Antiglycemic agent tied to 55 cases of life-threatening gangrene in diabetics

The FDA has tied sodium-glucose cotransporter-2 (SGLT2) inhibitors to 55 cases of a life-threatening form of gangrene in the past six years, according to a May 7 report in the Annals of Internal Medicine.

An FDA team spearheaded by Susan J. Bersoff-Matcha, MD, used case data from the agency’s adverse event reporting system and published reports to study how SGLT2 inhibitors fared against other conventional antiglycemic agents in patients with diabetes. The researchers focused on a window from March 2013 through January 2019—from the time the FDA first approved the SGLT2 inhibitor canagliflozin through present day.

Canagliflozin was initially cleared to reduce the risk of major adverse cardiovascular events in adults with type 2 diabetes and established CVD, and in 2016 the FDA approved empagliflozin, another SGLT2 inhibitor, for the same indication. Bersoff-Matcha et al. said researchers linked the drugs to genital mycotic and urinary tract infections in clinical trials, later adding urosepsis, pyelonephritis, ketoacidosis and acute kidney injury to the running list of side effects.

Eventually, the FDA issued a warning that canagliflozin was associated with an increased risk for lower-limb amputations, and in August 2018 the agency released a statement that the drug class was associated with Fournier’s gangrene (FG), a “rare but serious” urologic condition also known as necrotizing fasciitis of the perineum.

FG is characterized by necrotizing infection of the external genitalia, perineum or perianal region, Bersoff-Matcha and colleagues wrote. Although in its infancy it was considered an idiopathic disease that affected only men, more recent research has identified FG in both sexes and has traced the condition back to cutaneous, anorectal and urogenital infections.

After scrutinizing case files, Bersoff-Matcha’s team identified a total of 55 unique cases of FG in diabetic patients taking SGLT2 inhibitors. The majority were men but 16 were women, with patients’ ages ranging from 33 to 87 years.

Those individuals first presented with FG symptoms anywhere from 5 days to 49 months after initiating SGLT2 inhibitor therapy, and the prognosis wasn’t good. All patients reported surgical debridement and were “severely ill,” and eight suffered diabetic ketoacidosis, nine experienced sepsis or septic shock and four had acute kidney injury. An additional eight patients had to have fecal diversion surgery, two developed necrotizing fasciitis that required lower-extremity amputation and one required a lower-extremity bypass because of gangrenous toes. Three patients died.

The 55 FG cases linked to SGLT2 inhibitors stand in stark contrast to the 19 cases Bersoff-Matcha et al. found to be associated with other antiglycemic agents between 1984 and 2019. In those cases the culprits were metformin (eight patients), insulin glargine (six patients), short-acting insulin (two patients), sitagliptin plus metformin (two patients) and dulaglutide (one patient).

“We found only 19 cases in 35 years among patients receiving other classes of antiglycemic agents,” the authors wrote. “If FG were associated only with diabetes mellitus and not SGLT2 inhibitors, we would expect far more cases reported with other antiglycemic agents, considering the 35-year timeframe and the large number of agents.”

The researchers said the instance of FG in SGLT2 inhibitor patients was actually likely higher than they reported, too. They said they’d found 41 additional reports that mentioned FG in the case narrative, but those reports lacked a documented surgical procedure and couldn’t be used for analysis. The authors also said it’s possible healthcare providers reported FG as the result of another disease, like diabetes, rather than the result of a drug being used to treat it.

“Of importance, in the absence of a control group, causality cannot be established and such issues as confounding by indication cannot be excluded,” they wrote. “Despite these weaknesses, in the postmarket setting, spontaneous reporting may be the only way to learn about very rare drug-related adverse events.”