PDF: RE-ALIGN
(2013)Objective
To evaluate the safety and efficacy of dabigatran as an alternative to warfarin in patients with mechanical heart valves for the prevention of thromboembolic events.
Study Summary
• Dabigatran was associated with increased rates of thromboembolic and bleeding events in patients with mechanical heart valves, leading to early termination of the trial.
Intervention
Patients with recent or remote mechanical aortic or mitral valve replacement were randomized to receive dabigatran (150–300 mg twice daily) or dose-adjusted warfarin, with the goal of comparing thromboembolic and bleeding outcomes.
Inclusion Criteria
Patients aged ≥18 years with mechanical heart valve prostheses in the aortic or mitral position, either within 7 days of surgery (early cohort) or after 3 months post-surgery (late cohort).
Study Design
Arms: Dabigatran vs. Warfarin (INR 2.0–3.5 target)
Patients per Arm: Dabigatran: 126, Warfarin: 40
Outcome
• Thromboembolic events: 9 (7 strokes, 1 TIA, 1 MI) in dabigatran group vs. none in warfarin group; • Major bleeding: 7 in dabigatran group vs. 2 in warfarin; • All strokes and bleeding events occurred in the dabigatran arm; • Trial was terminated early due to excess events in dabigatran group; • No thromboembolic or major bleeding events in warfarin group.
Bottom Line
The use of dabigatran in patients with mechanical heart valves was associated with increased rates of both thromboembolic events (stroke) and bleeding complications compared to warfarin. The trial was terminated prematurely, demonstrating that dabigatran should not be used in this patient population.
Major Points
- RE-ALIGN is the definitive safety trial establishing that DOACs should NEVER be used in mechanical heart valve patients — a critical 'do not use' landmark that every clinician prescribing DOACs must know.
- Phase 2 dose-validation study terminated prematurely by DSMB due to excess of BOTH thromboembolic AND bleeding events with dabigatran — the worst possible combination of outcomes (more clots AND more bleeding simultaneously).
- Stroke: 5% (9/168) dabigatran vs 0% (0/84) warfarin — all strokes occurred in the dabigatran group. Additionally, 5 patients had asymptomatic valve thrombosis detected on echocardiography.
- Bleeding paradox: despite inadequate anticoagulation for valve thromboprevention, dabigatran caused MORE bleeding: any bleeding 27% vs 12% (HR 2.45, p=0.01); major bleeding 4% vs 2%. This paradox reflects that valve thrombosis and bleeding operate through different mechanisms.
- 252 patients across 39 centers in 10 countries. Two populations enrolled: Population A (79% of cohort) = recent mechanical valve surgery within 7 days; Population B = mechanical mitral valve >3 months post-surgery.
- Most thromboembolic events occurred in Population A (early post-operative) — suggesting dabigatran was particularly dangerous in the immediate post-surgical period when the valve-blood interface is most thrombogenic and requires broad-spectrum coagulation inhibition.
- Mechanism of failure (per paper): thrombin generation on mechanical valves is driven both by tissue factor released from surgically damaged tissue AND by contact-pathway activation on the artificial valve surface/sewing ring (which does not endothelialize for weeks). Warfarin suppresses both pathways by inhibiting synthesis of factor VII (tissue-factor pathway), factor IX (contact pathway), factor X, and thrombin, while dabigatran exclusively inhibits thrombin — so intense contact activation can overwhelm local dabigatran and generate thrombus on the valve.
- Dabigatran dosing was based on plasma trough levels (target ≥50 ng/mL) with doses up to 300 mg BID — even at these doses anticoagulation was inadequate for mechanical valves while simultaneously causing excess bleeding.
- Led to FDA black box warning (2012) and EMA contraindication for dabigatran in mechanical heart valves. Subsequently, ALL DOACs carry a contraindication for mechanical valves — a class-level prohibition based primarily on this single trial.
- For over a decade RE-ALIGN was the only randomized trial of a DOAC in mechanical valves; PROACT Xa (apixaban vs warfarin in mechanical On-X aortic valves) was subsequently launched and stopped early in 2022 for futility/harm, reinforcing the mechanical-valve DOAC contraindication.
Study Design
- Study Type
- Prospective, randomized, phase 2, open-label trial with blinded end-point adjudication.
- Randomization
- Yes
- Blinding
- Open-label with blinded endpoint adjudication.
- Sample Size
- 252
- Follow-up
- Planned for 12 weeks with an extension phase; terminated early.
- Centers
- 39
- Countries
- 10 countries, including Canada and nations in Western and Central Europe.
Primary Outcome
Definition: The primary endpoint of RE-ALIGN was the trough plasma level of dabigatran (a pharmacokinetic dose-validation endpoint), measured by HPLC–tandem mass spectrometry. The trial aimed to validate a dosing regimen that would keep <10% of patients below the 50 ng/mL trough target. The clinical composite (death, stroke, systemic embolism, or MI) was an additional pre-specified clinical outcome and is reported under Secondary Outcomes.
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| - | Trough plasma level ≥50 ng/mL achieved for a mean 86% of time overall (84% Population A, 96% Population B). Dose escalation or discontinuation of dabigatran required in 52/162 (32%) patients; 13/162 (8%) discontinued per protocol despite 300 mg BID because trough remained <50 ng/mL. | - | - |
Limitations & Criticisms
- Open-label design — though outcomes were adjudicated blindly, awareness of treatment assignment could have influenced monitoring intensity and clinical decision-making.
- Small sample size (252 patients) — the trial was never intended to be a definitive outcomes trial; it was a phase 2 PK/dose-validation study that happened to detect a clear harm signal.
- Only tested dabigatran (direct thrombin inhibitor) — cannot directly extrapolate to factor Xa inhibitors (rivaroxaban, apixaban, edoxaban), though the mechanism of failure (thrombin inhibition alone being insufficient against valve-surface contact activation) is plausibly shared.
- Population A (79%) dominated the results — most events occurred early post-op. It remains unclear whether chronic, stable mechanical valve patients might tolerate DOACs differently, though Population B also showed excess events.
- Dabigatran trough-guided dosing was novel and unvalidated — the target of ≥50 ng/mL was extrapolated from AF data and may have been inappropriate for the higher thrombogenic burden of mechanical valves.
- For over a decade the class-wide DOAC contraindication in mechanical valves rested on this single dabigatran trial; the subsequent PROACT Xa trial of apixaban vs warfarin in mechanical On-X aortic valves was also stopped early (2022, futility/harm), but only one factor Xa DOAC has been formally tested in this population.
- Very short planned follow-up (12 weeks) — even within this brief period, the harm was evident, suggesting the signal would only worsen with longer exposure.
- Industry-sponsored (Boehringer Ingelheim, manufacturer of dabigatran/Pradaxa) — to their credit, the manufacturer published the negative results promptly and updated the label with a contraindication.
- Does not address bioprosthetic valves — subsequent trials (RIVER, DAWA) have explored DOACs in bioprosthetic valves with more favorable results, highlighting the mechanical vs bioprosthetic distinction.
Citation
N Engl J Med 2013;369:1206-14.