SGLT2 Inhibitors: Now The Bad News

By Jeff Unger, MD, FACE

Director, Unger Primary Care Concierge Medical Group, Rancho Cucamonga, CA

Dr. Unger reports he is a consultant for and serves on the speakers bureau of Janssen Pharmaceuticals.

SYNOPSIS: Sodium-glucose cotransporter-2 inhibitors are approved as adjunctive therapy for the treatment of patients with type 2 diabetes. This drug class has been also used off-label to improve glycemic control in patients with type 1 diabetes. This article describes seven cases of diabetic ketoacidosis in patients with type 1 diabetes and two cases in patients with type 2 diabetes. Although the patients were ketotic their blood glucose levels were not significantly elevated. Thus, patients are referred to as having “euDKA.”

SOURCE: Peters AL, et al. Euglycemic diabetic ketoacidosis: A potential complication of treatment with sodium-glucose cotransporter 2inhibition. Diabetes Care 2015;38:1687-1693.

Diabetic ketoacidosis (DKA) occurs in approximately 5% of patients with type 1 diabetes and is defined by a triad of hyperglycemia (blood glucose > 250 mg/dL), anion-gap acidosis, and increased plasma ketones. Euglycemic DKA (euDKA) is considered rare, but may be underreported. Risk factors for euDKA include meal skipping, alcohol intake, and inhibition of gluconeogenesis. Patient 1 was a 40-year-old female with type 1 diabetes and a body mass index of 26.5 kg/m². Prior to canagliflozin initiation, her A1c was 11.4%. As her glycemic control improved within 2 weeks, the patient reduced her basal insulin dose by 50%. Two weeks later, she developed a febrile illness resulting in reduced oral intake. She was therefore consuming fewer carbohydrates and using less insulin. In the emergency department (ED) her blood glucose was 220 mg/dL and serum ketones were positive. Her blood gas showed a pH of 6.9.

A second patient with type 1 diabetes was prescribed canagliflozin as adjunctive therapy to her insulin pump regimen. Her baseline A1c was 7%. The patient walked through an amusement park for 12 hours before becoming ill. After reducing her basal insulin dose, the patient awoke the next morning with a migraine headache. Over the course of 5 days, her headache persisted. A neurologist prescribed steroids. Although her blood glucose was 120 mg/dL, the patient’s urine ketone test was positive. She self-managed the euglycemia with oral-glucose-containing fluids, insulin, and antiemetics. As her ketones resolved, so did her migraine, nausea, and vomiting.

COMMENTARY

Personally, I have discontinued my off-label use of sodium-glucose cotransporter-2 (SGLT2) inhibitors in patients with type 1 diabetes as I have experienced seven cases of euDKA. Five of my patients required overnight hospitalization and responded to hydration, insulin, and carbohydrate intake. Interestingly, the ED doctors were very confused by their presentation in each case. One 25-year-old patient on an insulin pump and canagliflozin helped a friend move. Knowing that his increased activity would likely result in hypoglycemia, the patient lowered his basal rate by 50% for several hours. Seven hours after the move was completed, the patient developed severe muscle pain, nausea, and vomiting. His blood glucose level in the ED was 184 mg/dL, but he was noted to have positive serum ketones. The ED staff was unfamiliar with the diagnosis of euDKA.

The pathophysiology of euDKA is thought to result from the following sequence of events: 1) As carbohydrate utilization is reduced (i.e., caloric restriction), lipolysis becomes the primary source of energy production for the body. 2) The patients with euDKA typically reduce their exogenous insulin dose, thereby enhancing lipolysis. 3) SGLT2 inhibitors are known to raise serum glucagon levels as a means of glucose counterregulation. 4) Exogenous insulin increases glycosuria, which may also increase glucagon production and secretion. 5) The hyperglucagonemia coupled with increased delivery of free fatty acids (product of lipolysis) to the liver promotes both the oxidation of free fatty acids and the production of ketone bodies. 6) SGLT2 inhibitors, as a class, tend to prevent renal excretion of ketone bodies.

EuDKA may be predictable, detectable, and preventable in patients using SGLT2 inhibitors. Patients using SGLT2 inhibitors (especially those using this class of drug as adjunctive therapy to insulin) should be informed to: 1)minimize their risk of dehydration; 2) avoid reducing insulin doses on “sick days; 3) minimize alcohol intake, which may increase hepatic ketone production; and 4) consume carbohydrates on sick days, which will reduce their dependency on lipolysis.

Finally, patients with type 2 diabetes who develop DKA while on SGLT2 inhibitors should be tested for latent autoimmune diabetes of adulthood. DKA is extremely rare in type 2 diabetes.