mmol/L (0-0.3), urine glucose of >500 mg/dL and urine ketones of 80 mg/dL. Lactate and renal function were normal. The patient was admitted for euglycemic DKA (eDKA) and started on an insulin infusion with dextrose 10%. AG closed and he was transitioned to subcutaneous insulin, but AG quickly reopened (16mmol/L) with recurrence of ketonemia (β-OH 2.9 mmol/L) requiring resumption of insulin drip on hospital day three. Hospital day five, transition to subcutaneous insulin was again attempted. AG remained closed and urinalysis was now negative for ketones, but continued to show heavy glucosuria. C-peptide was low at 0.61 ng/mL (0.80-3.85) and GAD-65 was positive indicating the patient had previously unrecognized latent autoimmune diabetes of adulthood (LADA).  

Conclusions: Glucosuria persisted longer than would have been expected given the half-life of dapagliflozin is 12.9 hours, suggesting a prolonged mechanism of action such as possible receptor binding irreversibility. SGLT2i are thought to increase ketonemia due to decrease in circulating insulin level but possibly also due to decreased renal clearance of ketones and by increasing glucagon secretion stimulating ketogenesis (1).  

This case highlights the need to screen patients who might be at high risk for eDKA prior to initiation of SGLT2i. In this case, history of other autoimmune disorders and poor response to multiple hypoglycemic agents were clues this patient had LADA. In addition, the prolonged biological effect of the SGLT2i beyond what is predicted by the pharmacological half-life needs further investigation to better guide clinical management. 


Nothing to Disclose: RV, AAL