The DANCE Trial is a multi-centre, randomized controlled trial comparing the safety of direct oral anticoagulants (DOAC) versus vitamin K antagonists (VKA) in the early period (30 days) after cardiac surgery in patients with atrial fibrillation requiring oral anticoagulation.

Approximately 36,000 Canadian adults undergo cardiac surgery annually. Of these patients, about 10% have a prior history of atrial fibrillation (AF). In the early post-operative period after cardiac surgery, 30-60% of patients develop AF and, by the time of discharge, 32% of patients who underwent cardiac surgery have an indication for oral anticoagulation (OAC). AF is associated with a significantly higher risk of stroke, even when transient, and OAC is the standard for thromboembolic prevention in these patients. In the post-operative period, the balance of benefits and risks of OAC may differ and the safest and most effective OAC in that patient population is uncertain.

Vitamin K antagonists (VKAs), such as warfarin or coumadin, are the most used anticoagulants after cardiac surgery. In the Left Atrial Appendage Occlusion Study (LAAOS) III that recruited 4811 patients from 105 centres in 27 countries, 77% of patients with AF on OAC were discharged on a VKA after cardiac surgery. Among patients taking a DOAC preoperatively, 55% were switched to a VKA after surgery. Over the first post-operative year, most of those patients were gradually transitioned back to a DOAC. Although effective, the use of VKAs is limited by a narrow therapeutic index requiring frequent international normalized ratio (INR) measurements to ensure appropriate levels of anticoagulation. This key limitation leads to non-compliance and discontinuation. In addition, in the first 3 months after cardiac surgery, time in the therapeutic range is low, even with close monitoring by experienced prescribers.

In the last decade, DOACs - inhibitors of factor Xa or thrombin- have become broadly used in patients with AF. Treatment with a DOAC in patients with AF has been demonstrated to yield a lower risk of stroke or systemic embolism and a similar risk of major bleeding when compared to VKAs during long-term follow-up. Moreover, DOACs are more convenient for both patients and clinicians. They have a rapid onset of effect, fixed dosage that obviates the need for regular monitoring, and few interactions with food and other medications. In the postoperative setting, DOACs may also lead to shorter length of stay and reduced costs.

The purpose of this study is to establish whether DOACs are as safe as VKAs in the first few weeks after heart surgery. The results of this study will impact the treatment of hundreds of thousands of patients in the world every year.

A subset of 910 DANCE participants with a recent bioprosthetic aortic and/or mitral valve replacement will be enrolled in the SUNDANCE substudy (Subclinical valve thrombosis in patients with surgical bioprosthetic valve replacement: An imaging substudy of the DANCE trial). SUNDANCE will examine the effects of DOACs versus VKAs on subclinical valve thrombosis and bioprosthetic valve function by conducting computed tomography (CT) scans and echocardiograms at 60 to 90 days after randomization.

Inclusion Criteria:

1. Age ≥18 years at the time of enrolment,
2. Open heart surgery in the last 10 days,
3. Atrial fibrillation requiring anticoagulation (including pre-existing or post-operative atrial fibrillation),
4. Informed consent from either the patient or a substitute decision-maker.

Exclusion Criteria:

1. Mechanical valve replacement,
2. Antiphospholipid syndrome (triple positive),
3. Severe renal failure (Cockcroft-Gault equation; creatinine clearance <15 ml/min),
4. Known significant liver disease (Child-Pugh classification B and C),
5. Left ventricular thrombus,
6. Ongoing bleeding, hemorrhagic disorders, or bleeding diathesis,
7. Known contraindication for any DOAC or VKA,
8. Women who are pregnant, breastfeeding, or of childbearing potential,
9. Surgery including left ventricular assist device implantation or cardiac transplantation,
10. Previously enrolled in this trial,
11. Follow-up not possible,
12. History of moderate or severe mitral valvular lesion (stenosis or regurgitation) that is not corrected during index cardiac surgery.

We will establish a plan for the full-scale study but there is no plan to make IPD available to other researchers.