The risk of developing diabetic ketoacidosis (DKA) among type 2 diabetes patients initiating a sodium–glucose cotransporter 2 (SGLT2) inhibitor medication is about double that seen among patients starting a dipeptidyl peptidase-4 (DPP-4) inhibitor, but the overall risk is still low, new research suggests.
Findings from the largest study conducted to date to investigate the issue were published as a research letter in the June 8 issue of the New England Journal of Medicine by Michael Fralick, MD, and colleagues at the Brigham and Women's Hospital, Boston, Massachusetts.
"We found a doubling in the risk of DKA, which sounds frightening, but the absolute risk is quite small....I still think this is a very good class of medications and for certain patients will continue to be. Now we just have a little more information to add to the discussion when the risks and benefits are being considered," Dr Fralick told Medscape Medical News.
He estimates that between 5 and 8 patients per 1000 initiating SGLT2 inhibitors will develop DKA.
And he advises that patients be monitored for signs of DKA or full information to thew patients of the symtoms of DKA to seek help if DKA appears after starting on SGLT2 inhibitors, noting, "This is something that can happen relatively quickly, so that's why I think it's important right after patients are started on these drugs that they're closely monitored and the clinician considers ordering bloodwork."
But overall, Dr Fralick, a general internist, supports use of the SGLT2 inhibitor class for selected patients with type 2 diabetes, given the recent results from the Empagliflozin Cardiovascular Outcome Event Trial in Type 2 Diabetes Mellitus Patients (EMPA-REG OUTCOME) study showing reduction in cardiovascular deaths, as well as renal protection, with empagliflozin (Jardiance, Boehringer Ingelheim/Lilly).
"I completely agree that these medications have significant benefits," he commented.
All eyes will be on the results from another cardiovascular-outcomes trial with a different SGLT2 inhibitor, canagliflozin (Invokana, Johnson & Johnson) to be reported on Monday at the American Diabetes Association (ADA) 2017 Scientific Sessions.
Results from the Canagliflozin Cardiovascular Assessment Study(CANVAS) program will reveal whether the cardiovascular protection observed with empagliflozin in EMPA-REG OUTCOME is a class effect or not.
And the findings will further inform on some of the side effects so far associated with this drug class, including DKA, as well as fracture risk and a doubling of amputations of the lower limb, already identified with canagliflozin compared with placebo in CANVAS, which resulted in the Food and Drug Administration adding a boxed warning to this effect to the product label.
Largest Study of Its Kind
For the current study, Dr Fralick and colleagues used a claims database of commercially insured US patients (Truven MarketScan) and identified 50,220 type 2 diabetes patients who had received a new prescription for an SGLT2 inhibitor and 90,132 initiating a DPP-4 inhibitor (chosen as the comparator class because it is used similarly to SGLT2 inhibitors, as second-line after metformin for type 2 diabetes, but has no known link to DKA).
The primary outcome was hospitalization for DKA — the unadjusted rate within 180 days of SGLT2 inhibitor initiation was 4.9 per 1000 person-years, compared with 2.3/1000 person-years following DPP-4–inhibitor initiation (hazard ratio, 2.1).
After propensity score matching with 38,045 patients in each arm to account for confounders such as age, comorbidities, use of other medications, and healthcare utilization, the hazard ratio for hospitalization for DKA with SGLT2 inhibitors vs DPP-4 inhibitors was still significant at 180 days (4.9 vs 2.2/1000 person-years; HR, 2.2), as well as at 30 days (7.5 vs 3.3/1000 person-years; HR, 2.3) and 60 days (5.6 vs 2.3; HR, 2.5).
The DKA risk at 180 days was also significantly higher with SGLT2 inhibitors among patients not taking insulin (2.5 vs 1.0; HR, 2.5).
"Still So Much Work to Be Done"
Once the investigators had the data, they strove to get them published as quickly as possible — hence in a research letter rather than a full paper, Dr Fralick told Medscape Medical News, adding, "There's still so much work to be done to identify specific risk factors."
Meanwhile, the group is using the same database to examine the risk of amputations with canagliflozin; results are expected in a few weeks.
The study was supported by the division of pharmacoepidemiology and pharmacoeconomics, department of medicine, Brigham and Women's Hospital, Harvard Medical School, Boston. Dr Fralick reports grants from the University of Toronto Clinician Scientist Program and Clinician Investigator program and grants from the Detweiller Traveling Fellowship funded by the Royal College of Physicians and Surgeons of Canada, outside the submitted work. Disclosures for the coauthors are listed on the journal website.
N Engl J Med. 2017;376:2300-2302.
Risk of Diabetic Ketoacidosis after Initiation of an SGLT2 Inhibitor
N Engl J Med 2017; 376:2300-2302June 8, 2017DOI: 10.1056/NEJMc1701990
Article To the Editor:
Inhibitors of sodium–glucose cotransporter 2 (SGLT2) decrease plasma glucose by blocking the reabsorption of glucose at the proximal tubule.1,2 Case reports have suggested that SGLT2 inhibitors may be associated with an increased risk of diabetic ketoacidosis, which led to a warning from the Food and Drug Administration (FDA) in May 2015.3,4 The objective of our study was to assess the risk of diabetic ketoacidosis after the initiation of an SGLT2 inhibitor.
Using a large claims database of commercially insured patients in the United States (Truven MarketScan), we identified a cohort of adult patients (≥18 years of age) who had newly started treatment with either an SGLT2 inhibitor or a dipeptidyl peptidase-4 (DPP4) inhibitor between April 1, 2013, and December 31, 2014 (before the FDA warning). DPP4 inhibitors were chosen as the comparator medication because they are similarly used as a second-line treatment for diabetes but have no known association with diabetic ketoacidosis. We excluded patients with human immunodeficiency virus infection, end-stage renal disease, cancer, type 1 diabetes, or past diabetic ketoacidosis. Our primary outcome was hospitalization for diabetic ketoacidosis (using the primary position code of the International Classification of Diseases, Ninth Revision) within 180 days after the initiation of an SGLT2 inhibitor or a DPP4 inhibitor. We censored data for patients at the time that they discontinued the initial medication, had the outcome, lost insurance coverage, or died.
We used 1:1 propensity-score matching to balance 46 characteristics of the patients and Cox regression to estimate hazard ratios and 95% confidence intervals for diabetic ketoacidosis within 180 days after treatment initiation. Predefined sensitivity analyses included shorter durations of follow-up (30 days and 60 days). All statistical analyses were performed with the use of the validated Aetion platform and R software, version
We identified 50,220 patients who had received a new prescription for an SGLT2 inhibitor and 90,132 who had received a new prescription for a DPP4 inhibitor. Patients who were receiving SGLT2 inhibitors were younger and had fewer coexisting illnesses than those receiving DPP4 inhibitors but were more likely to receive insulin. After propensity-score matching was performed, these differences were well balanced (Table 1TABLE 1Characteristics of the Patients at Baseline.). Before propensity-score matching, the unadjusted rate of diabetic ketoacidosis within 180 days after the initiation of an SGLT2 inhibitor was about twice the rate after the initiation of a DPP4 inhibitor (4.9 events per 1000 person-years vs. 2.3 events per 1000 person-years) (hazard ratio, 2.1; 95% confidence interval [CI], 1.5 to 2.9). After propensity-score matching, the hazard ratio was 2.2 (95% CI, 1.4 to 3.6) (Table 2TABLE 2Primary and Other Outcomes.). The results were robust across sensitivity analyses.
In conclusion, shortly after initiation, SGLT2 inhibitors were associated with approximately twice the risk of diabetic ketoacidosis as were DPP4 inhibitors, although cases of diabetic ketoacidosis leading to hospitalization were infrequent. The increased risk of diabetic ketoacidosis with SGLT2 inhibitors is among the factors to be considered at the time of prescribing and throughout therapy if patients present with symptoms suggestive of diabetic ketoacidosis.
www red DiabetologNytt