PerspectiveIn the Journals

Invokana increases rates of ketone-related adverse events in type 1 diabetes

Show Citation

April 14, 2016

In patients with type 1 diabetes insufficiently controlled with insulin, the addition of Invokana improved glycemic control but increased the incidence of ketone-related adverse events, including serious adverse events of diabetic ketoacidosis, according to recent findings.

In a phase 2 randomized placebo-controlled, double blind trial, Anne L. Peters, MD, of the Keck School of Medicine of the University of Southern California in Los Angeles, and colleagues evaluated data from 351 patients with type 1 diabetes for at least 1 years with baseline HbA1c 7% to 9%. Participants were on a stable insulin regimen consisting of multiple daily insulin injections or continuous subcutaneous insulin infusion for at least 8 weeks prior to screening.

Participants were randomly assigned to Invokana (canagliflozin, Janssen) 100 mg (n = 117) or 300 mg (n = 117) or placebo once per day taken before the first meal of the day. Researchers sought to determine the effect of canagliflozin as an add-on to insulin on glycemic control and weight as well as on the incidence of serious adverse events, including diabetic ketoacidosis (DKA).

At week 18, the incidence of any ketone-related adverse event was 5.1% with canagliflozin 100 mg, 9.4% with canagliflozin 300 mg and none in the placebo group. Most ketone-related adverse events took place after 1 month of treatment with canagliflozin. Serious adverse events of DKA necessitating hospitalization occurred in 4.3% of the canagliflozin 100 mg group and 6% of the canagliflozin 300 mg group. Of the 12 participants who had serious adverse events of DKA, the range of blood glucose levels at the time of hospitalization was 9.4 mmol/L to more than 44.4 mmol/L; five of these participants had blood glucose concentrations less than 13.9 mmol/L, which is the standard glucose threshold used to define DKA.

Participants in the canagliflozin 100 mg group had a median time of 116 days to a serious adverse event of DKA compared with a median 32 days in the canagliflozin 300 mg group. Coexisting conditions at the time of the event were present in all participants who experienced a serious adverse event of DKA, including influenza, pneumonia, infusion-site infection, food poisoning, insulin pump failure/malfunction, noncompliance with insulin regimen or decrease in food intake.

There were no differences in baseline characteristics between participants who experienced a ketone-related adverse event vs. those without a ketone-related adverse event that would appear to predict additional risk.

“People with type 1 diabetes who use an SGLT-2 inhibitor are at increased risk for DKA, which appears to be dose related,” Peters told Endocrine Today. “If [canagliflozin is] used in this off-label fashion, patients should be fully educated as to this risk and willing to monitor ketones at times of illness or other stress, and only the lowest dose of the SGLT-2 inhibitor should be used." – by Jennifer Byrne

For more information:

Anne L. Peters, MD, can be reached at

Disclosure: Peters reports various financial ties with Abbott Diabetes Care, Amgen, AstraZeneca, Becton Dickinson, Biodel, Bristol-Myers Squibb, Boehringer Ingelheim, CVS/Caremark, Eli Lilly, Janssen, Lexicon, Medtronic, Merck, Novo Nordisk, OptumRx, Sanofi, Takeda and Thermalin Diabetes. Please see the full study for a list of all other authors’ relevant financial disclosures.

itj+ Infographic

itj+ Perspective

John B. Buse

John B. Buse

The use of SGLT2 inhibitors in type 1 diabetes is an exciting proposition because of the efficacy to reduce HbA1c, glycemic variability, blood pressure and weight. However, the safety has been totally unresolved until now. The study by Peters and colleagues demonstrates as a best case scenario, patients enrolled in a short-term clinical trial, 5% to 10% of patients treated with an SGLT2 inhibitor developed diabetic ketoacidosis (DKA), mostly between 1 and 3 months. As is true for DKA in general, there were precipitating factors, but no baseline characteristics that were specific or sensitive risk markers. Furthermore, many cases were not accompanied by hyperglycemia. For now, SGLT2 inhibitor therapy in the setting of type 1 diabetes has to be viewed as potentially dangerous and needs to be considered only in the context of extensive and balanced patient education as well as careful follow-up with ready availability of clinicians experienced in diagnosing and treating DKA in this unusual setting with atypical presentation. Efforts to mitigate the risk for DKA will need to be developed to have any hope of leveraging the benefit of the drug class in patients with type 1 diabetes.

John B. Buse, MD, PhD
Verne S. Caviness distinguished professor
Chief, Division of Endocrinology
Director, Diabetes Care Center
University of North Carolina School of Medicine
Chapel Hill, NC

Disclosure: Buse reports receiving grants, research support, consulting fees or speakers’ fees from Amylin Pharmaceuticals, Andromeda, AstraZeneca, Astellas, Boehringer Ingelheim, Bristol-Myers Squibb, Dance Biopharm, Elcelyx Therapeutics, Eli Lilly, GI Dynamics, Halozyme Therapeutics, F. Hoffmann-La Roche, Intarcia Therapeutics, Johnson & Johnson, Lexicon, LipoScience, MacroGenics, Medtronic, Merck, Metavention, Novo Nordisk, Orexigen Therapeutics, Osiris Therapeutics, PhaseBio Pharmaceuticals, Quest Diagnostics, Sanofi, Santarus, Scion NeuroStim, Takeda, ToleRx and TransTech Pharma.