PDF: AZALEA–TIMI 71
(2025)Objective
Abelacimab (XI & XIa inhibitor) monthly subcutaneous injection versus rivaroxaban in patients with atrial fibrillation at moderate-to-high risk of stroke.
Study Summary
• Stroke occurred slightly more often with abelacimab (XI and XIa inhibitor) than rivaroxaban, though not statistically significant, however Abelacimab showed a favorable bleeding profile, particularly for gastrointestinal bleeding.
Intervention
Subcutaneous abelacimab 90 mg or 150 mg monthly vs. oral rivaroxaban 20 mg daily (or 15 mg if creatinine clearance ≤50).
Inclusion Criteria
Age ≥55 years, atrial fibrillation/flutter, CHA₂DS₂-VASc ≥4 or ≥3 with additional risk (antiplatelet use or CrCl ≤50). Most had prior anticoagulation use.
Study Design
Arms: Abelacimab 150 mg vs. Abelacimab 90 mg vs. Rivaroxaban
Patients per Arm: Abelacimab 150 mg: 430, Abelacimab 90 mg: 427, Rivaroxaban: 430
Outcome
• Major or clinically relevant nonmajor bleeding (CRNM) bleeding: Rivaroxaban 8.38, Abelacimab 150 mg 3.22 (HR 0.38), 90 mg 2.64 (HR 0.31); P<0.001. <br>• Major bleeding: Rivaroxaban 3.73, Abelacimab 150 mg 1.22 (HR 0.33), 90 mg 0.99 (HR 0.26). <br>• GI bleeding: Rivaroxaban 2.14, Abelacimab 150 mg 0.24 (HR 0.11), 90 mg 0.25 (HR 0.11). <br>• Stroke/systemic embolism: Rivaroxaban 0.83, Abelacimab 150 mg 1.21, 90 mg 1.36. <br>• Ischemic stroke: Rivaroxaban 0.59, Abelacimab 150 mg 1.21, 90 mg 1.24. <br>• Hemorrhagic stroke: Rivaroxaban 0.23, 150 mg 0, 90 mg 0.12. <br>• All-cause death: Rivaroxaban 3.52, Abelacimab 150 mg 2.65, 90 mg 3.20.
Bottom Line
Monthly subcutaneous abelacimab significantly reduced major and clinically relevant nonmajor bleeding versus rivaroxaban, but ischemic stroke rates were numerically higher and efficacy was not established.
Major Points
- AZALEA-TIMI 71 is the landmark trial of Factor XI inhibition for AF — demonstrated that abelacimab (anti-FXI monoclonal antibody) reduced bleeding by 62-69% vs rivaroxaban. Published NEJM 2025.
- Mechanism: abelacimab locks Factor XI in its zymogen form, blocking the intrinsic pathway (thrombosis amplification) while preserving the extrinsic pathway (hemostasis). This 'uncouples' thrombosis from bleeding — the holy grail of anticoagulation.
- Primary outcome (major or CRNM bleeding): 3.22/100 PY (150 mg) and 2.64/100 PY (90 mg) vs 8.38/100 PY (rivaroxaban) — p<0.001 for both doses. GI bleeding nearly eliminated: 0.24-0.25 vs 2.14/100 PY.
- Trial stopped early by DSMB due to overwhelming bleeding benefit — but this early stop means efficacy (stroke prevention) could NOT be adequately assessed.
- Critical safety signal: NUMERICAL increase in ischemic stroke with abelacimab (1.21-1.36 vs 0.83/100 PY) — not statistically significant but concerning. Raises the fundamental question: can Factor XI inhibition prevent stroke as well as Factor Xa inhibitors?
- Monthly subcutaneous injection (vs daily oral rivaroxaban) — potentially transformative for adherence in AF patients. Achieved 97-99% reduction in free Factor XI.
- Part of the Factor XI inhibitor revolution: alongside asundexian (small molecule, OCEANIC-AF failed), milvexian (small molecule, LIBREXIA-AF ongoing), and osocimab (antibody). AZALEA is the first positive phase 2b trial in this class.
- Contrasts with OCEANIC-AF (asundexian vs apixaban) which was STOPPED for excess stroke — suggesting that the degree of FXI inhibition matters. Abelacimab achieves near-complete (>95%) FXI suppression, while asundexian achieved ~90%.
- Implications for neurology: if Factor XI inhibitors can prevent stroke with dramatically less ICH, they could be preferred for AF patients with prior ICH, cerebral amyloid angiopathy, or high fall risk.
- Phase 3 LILAC-TIMI 76 trial is underway comparing abelacimab to apixaban in AF — the definitive test of whether Factor XI inhibition provides both bleeding safety AND stroke prevention.
Study Design
- Study Type
- Multinational, phase 2b, randomized, partially-blinded, active-controlled
- Randomization
- Yes
- Blinding
- Double-blind for abelacimab dose; open-label for rivaroxaban
- Sample Size
- 1287
- Follow-up
- Median 2.1 years
- Centers
- 95
- Countries
- USA, Canada, Hungary, Czech Republic, Poland, Taiwan, South Korea
Primary Outcome
Definition: Major or clinically relevant nonmajor bleeding (ISTH criteria)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 8.38 per 100 person-years (rivaroxaban) | 3.22 per 100 PY (150 mg), 2.64 per 100 PY (90 mg) | - | <0.001 for both doses |
Limitations & Criticisms
- Early termination for bleeding benefit means the trial was UNDERPOWERED for efficacy (stroke/systemic embolism). The fundamental question — does FXI inhibition prevent stroke? — remains unanswered.
- Numerical increase in ischemic stroke (1.21-1.36 vs 0.83/100 PY) is concerning. While not significant, the OCEANIC-AF trial (asundexian) was stopped for similar stroke signal — pattern recognition across the FXI class raises red flags.
- Partially blinded design: abelacimab doses were blinded to each other, but the rivaroxaban arm was open-label (oral vs SC). This could influence reporting of subjective bleeding events.
- Comparator was rivaroxaban, not apixaban (the most-prescribed DOAC for AF). Rivaroxaban has higher GI bleeding rates than apixaban — the bleeding reduction may be smaller when compared to apixaban.
- Limited racial diversity — predominantly White participants. AF management and bleeding risk profiles differ across ethnicities, particularly in Asian populations (where lower DOAC doses are standard).
- Phase 2b sample size (n=1,287) insufficient for reliable efficacy conclusions. Need LILAC-TIMI 76 (phase 3, ~15,000 patients) to determine true stroke prevention efficacy.
- Monthly subcutaneous injection may be an advantage for adherence but is a disadvantage if urgent reversal is needed — no FXI-specific reversal agent exists. Half-life of abelacimab is ~25-30 days.
- Injection-site reactions (2-3%) could affect long-term adherence in a chronic condition requiring lifelong treatment.
- Excluded patients with prior ICH — ironically the population most likely to benefit from a non-bleeding anticoagulant. Phase 3 should include this subgroup.
Citation
Ruff CT, Patel SM, Giugliano RP, et al. Abelacimab versus Rivaroxaban in Patients with Atrial Fibrillation. N Engl J Med. 2025;392:361–71.