Conference Coverage

ENGAGE AF-TIMI: Insulin linked to greater risk for stroke, CV death, bleeding



– Patients with diabetes had significantly higher adjusted risk of bleeding, cardiovascular-related death, and poorer net outcomes, particularly those treated with insulin, a subanalysis of the ENGAGE AF-TIMI 48 trial has shown.

Dr. Anna Plitt cardiology fellow, The Mount Sinai Hospital, New York. Doug Brunk/MDedge News

Dr. Anna Plitt

In addition, the pharmacokinetic and pharmacodynamic profile of the study drug, edoxaban – a novel oral anticoagulant drug and a direct factor Xa inhibitor – was generally similar in patients with and without diabetes.

“We know that atrial fibrillation is associated with a fivefold increased risk of stroke,” Anna Plitt, MD, said at the World Congress on Insulin Resistance, Diabetes & Cardiovascular Disease. “Type 2 diabetes is associated with a twofold increased risk of stroke, and longer duration of diabetes is associated with even higher ischemic event rates. The coexistence of [atrial fibrillation] and type 2 diabetes further increases thromboembolic risk.”

Dr. Plitt, a cardiology fellow at Mount Sinai Hospital, New York, noted that, although type 2 diabetes is characterized by a prothrombotic and inflammatory state, the mechanism of action by which hyperglycemia and/or insulin resistance leads to the development of atrial fibrillation (AFib) remains unknown. “Given the complex clinical interactions between AFib and type 2 diabetes, care for these patients remains challenging,” she said. “Recommendations for anticoagulation managements vary based on the presence of additional risk factors and which guidelines are followed.”

In the ENGAGE AF-TIMI 48 trial, 21,105 patients with documented AFib within the previous 12 months were randomized to standard-care warfarin or high-dose edoxaban (60 mg daily) or low-dose edoxaban (30 mg daily). The edoxaban dose was reduced by 50% if creatinine clearance reached 30-50 mL/min, patient weight reached 60 kg or less, or there was concomitant use of a P-glycoprotein inhibitor (N Engl J Med. 2013;369:2093-104). The median follow-up was 2.8 years, and the primary efficacy endpoint was stroke or systemic embolic events (SEEs). The primary safety endpoint was major bleeding, as defined by the International Society on Thrombosis and Haemostasis criteria.

The findings showed that edoxaban was noninferior to warfarin in preventing stroke/SEEs. It also significantly reduced major bleeding, cardiovascular death, and net outcomes. “Therefore, the higher dose of edoxaban was approved globally for treating patients with AFib,” Dr. Plitt said. “The lower-dose regimen was not approved because there was less protection from ischemic stroke, compared with warfarin.”

For the current subanalysis, Dr. Plitt and colleagues set out to further evaluate outcomes of patients enrolled in the ENGAGE AF-TIMI 48 trial, excluding those who were in the low-dose edoxaban group. The presence or absence of diabetes was determined by the local investigator at randomization. The investigators further stratified patients into insulin-treated and non–insulin treated groups and used multivariate Cox regression models to adjust for baseline characteristics across the groups stratified by diabetes status. Next, they analyzed edoxaban concentration, anti–factor Xa activity, and international normalized ratio data and compared outcomes of high-dose edoxaban with those of warfarin.

The primary endpoint and the primary safety endpoint of interest were the same as in the main ENGAGE AF-TIMI 48 trial. Key secondary endpoints included in the subanalysis were cardiovascular death, stroke/SEE, major adverse cardiovascular events (MACE, a composite of myocardial infarction, stroke, SEE, or death because of cardiovascular cause or bleeding), and all-cause death.

In all, 7,624 of the 21,105 patients in the ENGAGE AF-TIMI 48 trial had diabetes, for a rate of 36%. Most of the patients with diabetes did not require insulin (30%), while 6% did. There were fewer female patients with diabetes than without (37% vs. 39%, respectively). Of note was that history of prior stroke/transient ischemic attack was higher in the no-diabetes group than in the diabetes group (33% vs. 21%), as was congestive heart failure (63% vs. 48%).

The mean CHA2DS2-VASc score for predicting thromboembolic risk (0, low risk; greater than 1, high risk) was 4.6 in the diabetes group and 4.2 in the no-diabetes group. When diabetes was not included in the score, the mean CHA2DS2-VASc score was 3.6 in the diabetes group. “Because the trial entry criteria required a minimum CHADS2 score of 2, patients without diabetes were enriched with stroke risk factors other than diabetes,” Dr. Plitt said.

Adjusted outcomes from the subanalysis showed that the risk of stroke/SEE was similar between patients with and without diabetes (hazard ratio, 1.08). However, patients with diabetes were at higher adjusted risk for cardiovascular death than patients without diabetes (HR, 1.29), MACE (HR, 1.28), major bleed (HR, 1.28), and the net outcome of stroke, SEE, major bleed, or all-cause death (HR, 1.25).

The researchers also analyzed the pharmacodynamic and pharmacokinetic data of high-dose edoxaban, stratified by diabetes status. They found that the parameters were generally similar between patients with and without diabetes, including trough concentrations of edoxaban (34.3 and 37.2 ng/mL, respectively; P = .04), trough exogenous anti–factor Xa activity (0.59 and 0.68 IU/mL; P = .11), and the percentage change from baseline in the peak endogenous anti–factor Xa activity (P = .66). The percentage changes from baseline of the trough endogenous anti–factor Xa activity was slightly lower in patients with diabetes, compared with patients without diabetes (P less than .001). “However, these modest differences between the two groups are of unclear clinical significance,” Dr. Plitt said.

Results from the main ENGAGE AF-TIMI 48 showed that the rates of stroke/SEE were reduced by 13% on high-dose edoxaban. However, the subanalysis found no significant effect modification in the reduction in stroke/SEE with edoxaban, compared with warfarin, when stratified by diabetes status (reductions of 16% vs. 7% in the no-diabetes and diabetes groups, respectively; P for interaction = .54). The researchers also observed similar reductions with edoxaban in the risks of secondary outcomes when patients were stratified by diabetes status.

In another finding, patients with diabetes who were treated with insulin were at a higher adjusted risk for all outcomes, compared with those with diabetes who were not treated with insulin. This included stroke/SEE (HR, 1.44), cardiovascular-related death (HR, 1.83), MACE (HR, 1.78), major bleed (HR, 1.31), and net outcome (HR, 1.57).

Next, the researchers compared the study endpoints of high-dose edoxaban and warfarin, with and without insulin. “None of the efficacy, safety, or net outcomes demonstrated evidence of treatment effect modification related to the use of insulin among [patients with diabetes],” she said.

Dr. Plitt disclosed having received honoraria for educational activities from Bristol-Myers Squibb.

[email protected]

Recommended Reading

DAPA-HF results transform dapagliflozin from antidiabetic to heart failure drug
The Hospitalist
Machine learning–derived risk score predicts heart failure risk in diabetes patients
The Hospitalist
FDA adds diabetic kidney disease, heart failure indications to canagliflozin
The Hospitalist
Certain diabetes drugs may thwart dementia
The Hospitalist
Dapagliflozin approved for reducing HF hospitalization in diabetes
The Hospitalist
New model for CKD risk draws on clinical, demographic factors
The Hospitalist
Glycemic Control eQUIPS yields success at Dignity Health Sequoia Hospital
The Hospitalist
DAPA-HF: Dapagliflozin’s HFrEF efficacy confirmed in nondiabetics
The Hospitalist
DAPA-HF: Dapagliflozin benefits regardless of age, HF severity
The Hospitalist
New heart failure trial data presage guideline revisions
The Hospitalist
   Comments ()