BRIDGE Study: Perioperative Bridging Anticoagulation in Patients with Atrial Fibrillation

Douketis and the BRIDGE Investigators. NEJM 2015. Published on line first 22nd June 2015. doi:10.1056/NEJMoa1501035

Clinical Question

• In elective surgical patients with atrial fibrillation who are suspending their normal warfarin peri-operatively, does no bridging anticoagulation compared to bridging anticoagulation with dalteparin increase the incidence of thromboembolism or reduce the incidence of major bleeding?

Design

• Randomised, double-blind, placebo-controlled trial
• Randomisation was stratified according to the study centre
• A revised calculation based on 1882 patients provided 90% power for each primary end point based on an arterial thromboembolism rate of 0.46% and bleeding rate of 2.3% in all patients
• Eisau donated the study drug (dalteparin) and the University of Iowa Pharmaceuticals prepared the matching placebo

Setting

• 108 hospitals in US and Canada
• July 2009 – December 2014

Population

• Inclusion: patients with chronic (permanent or paroxysmal) atrial fibrillation or flutter; 18 years or older; received warfarin therapy for 3 months or longer with an INR therapeutic range of 2-3; undergoing an elective operation or procedure that required interruption of warfarin therapy; at least one CHADS2 stroke risk factor
  • congestive heart failure or left ventricular dysfunction, hypertension, age of 75 years or older, diabetes mellitus, or previous ischaemic stroke, systemic embolism, or transient ischaemic attack (TIA)
• Exclusion: presence of a mechanical heart valve; stroke, systemic embolism, or TIA within the previous 12 weeks; major bleeding within the previous 6 weeks; creatinine clearance of < 30mls/min; platelets < 100; planned cardiac, intracranial or intraspinal surgery.
• 1813 patients
  • baseline characteristics of patients were matched between groups
    • mean age 71.7
    • 2/3 male
    • mean CHADS2 score 2.35
    • 1/3 patients on aspirin
    • most common procedures were gastrointestinal (44%), cardiothoracic (17.2%) and orthopaedic (9.2%)

Intervention

• Placebo (no bridging therapy)
  • placebo administered subcutaneously twice daily

 Control

• Bridging anticoagulation therapy with low-molecular-weight heparin
  • 100 IU of dalteparin / kg of body weight administered subcutaneously twice daily

Management common to both groups

Outcome

• Primary outcome: incidence of arterial thromboembolism (stroke, TIA and systemic embolism) and major bleeding at 30 days
  • arterial thromboembolism: 3 patients (0.3%) in the bridging group vs. 4 patients (0.4%) in the no-bridging group. 95% CI -0.6 to 0.8, P=0.73
  • major bleeding: 29 patients (3.2%) in the bridging group vs. 12 patients (1.3%) in the no-bridging group. RR 0.41; 95% CI 0.20 to 0.78, P=0.05
    • no fatal events
• Secondary outcome:
  • minor bleeding at 30 days
    • 187 patients (20.9%) in the bridging group vs. 110 patients (12%) in the no-bridging group. P=<0.001
  • no statistical difference in the following at 30 days:
    • acute myocardial infarction
    • deep-vein thrombosis
    • pulmonary embolism
    • death

Authors’ Conclusions

• In patients with atrial fibrillation who had warfarin treatment interrupted for an elective operation or other elective invasive procedure, forgoing bridging anticoagulation was noninferior to perioperative bridging with low-molecular-weight heparin for the prevention of arterial thromboembolism and decreased the risk of major bleeding

Strengths

• Randomised and multi-centre
• Allocation concealment
• An important study which addresses useful outcome measures
• Both the study drug and placebo were provided in identical vials

Weaknesses

• The CHADS2 score was low (mean 2.3) and only 3% of patients had scores of 5 or 6
• Patients undergoing major surgical procedures associated with high rates of arterial thromboembolism and bleeding (e.g carotid endarterectomy, major cancer surgery, cardiac surgery or neurosurgery) were not included in this study
• The power study was recalculated during the study as the rate of arterial thromboembolism was lower than expected. The planned recruitment numbers at the start of the study was 3282
• As a non-inferiority trial, it is important to avoid biases that may favour a smaller observed difference between the groups. The statistical approaches used in this trial (e.g. intention-to-treat analysis, absolute non-inferiority margin) may have led to a false positive conclusion (i.e. rejecting the null hypothesis of inferiority when the intervention is actually inferior)
• The majority of patients underwent procedures that were considered to be low bleeding-risk (89.4% based on pre-specified classification; 69.1% based on site investigator decision)
• No data is available regarding the number of patients with paroxysmal vs. permanent atrial fibrillation
• No data is available regarding the investigators decision to continue or stop anti-platelet therapy. A 1/3 of the patients were taking aspirin