A case of perioperative euglycemic ketoacidosis in a patient without diabetes: are current guidelines enough?

Abstract

Background: Euglycemic ketoacidosis (eKA) is a serious and potential life-threatening complication of SGLT2-inhibitor (SGLT2i) use. eKA associated with SGLT2i has been increasingly reported in patients without diabetes likely due to more widespread use. The perioperative setting is a particularly vulnerable time for patients due to decreased carbohydrate intake, volume depletion, medication changes, and increased counterregulatory hormone activity due to surgical stress. Current guidelines recommend that patients with type 2 diabetes hold SGLT2i for at least 24-72 h prior to elective surgery.

Case presentation: We report a case of an 82-year-old woman without a history of diabetes who held empagliflozin for 72 h prior to planned bowel resection for management of colon cancer. The indication for empagliflozin was heart failure with reduced ejection fraction. Intraoperatively, she was found to have profound metabolic acidosis, high-normal anion gap, normal glucose, and elevated serum beta hydroxybutyrate. Given the high risk for decompensation intraoperatively, the patient was empirically given dextrose and insulin to treat eKA. The surgery was otherwise uncomplicated. Afterward, the patient was transferred to the intensive care unit for treatment of eKA with insulin and dextrose infusions. She recovered and was discharged home. Empagliflozin was not restarted.

Conclusions: Our patient's case demonstrates that there is not a one-size-fits-all approach to withholding SGLT2i in patients in the perioperative setting. Despite holding the SGLT2i preoperatively, our patient without a history of diabetes nonetheless developed eKA. Upon further review, this patient had risk factors for developing eKA, including age, sex, chronic kidney disease, and preoperative nutrition status. Additional preoperative workup may have been warranted due to suspected preoperative hypovolemia associated with bowel preparation, perhaps allowing us to identify eKA prior to surgery. Future considerations include obtaining a basic metabolic panel on day of surgery for patients taking SGLT2i to evaluate for acute renal dysfunction and metabolic acidosis prior to surgical intervention. Point-of-care serum beta-hydroxybutyrate could also be incorporated into clinical decision-making though this may not be widely available. Future investigation should examine risk factors that predispose patients to developing eKA and drive development of protocols to guide which patients require more monitoring perioperatively.