Postoperative Euglycemic Diabetic Ketoacidosis and Encephalopathy Related to SGLT-2 Inhibitors: A Case Report and Discussion of Diabetes Treatment and “Sweet Pee Encephalopathy” in Perioperative Hospital Management

Abstract

During preanesthesia evaluation, patient medications are reviewed and many are not administered on the day of surgery. Additionally, neurosurgical patients can develop postoperative encephalopathy from a variety of etiologies, including metabolic derangements. We report a case of postoperative neurosurgical euglycemic ketoacidosis which presented as unexplained encephalopathy and was the result of continued action of the patient’s serum glucose cotransporter-2 (SGLT-2) inhibitor combined with perioperative fasting. A 68-year-old woman with a history of type 2 diabetes mellitus was admitted to the neurocritical care service after resection of a left temporal meningioma. On postop day 1, she became lethargic and with worsening aphasia. Laboratory studies revealed blood glucose 140 to 160 mmol/L, bicarbonate 9 mmol/L, anion gap of 21, and pH of 7.2. Urine was positive for ketones and glucose, and serum was positive for β-hydroxybutyrate. Endocrinology was consulted and the patient was diagnosed with euglycemic diabetic ketoacidosis and treated with insulin until her anion gap closed. Over the next 2 days, her neurological examination improved to baseline. Although the patient did not take empagliflozin the day of surgery, the drug has a half-life of >12 hours, and other reports have described continued glycosuria for up to 10 days after drug discontinuation. This case illustrates the need for increased awareness of SGLT-2 inhibitors and “sweet pee encephalopathy” among neurosurgical and neurointensivist teams as well as potential modification of perioperative management of patients using newly emerging SGLT-2 inhibiting pharmaceuticals.

Introduction

Selective serum glucose cotransporter-2 (SGLT-2) inhibitors are a class of diabetes drugs that work by inhibiting the reuptake of glucose in the kidney, thereby lowering blood glucose levels through the induction of glycosuria.¹ The first of these drugs was canagliflozin, which received FDA approval in 2013.^1,2 This was quickly followed by dapagliflozin and empagliflozin, both approved in 2014.^3,4 Several randomized, placebo- or comparator-controlled studies have demonstrated the safety and efficacy of these drugs in glycemic control and other positive outcomes including decreased bodyweight, blood pressure, triglycerides, and cardiovascular events.^3–5 Due to their recent development and numerous trials demonstrating their efficacy, it is likely that the use of these drugs will increase as information disseminates, insurance coverage increases, and generic formulations become available.

Studies have found SGLT-2 inhibitors to be safe drugs for long-term use with generally mild toxicity profiles and minimal effects on organ function.^6,7 However, one serious potential complication is diabetic ketoacidosis, specifically euglycemic diabetic ketoacidosis (diabetic ketoacidosis with a normal or relatively normal blood glucose level).^8,9 This rare but serious complication has been noted by the FDA, and it is now listed as a black box warning for this class of drug. In their correspondence regarding this issue, the FDA noted that they had received numerous reports of delays in diagnosis and treatment of diabetic ketoacidosis due to its presentation with blood glucose levels much lower than what would be expected in a patient with diabetic ketoacidosis.¹⁰ Diabetic ketoacidosis presents heterogeneously, with common symptoms including nausea, vomiting, abdominal pain, fatigue, and difficulty breathing. A less common but well-documented presentation is encephalopathy including confusion, lethargy, and altered personality.¹¹

In the setting of neurocritical care postneurosurgery, the presence of newly onset encephalopathy can be very concerning due to potential sequelae of the surgery, including brain hemorrhage, infection, or reactions to general anesthetics. This is a report of a case in which postneurosurgery encephalopathy was observed secondary to diabetic ketoacidosis which was caused by continued action of the SGLT-2 inhibitor empagliflozin despite discontinuation on the day of surgery. As noted by the FDA, SGLT-2-mediated diabetic ketoacidosis can be difficult to diagnose due to its presentation with a relatively normal blood glucose level. In this case, the diagnosis was further obscured by its occurrence in a patient 1 day postcraniotomy. With the increasing use of SGLT-2 inhibitors in diabetes management, it is important for neurosurgeons, neurointensivists, and neurohospitalists to be aware of these drugs and their potential to cause diabetic ketoacidosis and encephalopathy, particularly in patients with decreased oral intake.¹² It should also be noted that this patient cleared preanesthesia evaluation with the specific instructions to continue empagliflozin up to the morning of surgery. Given the seriousness of this potential complication, further thought and discussion between surgery and anesthesiology teams regarding proper perioperative management of these drugs is warranted.

Case Description

This 68-year-old female with a past medical history notable for type 2 diabetes presented to our institution after diagnosis with a left temporal meningioma. After consultation with neurosurgery, she reported to her preanesthesia evaluation in preparation for craniotomy and removal of her tumor. The preoperative evaluation noted that the patient had been taking empagliflozin 10 mg once daily. She was instructed to hold oral antihyperglycemic and noninjectable insulin agents beginning “on the day of surgery.”

The patient’s operative note stated that the tumor was readily removed from the parenchyma of the brain. She awoke from anesthesia with no immediate complications. She was admitted to the neurocritical care service for postoperative observation and recovery.

On postoperative day 1, the patient was found to be increasingly lethargic with confusion and worsening expressive aphasia. Electroencephalogram was performed and showed diffuse, moderate slowing involving the left frontotemporal region, thought to be related to resection of the tumor. No clinical or subclinical seizure activity was detected. Computed tomography scan revealed no signs of hemorrhage.

Laboratory studies revealed blood sugar of 140 to 160 mmol/L, bicarbonate of 9 mmol/L, anion gap of 21, and blood pH of 7.2. Further investigation of blood acidosis revealed elevated serum β-hydroxybutyrate, positive urine ketones, and urine glucose >1000 mg/dL. Other urine laboratory test results were within normal limits. Blood urea nitrogen (BUN) and blood lactic acid were within normal limits, ruling out uremic and lactic acidoses. There was no exposure to methanol, ethanol, paracetamol, propylene glycol, ethylene glycol, salicylates, or other drugs known to cause metabolic acidosis, and liver function tests were all within normal limits. Osmolar gap was calculated at 9, further ruling out alcohol poisoning. Given these laboratory values and the patient’s lack of other exposures, and after consultation with endocrinology, the patient was diagnosed with euglycemic diabetic ketoacidosis. She was treated with an insulin drip per institutional diabetic ketoacidosis protocol until her anion gap closed. Over the following 2 days, her neurological examination returned to her baseline. After hospitalization, this patient gave her informed consent for the publication of this manuscript.

Discussion

Although diabetic ketoacidosis (DKA) is an established toxicity in patients taking SGLT-2 inhibitors, particularly in the setting of decreased oral intake, this case was challenging because (1) the onset of encephalopathy on a neurosurgery/neurocritical care service immediately raises concerns for neurologic and surgical etiologies, and (2) neurologists and neurosurgeons do not routinely manage the care of diabetes or prescribe these pharmaceuticals. Furthermore, the patient’s encephalopathy may have masked her ability to report DKA symptoms such as pain and nausea. For these reasons as well as the rising incidence of diabetes in the United States, increased awareness of SGLT-2 inhibitors is needed among these providers as well as patients taking these drugs who may undergo surgery.

The propensity for SGLT-2 inhibitors to cause diabetic ketoacidosis even after cessation on the day of surgery warrants further discussion regarding perioperative management of patients using these drugs. The half-life of empagliflozin is approximately 12.4 hours.¹³ Using the commonly accepted formula that drugs are removed from the bloodstream after approximately 5 half-lives, one would expect empagliflozin to remain in the blood for approximately 62 hours, or just over 2.5 days. Moreover, measurements of the drug in the bloodstream do not preclude its continued binding to SGLT-2 receptors and therefore continued action beyond 62 hours. Reports have documented cases in which patients continued to experience glycosuria, sometimes leading to ketoacidosis, for up to 10 days after cessation of an SGLT-2 inhibitor.^14,15 It is known that genetic mutations in uridine diphosphate glycosyltransferase enzymes that metabolize drugs, as well as mutations in SGLT-2 itself, can modulate duration of glycosuria.^16,17 These genetic differences in drug metabolism may explain prolonged glycosuric effects and acidosis in a subset of this patient population. Of note, our patient continued taking her SGLT-2 inhibitor until the day before her surgery (exact timing unknown).

In 2016, the American Association of Clinical Endocrinologists and the American College of Endocrinology (AACE/ACE) released a joint position statement regarding SGLT-2 inhibitors. In the paper, they noted the risk of DKA and its increased likelihood in patients undergoing surgery; however, they recommended stopping the drugs 24 hours before surgery.¹⁸ Conversely, the Australia/New Zealand College and Anesthetists recommend stopping SGLT-2 inhibitors 3 days before planned surgery,¹⁹ and a case series published by Canadian cardiac surgeons recommends cessation at least 2 days and in some cases 1 week before surgery.²⁰ These groups are unified in their acknowledgment of the risk of DKA when these patients cease oral intake and undergo surgery, and each of them note the need for increased perioperative monitoring of these patients. The AACE/ACE noted that patients who develop SGLT-2-mediated diabetic ketoacidosis tend to be insulin deficient at the time of this diagnosis. If this is the case, it may be possible to mitigate the risk of DKA through perioperative monitoring and increased insulin doses. Surgical trauma is a well-documented cause of disruption in insulin regulation which can present as hyperglycemia and hypoinsulinemia/marked insulin resistance.^21,22 It appears that this was the status of our patient, but excessive glycosuria from the SGLT-2 inhibitor masked its appearance on laboratory glucose assessments.

In conclusion, we present a case of euglycemic ketoacidosis in a type 2 diabetic patient, which was caused by continued action of SGLT-2 inhibiting diabetes medication in combination with perioperative fasting and surgery. The possibility of DKA in patients taking SGLT-2 inhibitors is listed as a black box warning on drug labels, with low nutrition states noted as a notable concomitant risk. Due to the potential presentation of this problem as encephalopathy, it is important for anesthesiologists, neurosurgeons, neurointensivists, and neurohospitalists to be aware of these “-flozin” drugs and their prolonged mechanism of “sweet urine flow.” Given this duration of action, perioperative management of these drugs should be further considered and patients should either cease these drugs earlier than day of surgery or be more strictly monitored and treated with increased insulin doses during the perioperative period. Early cessation may present a challenge at institutions in which preoperative visits tend to occur the day before surgery, but it is important and may become increasingly important for surgeons to be aware of this potential adverse event postsurgery, particularly neurosurgery.