Patient-Specific Risk Factors and Clinical Correlates of Euglycemic Diabetic Ketoacidosis in Patients on Sodium-Glucose Co-Transporter-2 Inhibitors

Abstract

Abstract Background: Sodium-Glucose Co-Transporter-2 Inhibitors (SGLT-2i) association with euglycemic diabetic ketoacidosis (EDKA) has been well reported. The underlying mechanism is mainly enhanced lipolysis and ketone bodies’ reabsorption. They also stimulate the pancreatic alpha cells and inhibit the beta cells, thereby causing an imbalance in glucagon/insulin levels, further contributing to lipolysis and ketogenesis. SGLT-2i were also found to cause EDKA in all types of diabetes, even uncovering undiagnosed Latent Autoimmune Diabetes of the Adult (LADA). Methods: Numerous electronic databases were systematically searched to identify patient-specific risk factors and clinical characteristics of EDKA in patients on SGLT-2i. The patient’s symptoms, clinical profile, laboratory results, and precipitants for EDKA were reviewed. Results: A total of 96 case reports identifying 116 patients with EDKA was fully reviewed. EDKA was twice prevalent in females (66.3%) than males (33.6%); median age was 52.15 ± 13.47, BMI was 29.3 ± 7.0. Among the 116 DKA events in SGLT-2i 92 (79.3%) were associated with Type-2 DM, 15 (12.9%) were Type-1 DM, 8 (6.9%) in LADA. Common symptoms were nausea (48.7%), vomiting (47%), and abdominal pain (28.2%). Canagliflozin was the most common SGLT-2i (40.5%), followed by Empagliflozin (29.3%) and Dapagliflozin (25.9%). The most common precipitant was surgery (17.2%), followed by infection (14.7%), fasting (11.2%), and Keto Diet (9.5%); others being reduced insulin use, alcoholism, and cancer. At presentation, average blood glucose was 196.8 ± 96.5, pH 7.1 ± 0.16, HCO3 8.7 ± 5.7 mmol/L, potassium 4.3 ± 1.03, anion-gap 24.2 ± 6.8 mmol/L, and the average HbA1C was 9.24 ± 2.08. Urine Ketones were positive in 81.89% of patients. 17 patients had pancreatic autoantibodies testing, and 7 were positive (41.2%) for glutamic acid decarboxylase-65 antibodies (anti-GAD-65). As a result, 7 patients were newly diagnosed with LADA who were previously misdiagnosed with type-2 DM. Conclusion: SGLT2i induced EDKA was found to be more predominant in females and type-2 DM. Diabetics should be educated on risk factors and consult physicians before commencing a dietary or exercise change. Physicians should be vigilant in diagnosing EDKA by thoughtful measurement of urine ketones and anti-GAD-65 testing can help diagnose underlying LADA.