• Canvas
• Duke Hub
• Program Website
• Application Website

• CRP 271, 13:30 - 15:00
• CRP 248, 15:15 - 16:30

• CRP 254
• CRP 245

Project

Factor XIa Inhibition (Abelacimab) for ECMO and Mechanical Support

• The Drug:
  • Abelacimab (a novel Factor XIa inhibitor).
• The Concept:
  • A Phase 3 trial evaluating whether a Factor XIa inhibitor reduces circuit thrombosis and bleeding compared to standard unfractionated heparin in patients on mechanical circulatory support.

Title: Factor XIa Inhibition with Abelacimab versus Unfractionated Heparin for the Prevention of Circuit Thrombosis and Bleeding in Adults Requiring Venoarterial Extracorporeal Membrane Oxygenation: A Phase 3, Randomized, Active-Controlled Trial (the XI-SUPPORT VA Trial)

Phase: 3 | Design: Multicenter, randomized, active-controlled, blinded-endpoint (PROBE) | Population: Adult VA-ECMO | Investigational agent: Abelacimab (dual Factor XI/XIa inhibitor)

[Principal Investigator, Institution, Contact — to be inserted]

Venoarterial extracorporeal membrane oxygenation (VA-ECMO) sustains patients with refractory cardiogenic shock and cardiac arrest, but blood contact with the large non-endothelial surfaces of the oxygenator, pump, and cannulae rapidly activates platelets and coagulation. Systemic anticoagulation is therefore obligatory: in the VA configuration, blood is returned directly to the systemic arterial circulation, making thromboembolic stroke a first-order risk, and the poorly-ejecting supported ventricle is prone to intracardiac stasis and thrombus. Unlike venovenous ECMO—where anticoagulation-free strategies are increasingly used and where systemic anticoagulation has not been clearly associated with improved survival—VA-ECMO retains clear equipoise for the type, but not the presence, of anticoagulation. This makes VA-ECMO the cleanest population in which to test a novel anticoagulant against standard care.

Despite anticoagulation, both circuit/device thrombosis and hemorrhage remain leading, independently lethal complications of VA-ECMO. The standard of care, unfractionated heparin (UFH), acts indirectly via antithrombin to inhibit thrombin and Factor Xa—central effectors of physiologic hemostasis—so the intensity required to protect the circuit inherently impairs the patient's protective clotting, producing the coexistence of thrombosis and bleeding. UFH is further limited by an unpredictable dose response, heparin resistance in inflammatory/critical illness, heparin-induced thrombocytopenia, and the absence of a monitoring assay that reliably reflects the in vivo effect.

A mechanistically distinct strategy targets the contact activation pathway. Blood contact with artificial surfaces activates Factor XII, which activates Factor XI; amplification of Factor XIa drives the surface-mediated thrombin generation responsible for device-associated thrombosis. Critically, Factor XI plays at most a supporting role in normal hemostasis. Inhibiting Factor XI/XIa therefore offers hemostasis-sparing anticoagulation: suppression of pathological, device-triggered clotting without the bleeding penalty of thrombin- and Xa-directed agents. Preclinical extracorporeal-circuit and ECMO models using Factor XI or Factor XII inhibition show antithrombotic efficacy comparable to heparin with substantially less bleeding, and Factor XIa–directed antibodies potently inhibit clotting across simulated hemodialysis, mechanical left-ventricular support, and ECMO circuits.

Abelacimab is a fully human monoclonal antibody that binds Factor XI, locking it in the inactive zymogen conformation and providing dual Factor XI/XIa inhibition. Intravenous administration produces near-maximal (~99%) Factor XI inhibition within roughly one hour and sustains it for up to 30 days—pharmacology suited to the acute initiation and multi-day duration of VA-ECMO. The promise of hemostasis-sparing anticoagulation was substantiated in the Phase 2b AZALEA–TIMI 71 trial, in which abelacimab reduced major or clinically relevant non-major bleeding by ~62% versus rivaroxaban in atrial fibrillation, prompting early termination for benefit. No adequately powered trial has tested Factor XIa inhibition against heparin in VA-ECMO, where the pathophysiology most directly implicates the contact pathway. XI-SUPPORT VA addresses this high-priority, evidence-poor gap.

Primary objective: To determine whether abelacimab, versus UFH, reduces net adverse clinical events (a composite of major thrombotic and major bleeding events) in adults on VA-ECMO.

Primary hypothesis: Abelacimab is superior to UFH for the reduction of net adverse clinical events through the treatment period.

Secondary objectives: To compare abelacimab and UFH for (a) circuit/device thrombosis and unplanned circuit/component exchange; (b) patient-level thromboembolism, including ischemic stroke and systemic/arterial embolism; © major bleeding (ISTH); (d) transfusion burden; (e) all-cause mortality and successful weaning from VA-ECMO; and (f) safety, including decannulation-site bleeding and events attributable to sustained Factor XI inhibition.

Secondary hypotheses: Abelacimab reduces thrombotic endpoints without increasing—and plausibly while reducing—bleeding endpoints relative to UFH.

Inclusion criteria 1. Age ≥ 18 years. 2. Initiation of VA-ECMO, or randomization at the transition from the initial anticoagulation bolus to maintenance anticoagulation (including extracorporeal CPR [ECPR] cannulations). 3. Anticipated duration of support ≥ 72 hours. 4. Requirement for systemic anticoagulation per standard practice. 5. Written informed consent from the patient or legally authorized representative.

Exclusion criteria 1. Active clinically significant bleeding, defined as bleeding requiring surgical or interventional-radiology hemostasis, an ongoing transfusion requirement (≥ 2 units of red cells in the preceding 12–24 hours), or bleeding at a critical site—with a residual investigator-judgment clause. 2. Confirmed or suspected heparin-induced thrombocytopenia (4T score ≥ 4 and/or a HIT workup sent with heparin held); such patients require a systemic direct thrombin inhibitor and are addressed in a separate exploratory pathway rather than the primary cohort. 3. Requirement for therapeutic anticoagulation for a separate indication (e.g., mechanical valve, acute venous thromboembolism). 4. Concomitant durable ventricular assist device or a second/parallel extracorporeal circuit. (A concurrently running Impella for LV unloading is permitted—see §4—as are surgical LV vents and intra-aortic balloon pumps.) 5. Platelet count < 30 × 10⁹/L, or severe coagulopathy not attributable to the qualifying condition. 6. Recent (< 14 days) major surgery unrelated to cannulation, or intracranial hemorrhage; known intracranial vascular malformation or neoplasm. 7. Known hypersensitivity to abelacimab or excipients. 8. Pregnancy or lactation. Rationale: abelacimab is an investigational IgG monoclonal antibody with no reproductive-safety data and undergoes active transplacental transfer, posing unquantified fetal risk from a long-acting, non-reversible anticoagulant; UFH does not cross the placenta and remains available to these patients as standard care. 9. Anticipated survival < 48 hours or transition to comfort-focused care. 10. Participation in another interventional anticoagulation trial.

Eligible patients are randomized 1:1 to abelacimab or UFH, continued for the duration of VA-ECMO up to a maximum of 30 days. A standard-of-care heparin bolus at cannulation is permitted before randomization (patients are frequently emergent), and randomization occurs at the transition to maintenance anticoagulation.

• Abelacimab arm: The maintenance heparin infusion is stopped and a single intravenous abelacimab load is administered, achieving near-complete Factor XI inhibition within ~1 hour; no heparin washout interval is required or imposed. Because heparin self-clears over ~1–2 hours as abelacimab reaches maximal effect, the brief overlap maintains circuit protection during transition without creating an anticoagulation gap or an arm-asymmetric thrombosis window. Re-dosing follows protocol pharmacokinetics for support exceeding the dosing interval. Given the prolonged half-life and absence of a specific reversal agent, protocol-specified peri-decannulation and rescue-hemostasis management (factor concentrates and, where indicated, platelet support) is prescribed.
• Control (UFH) arm: Weight-based UFH by continuous infusion, titrated to a protocol-specified target (a common anti-Xa and/or aPTT band applied at all sites). Standardizing the comparator is essential: site-level variation in anticoagulation intensity would render the control condition undefined and the effect estimate uninterpretable.
• Concurrent Impella (ECpella): Permitted, but any running Impella in both arms must use a sodium-bicarbonate-based purge solution, so that no patient receives device-delivered heparin. Systemic direct thrombin inhibitors are not permitted in either arm. Surgical LV vents and intra-aortic balloon pumps require no special handling beyond the systemic protocol.

Concomitant antiplatelet therapy (including aspirin and P2Y12 inhibitors, common after PCI/ECPR) is permitted, recorded, and managed identically across arms; transfusion thresholds and circuit management follow institutional protocols applied identically across arms. Bridging to durable devices, transplantation, or recovery is permitted and captured as an outcome.

Method of randomization. Central, computer-generated 1:1 randomization using permuted blocks of randomly varied size drawn from a pre-specified set (e.g., sizes of 4 and 6), implemented through a concealed interactive-response system; random block sizes prevent allocation prediction in an open-label design, while the maximum block size bounds within-stratum imbalance. Stratification: enrolling site, ECPR versus non-ECPR indication, cannulation strategy (peripheral vs. central), baseline bleeding-risk category, and baseline antiplatelet/DAPT use. Because the anticoagulants differ in monitoring, the trial uses a PROBE design—open-label treatment with all endpoints adjudicated by an independent Clinical Events Committee blinded to allocation.

Definition of endpoints. - Primary — net adverse clinical events (NACE): first occurrence of any adjudicated major thrombotic or major bleeding event during the treatment period (support initiation through 48 hours after VA-ECMO discontinuation or day 30, whichever is first). - Major thrombotic event: circuit/oxygenator thrombosis requiring unplanned circuit or component exchange or causing circuit failure; or a patient-level arterial or venous thromboembolic event (ischemic stroke, systemic/arterial embolism, symptomatic pulmonary embolism, or symptomatic deep-vein thrombosis) confirmed by imaging or direct inspection. - Major bleeding: ISTH criteria adapted for critical illness—fatal bleeding; symptomatic critical-site bleeding (e.g., intracranial, intraspinal, pericardial); a hemoglobin fall ≥ 2 g/dL; or transfusion of ≥ 2 units of red cells. - Key secondary endpoints: individual components above; clinically relevant non-major bleeding; total transfused red-cell, platelet, and plasma units; number of unplanned circuit exchanges; all-cause mortality; successful weaning; and decannulation-site major bleeding.

Sample size justification. VA-ECMO carries a high combined thrombosis-plus-bleeding burden; a control NACE incidence of ~45% is a reasonable base-case assumption. To detect a clinically meaningful 30% relative reduction (to 31.5%) at two-sided α = 0.05 with 90% power requires ~298 patients per arm; with ~10% inflation for withdrawal, endpoint-unrelated early death, and incomplete follow-up, the target is ~300 per arm (≈ 600 total). A more conservative scenario (40% → 28%) would require ~360 per arm (~720 total). The final estimate will be confirmed by blinded review of the aggregate event rate during enrollment.

Methods of data analysis. The primary analysis is intention-to-treat, comparing time to first NACE with a Cox proportional-hazards model stratified by the randomization strata; the hazard ratio with 95% CI is the primary estimand and a stratified log-rank test the primary significance test. Because early death is common in VA-ECMO, cumulative incidence will be assessed with death as a competing risk (Fine–Gray) in sensitivity analyses. Individual thrombotic and bleeding components are analyzed as key secondary endpoints within a pre-specified hierarchical sequence controlling family-wise error; a win-ratio prioritizing thrombosis over bleeding is pre-specified as a supportive hierarchical composite. Recurrent events (circuit exchanges, transfusion counts) use Andersen–Gill or negative-binomial models. Pre-specified subgroups include ECPR status, cannulation strategy, sex, renal function, baseline DAPT, and heparin-resistance status, with interaction testing; a pre-specified sensitivity analysis excludes patients with a concurrently running Impella to address residual device heterogeneity. Safety analyses use the as-treated population. An independent DSMB reviews pre-specified interim efficacy/futility analyses using group-sequential α-spending, with heightened attention to bleeding, thrombosis, and mortality given abelacimab's prolonged pharmacodynamic effect. Missing data are handled by multiple imputation with tipping-point sensitivity analyses.

This letter of interest summarizes the scientific rationale and proposed design for discussion with the sponsor and coordinating center. Detailed dosing, pharmacovigilance, and operational procedures will be specified in the full protocol.

1. Yi BA, et al. Pharmacokinetics and pharmacodynamics of abelacimab (MAA868), a dual inhibitor of FXI and FXIa. J Thromb Haemost. 2022;20:307–315. 2. AZALEA–TIMI 71 investigators. Abelacimab versus rivaroxaban in atrial fibrillation. N Engl J Med. 2025;392(4):361–371. 3. Contact-pathway (FXI/FXII) inhibition in ECMO and extracorporeal-circuit models — simulated ECMO, rabbit/baboon, and FXIa-antibody (osocimab) circuit studies (2020–2025). 4. Suzuki G, et al. Systemic anticoagulation and outcomes in ARDS patients receiving VV-ECMO: multicenter propensity-weighted study (basis for VA-only restriction). 5. SCAI/consensus guidance on Impella anticoagulation and sodium-bicarbonate purge; FDA PMA approval of bicarbonate-based purge (2022). 6. NCT03707418 (Bivalirudin vs Heparin in ECMO) — precedent for randomization at transition to maintenance anticoagulation with a standard-of-care bolus at cannulation. 7. ISTH major-bleeding criteria; ELSO/registry epidemiology of VA-ECMO thrombosis and bleeding.