Patients with type 2 diabetes who took saxagliptin or sitagliptin had a similar risk for hospitalized heart failure compared with those who received antihyperglycemic agents, according to a retrospective cohort study.

Lead researcher Sengwee Toh, ScD, of Harvard Medical School and the Harvard Pilgrim Health Care Institute in Boston, and colleagues published their results online in the Annals of Internal Medicine on April 25.

Saxagliptin and sitagliptin are dipeptidyl peptidase-4 (DPP-4) inhibitors, a class of oral antihyperglycemic medications. The researchers noted that postmarketing trials have found conflicted results about the cardiovascular safety of DPP-4 inhibitors. For instance, the SAVOR-TIMI 53 study found patients who received saxagliptin had a 27 percent higher relative incidence of hospitalized heart failure compared with a placebo group. However, two other trials did not find significant differences in the risk of hospitalized heart failure among patients who received alogliptin or sitaglipton versus placebo.

This analysis was part of an ongoing active surveillance project to complement SAVOR-TIMI 53 and was conducted within Mini-Sentinel, a pilot program that assists the FDA in developing a national active surveillance system of FDA-regulated products. As of August 2014, the Mini-Sentinel database included data on 178 million persons between 2000 and 2014 from 18 administrative claims and clinical data partners.

The FDA funded the study and was involved in its design, conduct and reporting.

The researchers evaluated patients who were at least 18 years old with type 2 diabetes who initiated treatment with saxagliptin, sitagliptin, pioglitazone, second-generation sulfonylureas or long-acting insulin products from 2006 to 2013. This analysis included 788,553 saxagliptin users and 298,124 sitagliptin users.

The researchers said they chose the comparators because they were common alternatives to saxagliptin in clinical practice. They defined type 2 diabetes as patients who had at least one prescription for an oral antihyperglycemic medication other than metformin or at least one diagnosis of diabetes plus at least one prescription for metformin.

The mean age of patients was approximately 60 years old, and approximately 55 percent were males. The age and sex distributions were similar between the new users of the study medications. The proportion of patients with a prior heart failure diagnosis was 5 percent for saxagliptin users, 7 percent for sitagliptin users, 7 percent for sulfonylurea users and 11 percent for insulin users.

The researchers said the risk for hospitalized heart failure was not higher with DPP-4 inhibitors compared with the other medications. They said the hazard ratios from the disease risk score-stratified analyses were 0.83 for saxagliptin versus sitagliptin, 0.63 for saxagliptin versus pioglitazone, 0.69 for saxagliptin versus sulfonylureas, 0.61 for saxagliptin versus insulin, 0.74 for sitagliptin versus pioglitazone, 0.86 for sitagliptin versus sulfonylureas and 0.71 for sitagliptin versus insulin.

They added that results were similar from 1:1 propensity score-matched analyses and subgroup analyses of patients with and without prior cardiovascular disease.

“By comparing DPP-4 inhibitor users and users of other antihyperglycemic agents who received these treatments in routine clinical practice, our study provides information that complements recently completed postmarketing placebo-controlled trials,” the researchers wrote. “Our findings are clinically relevant because patients and physicians often choose among various treatment alternatives (including no treatment) for [type 2 diabetes] in practice.”

The researchers provided a few explanations for why their results differed from the SAVOR-TIMI 53 trial, including population differences and limitations of an observational study design. They also obtained data from patients who received antihyperglycemic treatments in routine ambulatory clinical settings, which may differe from participants in other studies.

In this analysis, patients who received saxagliptin were younger, were less likely to have a prior history of heart failure or MI and had less concurrent insulin use at baseline compared with saxagliptin users from the SAVOR-TIMI 53 study. Further, the mean follow-up in this analysis was less than one year compared with 2.1 years in the SAVOR-TIMI 53 trial.

“Additional investigations are needed to better understand the relation between DPP-4 inhibitors and [hospitalized heart failure] risk,” the researchers wrote. “Well-designed randomized trials with [hospitalized heart failure] as the main end point or observational studies that address the limitations of our study will help provide more definitive evidence on the topic.”