Case Presentation: A 62 YO woman with PMHx significant for type 2 DM, HTN, CAD with a previous MI, and OSA was recovering as expected from elective decompressive laminectomy. On POD 4 she developed nausea, confusion, and lethargy. The hospitalist service was consulted, prompting a metabolic work-up that suggested anion gap metabolic acidosis. There were no remarkable findings on examination.
Laboratory results showed an anion gap of 32, a bicarbonate level of 5 mmol/L, an arterial pH 7.14, and WBC elevated to 14,000. A lactate level was normal at 1.1 mmol/L, ruling out lactic acidosis. There was concern for possible diabetic ketoacidosis, but serum glucose was only mildly elevated at 150 mg/dL and urinalysis was negative for ketones, although positive for glucose (>1000 mg/dL). A serum beta-hydroxybutyrate was measured at 3.2 mmol/L.

Last administration of anxiolytic or opioid analgesia was 18 hours before nursing reported her symptoms. She had been wearing her CPAP as directed. The patient was restarted on her oral diabetes medication (a SGLT2-Biguanide combination) on POD 2 by her primary team with no further laboratory data. Her symptoms were thought to be secondary to a suspected euglycemic diabetic ketoacidosis.

Discussion: Diabetic ketoacidosis is often a typical serious complication from an absolute insulin deficiency. Precipitants of DKA include infection, inflammation, poor compliance, glucocorticoids, ischemia or infarction. The body increases lipolysis and increases ketone formation in the liver. Clinical manifestations include nausea, vomiting, abdominal pain, polyuria, dehydration, altered mental status, and Kussmaul’s respirations. Laboratory studies typically reveal high anion gap acidosis, positive urine and serum ketones, and marked hyperglycemia.

Cases have been reported of euglycemic DKA (plasma glucose < 250 mg/dL) as an adverse event in the setting of insulin-independent sodium-glucose cotransporter 2 (SGLT2) inhibitor therapy. Special consideration should be given for post- operative patients using SGLT2 inhibitors. It is thought that these leads to falsely low blood glucose levels with subsequent reduction in insulin in the setting of illness. Although the mechanism is not fully understood, SGLT2 inhibitors are associated with an increase in glucagon levels, further increasing the propensity towards ketone production. Treatment algorithms for euglycemic DKA follow the same protocols as traditional DKA with the correction of electrolyte abnormalities and fluid balance. Appropriate treatment will require frequent laboratory monitoring and clinical reassessment as advised by the American Diabetes Association. She was transferred to ICU for closer monitoring and required insulin and dextrose infusion to correct her acidosis.

Conclusions: The patient had been on Canagliflozin-Metformin 150-1000mg BID for diabetes management prior to admission. During her pre-operative assessment plans for her inpatient diabetes management were not addressed. Oral diabetes medication were appropriately held the day prior to her planned surgery.

The patient’s euglycemic diabetic ketoacidosis was felt to be secondary to the SGLT2 inhibitor, Canagliflozin, which had been restarted on POD 2 during a post-operative stress period. Development of her acidosis had not been monitored, as a metabolic panel was last obtained on POD 2. It is recommended that oral diabetic medications should be held in the hospital setting in which a change in medical condition and nutritional intake is anticipated.