Case Presentation: Euglycemic diabetic ketoacidosis (euDKA) is increasingly recognized in the perioperative setting in patients with diabetes who are treated with sodium-glucose cotransporter-2 inhibitors (SGLT-2i). Here, we describe a rare case of a patient with prediabetes not on pharmacotherapy who developed euglycemic ketoacidosis postoperatively.A 71-year-old man with history of prediabetes (A1c 5.8%), hypertension, and childhood poliomyelitis with paraplegia was admitted by orthopedic surgery for right shoulder arthroplasty. Hospital medicine was consulted on postoperative day 2 for persistent sinus tachycardia to 120s and frequent premature ventricular complexes (PVCs) despite fluid resuscitation. Physical exam revealed a thin man with lower extremity paraplegia and sustained tachycardia with frequent skipped beats. Labs were notable for leukocytosis (white blood cells 19.4 x 10^9/L), high anion gap metabolic acidosis with respiratory compensation, a normal BUN and serum lactate. Computed tomography of the chest and abdomen was negative for pulmonary emboli, consolidates, or intra-abdominal infection. Blood cultures were negative. A urinalysis showed no evidence of infection but significant glucosuria (>500 mg/dL) and ketonuria (>80 mg/dL). Serum beta-hydroxybutyrate was elevated (3.95 mmol/L) and point of care glucose levels ranged from 112-176 mg/dL. In consultation with Endocrinology and Nephrology, the patient was presumptively treated for euglycemic ketoacidosis with 1L isotonic crystalloid with bicarbonate, continuous fluid with dextrose, and an insulin sliding scale. This resulted in closure of the anion gap, improved acidosis, and resolution of ketonuria. By postoperative day 6, the WBC normalized and glucosuria, tachycardia, and PVCs resolved.
Discussion: Clinicians are increasingly sensitive to postoperative euDKA in diabetic patients with SGLT2i exposure, however this case highlights that euglycemic ketoacidosis can also present in other patient populations such as patients with prediabetes. Although this patient consumed all of his meals postoperatively and before onset of his metabolic disturbances, he would later be diagnosed with protein calorie malnutrition. Prior case reports and small cohort studies have also described euglycemic ketoacidosis in patients following bariatric and cardiac surgery, acute pancreatis, and pregnancy with malnutrition – though mostly in patients with pre-existing diabetes. Our patient’s relative insulin deficiency in the setting of insulin resistance (prediabetes), coupled with excess of counterregulatory hormones in the setting of malnutrition may have led to ketoacidosis following surgery. He also had transient glucosuria – with physiology briefly mimicking SGLT2i use – promoting euglycemia despite ongoing ketogenesis, gluconeogenesis, and glycogenolysis.
Conclusions: Perioperative medicine consultants and co-management hospitalists should have a high index of suspicion in diagnosing euglycemic ketoacidosis in patients with postoperative anion gap acidosis, glucosuria, and ketosis. Although patients at highest risk for this condition may be identified by risk factors such pre-existing diabetes and SGTL2i use, euglycemic ketoacidosis may also occur in other at-risk populations. Our case also demonstrates the value of a hospital medicine-based co-management service to navigate peri-operative diagnostic uncertainties, including identification of post-operative metabolic disturbances.