Limited efficacy of existing agents

VKAs remain more efficacious than DOACs for several indications (Table 1). In patients with mechanical aortic or mitral valves, the RE‑ALIGN trial (Randomized, Phase II Study to Evaluate the Safety and Pharmacokinetics of Oral Dabigatran Etexilate in Patients after Heart Valve Replacement) found that dabigatran was associated with a higher risk of thromboembolic and bleeding events than warfarin.⁴ On‑X valves were purported to reduce thrombogenicity in comparison with conventional bileaflet valves, but when apixaban was compared with warfarin in patients with previously implanted On‑X aortic valves in the PROACT‑XA trial (A Trial to Determine if Participants With an On‑X Aortic Valve Can be Maintained Safely on Apixaban), apixaban was associated with a higher incidence of valve thrombosis or valve‑related thromboembolism.⁵ Superiority of VKAs relative to DOACs was also demonstrated for stroke prevention in AF in patients with rheumatic heart disease. The INVICTUS (Investigation of rheumatic AF Treatment Using Vitamin K Antagonists, Rivaroxaban or Aspirin Studies, Non‑Inferiority) trial found that in comparison with rivaroxaban, VKAs reduced the composite end point of stroke, systemic embolism, myocardial infarction (MI), and all‑cause mortality.⁶ Also, in antiphospholipid syndrome, pooled data from 4 randomized trials found a 5‑fold increased risk of arterial thrombosis with DOACs as compared with VKAs.⁷ Therefore, VKAs remain the standard of care for patients with triple positive disease antiphospholipid antibody syndrome, or a history of previous arterial events. There have also been concerns regarding the efficacy of standard DOAC dosing for stroke prevention in AF in patients with upper extremes of body weight. A recent patient‑level meta‑analysis, which included data from the 4 pivotal phase III trials found DOACs to be inferior to warfarin in patients with body mass index greater than or equal to 40 kg/m² with respect to all‑cause death and the net clinical outcome of stroke, systemic embolism, major bleeding, or death.⁸

High risk of bleeding

Despite their lower risks of bleeding as compared with VKAs, DOACs are relatively contraindicated in patients at the highest risk for bleeding. This includes patients with advanced liver disease, end‑stage renal disease, uncontrolled hypertension, severe thrombocytopenia, or cerebral amyloid angiopathy. Similarly, patients with a history of recurrent major bleeding or intracranial hemorrhage may not be suitable candidates for standard or even reduced doses of DOACs.

Patients with an indication for anticoagulation who also require concomitant antiplatelet therapy, for example, patients with AF who undergo coronary or peripheral revascularization, are at elevated bleeding risk, particularly in the acute setting.

Combined anticoagulant and antiplatelet therapy, or dual pathway inhibition (DPI) reduces the risk of thromboembolic events vs antiplatelet therapy alone in patients with acute and chronic cardiovascular disease. The ATLAS‑ACS (Rivaroxaban versus Placebo in Patients with Acute Coronary Syndromes) trial found that in patients with recent MI, the combination of rivaroxaban 2.5 mg twice daily and antiplatelet therapy compared with antiplatelet therapy alone reduced the risk of MI, stroke, or cardiovascular mortality.⁹ The VOYAGER‑PAD (Efficacy and Safety of Rivaroxaban in Reducing the Risk of Major Thrombotic Vascular Events in Subjects With Symptomatic Peripheral Artery Disease Undergoing Peripheral Revascularization Procedures of the Lower Extremities) trial demonstrated that, in patients with peripheral arterial disease after lower‑limb revascularization, the combination of rivaroxaban 2.5 mg twice daily and antiplatelet therapy compared with antiplatelet therapy alone reduced the rate of acute limb ischemia, amputation, MI, ischemic stroke, and cardiovascular mortality.¹⁰ However, both trials reported an increased risk of major bleeding, and in ATLAS‑ACS there was also an increase in intracranial hemorrhage risk.

In chronic cardiovascular disease, the COMPASS (Rivaroxaban for the Prevention of Major Cardiovascular Events in Coronary or Peripheral Artery Disease) trial found that DPI with low‑dose rivaroxaban and aspirin compared with aspirin alone also provided important benefits for the prevention of major adverse cardiovascular events but increased the risk of major bleeding, mainly from the gastrointestinal tract.¹¹ As such, patients with acute and chronic cardiovascular disease at high bleeding risk receiving DPI may experience an attenuated net clinical benefit from this therapy.

Other areas of unmet need

Patients with contraindications to DOACs include those with significant drug‑drug interactions, including certain anticonvulsant medications, antiretroviral drugs, chemotherapeutic agents, antifungal drugs, as well as patients who are pregnant or breastfeeding.

Mechanistic considerations

Traumatic vessel injury results in binding of extravascular tissue factor (TF) exposed at the site of injury to factor VII (FVII), thereby activating coagulation and driving hemostasis. Plaque rupture exposes intravascular TF to bind FVII and thereby also activates coagulation, but the amount of thrombin generation is insufficient to sustain thrombus formation and requires amplification via FXI. Thrombin also amplifies factors V and VIII, and the relative importance of FXI amplification as compared with other factors is unknown. FXI can also be activated by activated factor XII (FXIIa) via the contact pathway. FXII activation occurs after exposure to naturally occurring negatively charged entities (eg, neutrophil extracellular traps, polyphosphates, nucleic acids, activated platelets) or negative charge accumulation on medical devices devoid of a protective layer of endothelial cells, including mechanical heart valves, vascular grafts, stents, catheters, and dialysis circuits.^12,13

The relative importance of FXI amplification of thrombin generation in thrombus formation compared with hemostasis suggests that blocking thrombin amplification by targeting FXI could prevent pathologic thrombus formation after plaque rupture with little or no increase in bleeding. An additional potential benefit of FXI inhibition is to prevent thrombus formation on medical devices.

Epidemiologic data

First described in 1953 by Robert Rosenthal, inherited FXI deficiency (hemophilia C) is an autosomal dominant condition affecting up to 9% of Ashkenazi Jews.^14,15 Cohort studies have demonstrated that persons with inherited FXI deficiency have lower rates of thrombosis than the general population, with little or no increase in bleeding. Affected patients may be at increased risk of mucosal bleeding¹⁶ but have lower risk of both VTE and stroke,¹⁷ although not MI.¹⁸ These observations provide an epidemiologic basis for FXI as a potentially safe and effective antithrombotic target.

Experiments in animals

Animal models demonstrate that factor XI inhibition reduces thrombosis rates. FXI‑deficient mice have decreased rates of spontaneous, chemically‑induced, and catheter‑provoked thrombosis without obvious hemostatic abnormalities.¹⁹

Antisense oligonucleotides

ASOs bind to mRNA in hepatocytes, preventing its translation into the FXI protein, thereby reducing FXI levels to mimic a state of inherited FXI deficiency. The prolonged duration of action of ASOs makes them particularly attractive in patients requiring a short course of anticoagulation (eg, for total hip or knee arthroplasty), as a single dose may cover the entire treatment duration. The second‑generation ASO, fesomersen, has an N‑acetyl galactosamine (GalNAc) conjugation promoting maximum drug delivery to hepatocytes resulting in a longer half‑life. The onset of action of ASOs, however, remains limited by their mechanism of action; time is required to achieve a state of FXI deficiency after arresting protein synthesis by mRNA inhibition (circulating FXI half‑life is 50 h).

Antibodies

Monoclonal IgG antibodies under development include abelacimab, osocimab, and xisomab. These agents competitively or noncompetitively inhibit FXI and / or FXIa and have a more rapid onset of action than ASOs enabling their use in acute indications.

Small‑molecule inhibitors

Asundexian and milvexian are low‑molecular‑weight chemical compounds that competitively inhibit FXIa. With a rapid onset of action and oral administration, they are potentially advantageous for patients with chronic indications for anticoagulation. Their shorter half‑life may be desirable in patients who are at high risk of bleeding or require urgent surgery.

Hemostatic agents and specific reversal agents in patients with factor XI inhibitor–associated bleeding

FXI inhibitors prolong activated partial thromboplastin time (aPTT) in a dose‑dependent manner.²⁰ In patients treated with a FXI inhibitor who experience bleeding or require urgent surgery, clinicians may consider treatments to correct the aPTT or enhance hemostasis. In patients treated with an ASO, FXI replacement can correct the hemostatic defect. In patients treated with an antibody or small‑molecule inhibitor, FXI replacement is unlikely to be helpful but general hemostatic agents, such as prothrombin complex concentrates or recombinant factor VIIa (FVIIa), may provide benefits. In patients with inherited FXI deficiency, recombinant FVIIa has been shown to prevent bleeding complications and the need for perioperative blood products, even in severely deficient patients.²¹ Tranexamic acid appears to be a useful adjunctive agent, because FXI inhibitors may reduce activation of thrombin‑activatable fibrinolysis inhibitor leading to increased fibrinolysis.²² Defective fibrin clot structure has been identified as a possible contributor to bleeding in patients with FXI deficiency.²³

Although it is relevant to consider the need for specific reversal agents for FXI inhibitors, no patient in any of the published phase II FXI inhibitor trials required administration of FXI concentrate or a general hemostatic agent, such as a prothrombin complex concentrate or recombinant FVIIa.

Phase II clinical trials of factor XI inhibitors

Promising proof‑of‑concept studies have been published for several FXI inhibiting agents, including ASOs, monoclonal antibodies, and small‑molecule inhibitors.^24-31 Although none of the individual phase II trials were powered for efficacy, most showed similar bleeding rates as compared with placebo and lower bleeding rates as compared with active control (Table 3).

Further insights into the potential for FXI inhibitors come from pooled phase II trial data. A meta‑analysis of 4 phase II trials of FXI inhibitors in patients undergoing total knee replacement (TKR) demonstrated superiority of FXI inhibitors over prophylactic doses of LMWH with respect to efficacy (relative risk [RR], 0.62; 95% CI, 0.49–0.79), with greater efficacy with higher doses (P interaction <⁠0.001). FXI inhibitors were also associated with less bleeding (RR, 0.49; 95% CI, 0.31–0.77), although there was no apparent difference in safety between higher and lower doses (P interaction = 0.73). When data from 2 phase II trials comparing FXI inhibitors with apixaban after TKR or for stroke prevention in AF were pooled, results revealed no differences in efficacy or bleeding. Finally, when 3 phase II trials of small‑molecule inhibitors compared with placebo in patients with recent noncardioembolic stroke or MI were combined, there was no difference in the trial‑defined efficacy end point (RR, 1.02; 95% CI, 0.92–1.13) or major bleeding (RR, 1.21; 95% CI, 0.75–1.93), but there was an increase in clinically relevant nonmajor bleeding (RR, 1.25; 95% CI, 1.08–1.43).³²

Treatment indications

The AZALEA‑TIMI 71 study was not powered to demonstrate efficacy of abelacimab, and although OCEANIC‑AF demonstrated that the small‑molecule inhibitor, asundexian, was less effective than apixaban, it may still be more effective than no anticoagulation for stroke prevention. If so, this would be consistent with the epidemiologic data suggesting that FXI deficiency is associated with a lower risk of stroke.

Drug classes and agents

Different drug classes or specific agents within a class may have different effects. ASOs prevent production of FXI, antibodies prevent activation of FXI and / or inhibit FXIa, and small‑molecule inhibitors inhibit FXIa. It is unclear which of these mechanisms is associated with greatest benefit.

Dosing

Determining the optimal dose of a new anticoagulant to test in a phase III trial is always challenging, because earlier‑phase trials are never powered for efficacy. Some have suggested that inadequate asundexian dosing in OCEANIC‑AF trial (50 mg daily) explains its inferior efficacy in comparison with apixaban. Asundexian dose selection was based on the results of phase II dose‑finding studies and the results of laboratory testing demonstrating 91% and 92% inhibition of FXIa activity at trough levels in phase II trials PACIFIC‑AMI (Study to Gather Information About the Proper Dosing and Safety of the Oral FXIa Inhibitor BAY 2433334 in Patients Following an Acute Heart Attack) and PACIFIC‑AF (Study to Gather Information About the Proper Dosing of the Oral FXIa Inhibitor BAY 2433334 and to Compare the Safety of the Study Drug to Apixaban, a Non‑vitamin K Oral Anticoagulant [NOAC] in Patients with Irregular Heartbeat [Atrial Fibrillation] that Can Lead to Heart‑related Complication), respectively.^25,26 In the phase II AZALEA‑TIMI 71 trial, abelacimab 90 mg and 150 mg achieved 97% and 99% inhibition of FXI activity, respectively, suggesting that even higher levels of FXI inhibition do not compromise safety. However, the 2 drugs work through distinct mechanisms, and it is not possible to directly compare the assays, one of which measured FXIa activity, and the other which measured FXI activity using a novel assay with a kaolin trigger and fluorogenic substrate readout.

In summary, the AZALEA‑TIMI 71 and OCEANIC‑AF trials were both performed in patients with AF but they tested different drugs, dose selection was based on different laboratory assays, the comparator drugs were different, and perhaps most importantly, AZALEA‑TIMI 71 was a phase II safety trial that was not powered for efficacy, whereas OCEANIC‑AF was a phase III trial powered for efficacy.

Asundexian for prevention of recurrent stroke (OCEANIC‑STROKE)

The Independent Data Monitoring Committee which recommended stopping the asundexian trial for AF recommended continuation of the phase III OCEANIC‑STROKE trial. OCEANIC‑STROKE is recruiting 9300 patients to examine efficacy of asundexian 50 mg daily compared with placebo (on a background of standard antiplatelet therapy) for secondary prevention after noncardioembolic stroke or high‑risk transient ischemic attack (TIA).

Milvexian for stroke prevention in atrial fibrillation, prevention of recurrent ischemic stroke, and prevention of recurrent cardiac events (LIBREXIA program)

Three ongoing phase III trials, LIBREXIA‑AF, LIBREXIA‑STROKE, and LIBREXIA‑ACS, are examining the efficacy of another oral small‑molecule drug, milvexian.

LIBREXIA‑AF, which compares milvexian to apixaban for stroke prevention in AF, will be of particular interest, as an indirect comparator between milvexian (at a higher dose of 100 mg twice daily) and asundexian, given the termination of the OCEANIC‑AF, where asundexian was tested at a dose of 50 mg daily.

LIBREXIA‑STROKE compares milvexian 25 mg twice daily with placebo on a background of standard antiplatelet therapy in patients post noncardioembolic stoke or high‑risk TIA. Notably, the dose being tested in this trial is lower than the 100 mg twice daily dose being studied in the LIBREXIA‑AF; however, patients will also be receiving concurrent antiplatelet therapy, which may allow them to benefit from lower doses of anticoagulant in the context of DPI.

LIBREXIA‑ACS is comparing milvexian 25 mg twice daily against placebo on a background of standard antiplatelet therapy in patients with recent ACS (within 7 days) undergoing medical management or percutaneous coronary intervention. As with the LIBREXIA‑STROKE trial, this study uses a lower dose of milvexian in the context of concomitant antiplatelet therapy. While there was no signal for benefit of asundexian in the phase II PACIFIC‑MI trial, if milvexian proves effective in the phase III LIBREXIA‑ACS trial, it may provide a lower bleeding risk DPI therapy alternative in this population who remain at high risk for recurrent cardiovascular events despite DPI.

Abelacimab for stroke prevention in atrial fibrillation

LILAC‑TIMI 76 trial is comparing abelacimab 150 mg monthly to placebo in patients with AF deemed unsuitable for oral anticoagulation or left atrial appendage closure.

Patients with AF at high bleeding risk may not be able to tolerate currently available anticoagulants at doses tested in the pivotal phase III trials for optimal stroke prevention. Elderly individuals form an increasingly large and important population of patients with AF at high bleeding risk who are often also at elevated stroke risk due to shared risk factors.³³ ELDERCARE‑AF (Edoxaban Low‑Dose for Elder Care Atrial Fibrillation Patients) trial demonstrated that reduced‑dose edoxaban (15 mg daily) in elderly patients deemed inappropriate for anticoagulation at standard doses approved for stroke prevention in AF was effective for stroke prevention, but still numerically increased major bleeding compared with placebo (3.3% vs 1.8%; HR, 1.87; 95% CI, 0.9–3.89; P = 0.09).³⁴ If abelacimab is effective for stroke prevention and associated with an acceptable bleeding profile in patents with AF not suitable for treatment with existing therapies, it would address an important therapeutic gap.

Conclusions

There is cautious optimism with respect to the potential of FXI agents to provide effective anticoagulation without increasing bleeding. Ongoing studies will establish whether FXI agents could replace VKAs for indications in which DOACs are less effective or contraindicated, or whether they could be safely used in patients at the highest risk for bleeding. However, many outstanding questions remain (Table 5), which await the results of ongoing randomized trials.