Recent reports have highlighted the relatively rare yet potentially serious risk of diabetic ketoacidosis (DKA) with SGLT2 inhibitor use. In May 2015, the FDA issued a safety announcement after discovering 20 cases of DKA reported to the FDA Adverse Events Reporting System between March 2013 and June 6, 2014. These cases were atypical because most occurred in patients with type 2 diabetes (T2DM) who had only slightly increased blood glucose levels (termed euglycemic DKA [euDKA]).1
Now, two recently published articles have added more background to the issue.
In the first published case series2 about the link between SGLT2 inhibitors and euDKA, experts describe 13 cases of euDKA among patients with T1DM and T2DM. All patients had near normal blood glucose levels that delayed diagnosis and treatment. Importantly, because most patients did not know that they had DKA, they kept their insulin doses the same or even decreased them, likely worsening the acidosis.
The case series was published online on June 15, 2015. EuDKA cases linked to SGLT2 inhibitor use were identified incidentally without a systematic review of databases or medical records.
Key findings included:
• Most cases were women.
• Nine cases occurred in patients with type 1 diabetes (T1DM).
– Most of these cases were linked to decreased insulin dose.
– Other contributing factors were recent illness, increased exercise, decreased food intake, and alcohol consumption.
– A few patients had no identifiable contributing factors.
• 2 cases occurred in patients with T2DM, both of which developed postoperatively.
• After re-starting SGLT2 inhibitors, euDKA recurred in 3 patients.
• All patients responded well to intravenous fluids and insulin.
Patients with T1DM or T2DM who are taking SGLT2 inhibitors and who present with DKA symptoms should have urine and/or serum ketones checked, the authors advised. They also advise “great caution, extensive counselling, and close monitoring” if clinicians choose to prescribe SGLT2 inhibitors off-label to patients with T1DM.
The second article reviews possible mechanisms explaining why SGLT2 inhibitors could predispose to euDKA.3 Authors Simeon Taylor, MD, PhD (University of Maryland School of Medicine), Jenny Blau, MD, (NIH), and Kristina Rother, MD, MHSc (NIH) searched PubMed and Google for publications about the molecular mechanisms behind SGLT2 inhibitors and ketone body metabolism.
They identified three main mechanisms whereby increased production and decreased clearance of ketone bodies may be additive and lead to increased serum ketone levels:
• Decreased insulin dose when combining insulin with SGLT2 inhibitors, in order to avoid hypoglycemia
– Insulin normally suppresses lipolysis and ketogenesis.
– The lowered insulin dose may not be enough to suppress these processes, leading to increased ketone body production.
• Decreased renal clearance of ketone bodies
– The parent drug, phlorizin, nonselectively inhibits SGLT1/SGLT2 and decreases renal clearance of ketone bodies.
– SGLT2 inhibitors could act in a similar way, leading to increased ketone body levels in the blood.
• Pancreatic alpha cells express SGLT2 receptors and SGLT2 inhibitors increase glucagon secretion
– This may lead to increased hepatic ketogenesis and decreased endogenous insulin in some patients.
The risk of euDKA remains quite low, however. Dr. Taylor estimates that euDKA among patients with T2DM may number less than 1:1000, and is probably closer to 1:10,000. Most of the concern centers on off-label use of SGLT2 inhibitors in patients with T1DM.
Doctors may be motivated to use SGLT2 inhibitors off label in patients with T1DM in order to decrease the insulin dose and possibly decrease the risk of hypoglycemia, Dr. Taylor explained. In addition, early research has suggested that empaglifozin can decrease glomerular hyperfiltration, which could be renoprotective in patients with T1DM. Dr. Taylor needs more convincing, though.
"I really do believe that off-label use of these drugs [SGLT2 inhibitors] in T1DM is inappropriate," commented Dr. Taylor, "As far as I can tell, there is no compelling evidence suggesting meaningful clinical benefit in terms of reduced risk of hypoglycemia or renoprotection. And there certainly are risks in terms of ketoacidosis for T1DM patients."
"If and when clinical trials provide convincing evidence of a favorable benefit-to-risk profile, then I will change my mind," he added, "But I don't think that clinical practice doctors should jump the gun and start using these drugs off-label at the present time."