Based on their chemical structure, several categories of FXI inhibitors have been developed up to now, including monoclonal antibodies, antisense oligonucleotides (ASO), small molecules, aptamers or natural inhibitors, and some of them have reached phase II and phase III of clinical development (Table 1).

Regarding monoclonal antibodies, the first phase I and phase II trials were conducted with abelacimab (MAA868). Abelacimab is a fully human monoclonal antibody that binds to the catalytic domain of FXI, blocking it in the conformation of an inactive zymogen [28, 29].

Abelacimab, which has an intravenous or subcutaneous administration route, has a fast onset of action (peak plasma concentration of 1.75–2 h) and a relatively long half-life (25–30 days), allowing monthly administration. Furthermore, it should be noted that monoclonal antibodies such as abelacimab lack renal clearance and are not removed by dialysis, which differentiates them from small molecules [10, 28].

In a first-in-human phase I study, single subcutaneous administration of abelacimab up to the 240 mg dose was found to be safe and well-tolerated leading to dose dependent reduction in free FXI levels and prolongation of aPTT with abelacimab doses 150 mg and higher [29]. A phase I study of single ascending doses of intravenous abelacimab (30–150 mg) or placebo in healthy volunteers and healthy obese people has been recently completed (ANT-003) [29].

In the multiple ascending doses phase IIa ANT-004 trial (NCT04213807), abelacimab (120 or 180 mg) or placebo was subcutaneously administered once monthly for 3 months in patients with NVAF or flutter at low cardioembolic risk (CHA2DS2-VASc score 0–1 in male and 1–2 in female) in whom the use of an anticoagulant for stroke prevention was not suitable, according to the investigator’s judgment. Twenty-eight participants were screened; 5 participants did not meet inclusion/exclusion criteria; 5 eligible participants failed randomization. The primary outcome was the number of participants that achieved more than or equal to 50, 80, and 90% FXI inhibition after the third administration (Day 91) of the two different doses of abelacimab. From preliminary results, this monoclonal antibody appears to effectively inhibit more of 50% of FXI activity in the larger part of patients randomized to the 180 mg dose, leading to reductions in free FXI levels associated with rapid and sustained prolongation of the aPTT (Table 2) [29]. Furthermore, the ANT-003 and ANT-004 studies revealed that no patient developed a confirmed, treatment-emergent antidrug antibodies (ADA) response.

The results of these clinical studies contributed to clarify the tolerability, safety, pharmacokinetic and pharmacodynamic of abelacimab, paving the way to the much larger AZALEA-TIMI 71 trial (NCT04755283).

The AZALEA-TIMI 71 is a phase IIb, multicenter, randomized, active-controlled study comparing two blinded doses of abelacimab (90 or 150 mg given by subcutaneous injection once monthly) with rivaroxaban 20 mg daily in patients with NVAF, who are at moderate to high risk for stroke, defined by age ≥ 55 years and CHA2DS2-VASc of ≥4 or a CHA2DS2-VASc of ≥3 in association with concomitant antiplatelet medication use or Creatinine Clearance (CrCl) ≤ 50 ml/min. The study enrolled 1287 patients from 95 centers in United States (U.S.), Canada, Europe and Asia. The median age was 74 years, 44% of patients were female and the median CHA2DS2-VASc score was 5; 92% of patients had already taken anticoagulants and 66% had taken DOACs. Patients were followed up for a median of 1.8 years [31]. In the overall trial cohort, ClCr < 50 ml/min was present in 21% of the participants; in patients with CrCl ≤ 50 ml/min at randomization or during the study, rivaroxaban 15 mg/die was used. The primary outcome was the effect of abelacimab compared to rivaroxaban on the rate of major bleeding or clinically relevant non-major (CRNM) bleeding as defined by International Society on Thrombosis and Haemostasis (ISTH) criteria.

In September 2023, Anthos Therapeutics announced that AZALEA-TIMI 71 study was stopped early due to an “overwhelming reduction” in bleeding with both the 150 and 90 mg doses [30]. Following the preliminary results reported at the American Heart Association-Scientific Sessions 2023 on November 12th in Philadelphia, abelacimab demonstrated excellent anticoagulant activity with a median change in FXI from baseline to 3 months of −97% [interquartile range (IQR) −50 to −99] in the 90 mg group and −99% (IQR −98 to −99) in the 150 mg group. The composite rate of major and CRNM bleeding was 2.7 per 100 patient-years in patients taking abelacimab 150 mg (HR 0.33, 95% CI 0.19–0.55), 1.9 per 100 patient-years in the 90 mg abelacimab group (0.23, 0.13–0.42) and 8.1 per 100 patient-years in the rivaroxaban group (p < 0.001 for both doses of abelacimab as compared to rivaroxaban) with a reduction in the events of 67 and 77% respectively with abelacimab. Gastrointestinal (GI) bleeding occurred at a rate of only 0.1 per 100 patient-years with abelacimab at both doses, vs. 2.1 per 100 patient-years with rivaroxaban (HR 0.07, 95% CI 0.01–0.50 and 0.07, 0.01–0.51, respectively) indicating the almost total prevention of GI bleeding, frequently occurring in patients treated with DOACs and often leading to undertreatment. Major bleeding occurred at a rate of 1.0 per 100 patient-years in the 150 mg abelacimab group (HR 0.26, 95% CI 0.11–0.61) and 0.7 per 100 patient-years in the 90 mg abelacimab group (0.19, 0.07–0.50) vs. 3.7 events per 100 patient-years in the rivaroxaban group (p < 0.001 for both doses of abelacimab as compared to rivaroxaban). Similarly, CRNM bleeding occurred at rates of 1.8 per 100 patient-years (HR 0.39, 95% CI 0.21–0.75) and 1.1 per 100 patient-years (0.25, 0.11–0.54) in abelacimab 150 and 90 mg respectively, vs. 4.6 per 100 patient-years in the rivaroxaban group (p < 0.001 for both doses of abelacimab) (Table 2). Intracranial hemorrhage was uncommon and not significantly different between the groups. As secondary outcomes are concerned, such as the incidence of stroke and systemic embolism, ischemic or hemorrhagic stroke and all-cause death, no statistically significant differences were observed among the three treatment arms, but the trial was underpowered to evaluate an effect of the treatment on stroke and death (Table 2). Anyway, the rates of ischemic stroke were numerically but not significantly higher with abelacimab: 1.1 per 100 patient-years in the 150 mg group and 1.3 per 100 patient-years in the 90 mg group vs. 0.7 per 100 patient-years in the rivaroxaban group (p = 0.42 and 0.28, respectively). Considering a median CHA2DS2-VASc score of 5, one would expect an annual stroke rate of 6–7% without treatment [32], which suggests a comparable efficacy of DOACs and abelacimab in preventing around 70–80% of stroke events. As regards the net clinical outcome (a composite of ischemic stroke, systemic embolism, major or CRNM bleeding and all-cause death), statistically significant differences were found in favor of both abelacimab doses (150 and 90 mg) compared to rivaroxaban: 5.5 per 100 patient-years and 5.6 per 100 patient-years vs. 11.3 per 100 patient-years respectively (p < 0.001 for both comparisons).

In conclusion, the results of this phase II trial indicate that both doses of abelacimab (150 and 90 mg monthly) are superior to rivaroxaban 20 mg daily in reducing bleeding events among patients with NVAF and a high CHA2DS2-VASc score. For this reason, an extension study has started to enable patients to switch from rivaroxaban to abelacimab to benefit from the improved bleeding profile.

In the meantime, a phase III trial with abelacimab in AF patients is ongoing. The LILAC-TIMI 76 (NCT05712200) study is an event-driven, randomized trial aimed to evaluate the efficacy and safety of abelacimab as compared to placebo on the rate of ischemic stroke or systemic embolism and of BARC (Bleeding Academic Research Consortium) major bleeding (3c/5) in AF patients aged 65–74 years and with CHA2DS2-VASc ≥ 5, or age ≥ 75 years and with CHA2DS2-VASc ≥ 4, judged by the responsible physician or by their own decision to be unsuitable for currently available anticoagulation therapy. Patients are randomized to receive abelacimab 150 mg or matching placebo subcutaneously once monthly (Table 2). The researchers aim to enroll approximately 1900 patients from North America, Europe, Latin America, the Middle East, and Asia. This study is planned to end in March 2025.

Based on the available evidence, abelacimab appears to epitomize its promise as a hemostasis-sparing anticoagulant and could represent a paradigm shift in the prevention of stroke and other thrombotic conditions. Indeed, if an anti-FXIa therapeutic strategy will provide effective anticoagulation with a minor effect on hemostasis, we might be able to extend our ability to prevent unwanted thrombotic events to broader patient populations, including those at high risk for both bleeding and thrombosis who, because of their extreme fragility, often remain off anticoagulant treatment [33].

Beside monoloclonal antibodies, four anti-FXI small molecules are in advanced stages of clinical evaluation: asundexian, milvexian, SHR2285 and frunexian (EP-7041). The first three have a reversible effect and are administered orally; frunexian is administered intravenously.

To date, SHR2285 and frunexian have never been tested in patients with AF and non-cardioembolic ischemic stroke, so there are no data on their use in the clinical settings analyzed in the present review.

Regarding the small oral molecules asundexian and milvexian, thanks to their small size, rapid diffusion across the membranes is possible, with a rapid onset and offset of action (peak plasma concentration of 2–4 h, and fairly short half-life of 8–16 h), so as to require once or twice daily administration [10, 12, 13, 18, 34].

Asundexian was investigated in the PACIFIC-AF trial (NCT04218266), a multicenter, randomised, double-blind, phase II trial that compared two doses of asundexian (20–50 mg/day) with a standard regimen of apixaban (5 mg twice day, with dose reductions where necessary). The study was conducted at 93 sites in 14 countries, including 12 European countries, Canada, and Japan [35]. The aim of this study was to find the optimal dose of asundexian also comparing the incidence of bleeding with that of apixaban in patients with NVAF. In this study, 753 patients aged ≥45 years with NVAF, a CHA2DS2-VASc score of at least 2 if male or at least 3 if female, and an increased bleeding risk (history of previous bleeding requiring medical attention within 12 months, estimated glomerular filtration rate of 30–50 ml/min, or current indication for aspirin) were enrolled. Patients with another requirement for chronic anticoagulation, or on antiplatelet therapy (apart from up to 100 mg aspirin) were excluded from the study. Enrolled subjects had a mean age of 73.7 years; 351 (46%) were over 75 years, 309 (41%) were women, 336 (45%) were previously on DOACs; the mean CHA2DS2-VASc score was 4 (with 8.5% with a prior stroke). Overall, only 3% had a prior major bleeding event (slightly unbalanced, with up to 5.5% in the asundexian 50 mg group), and 28% had chronic kidney disease (slightly unbalanced, with only 21% in the asundexian 20 mg group).

After 4 weeks of treatment, asundexian 20 and 50 mg resulted in a 81 and 92% reduction in baseline FXIa at trough concentrations, and 90 and 94% at peak concentrations, respectively. After 12 weeks of therapy, both doses of asundexian were superior to apixaban in the primary safety outcome (major bleeding or CRNM bleeding according to ISTH criteria): three primary composite endpoint events occurred in the asundexian 20 mg arm, one in the asundexian 50 mg arm, and six in the apixaban arm. Overall, there were no episodes of ISTH major bleeding in any treatment arm. Ten patients experienced a CRNM bleeding event and 48 had any bleeding event. In general, bleeding rates were lower than expected during the planning of the study. The ratio of incidence proportions was 0.50 (90% CI 0.14–1.68) for asundexian 20 mg, 0.16 (CI 0.01–0.99) for asundexian 50 mg and 0.33 (0.09–0.97) for pooled asundexian vs. apixaban. The annualized incidence rate per 100 patient-years for the primary endpoint (major bleeding or CRNM bleeding) was 5.47 (90% CI 1.49–11.48) for asundexian 20 mg, not calculable for asundexian 50 mg, 3.61 (1.23–7.00) for asundexian total, 11.10 (4.83–19.45) for apixaban. Interestingly, these incidence rates found in the apixaban arm were similar to those of rivaroxaban in the AZALEA TIMI-71 trial, where the primary endpoint was the same. It is worth considering that this trial was designed as a dose-finding phase II clinical study, and was not powered to test differences in the rates of thrombotic events (Table 2). The rates of adverse events leading to discontinuation of the study drug were also similar in the three treatment arms.

Following the encouraging results of the PACIFIC-AF study, the phase III OCEANIC-AF study (NCT05643573) was started, aiming at evaluating the efficacy and safety (in terms of prevention of ischemic stroke, systemic embolism, and major bleeding) of asundexian (50 mg/day, orally) vs. apixaban (5 mg twice day) in adult patients with NVAF on a larger scale. The study was expected to enroll 18,000 patients aged ≥18 years and NVAF with CHA2DS2-VASc score ≥ 3 if male or ≥4 if female, or CHA2DS2-VASc score of 2 if male or 3 if female and enrichment criteria. This trial was expected to end in August 2025. However, on November 19th 2023, Bayer announced the early interruption of the OCEANIC-AF study: this decision by an independent data monitoring committee was based on ongoing surveillance findings showing that asundexian was inferior to apixaban for the prevention of stroke and systemic embolism in patients with AF (Table 2), marking the first unexpected negative result for this drug [36].

It is important to consider that the dosage chosen for the phase III trial was based on a phase II trial that was not powered to test differences in the rates of thrombotic events. It is therefore possible that the dosage chosen for the OCEANIC-AF trial was too low. Furthermore, when focusing on the early interruption of the OCEANIC-AF study with asundexian due to its low efficacy, as compared to the early interruption of the AZALEA-TIMI 71 trial with abelacimab for its superior safety with possible maintained efficacy, it could be speculated that a more intense and persistent inhibition of FXI occurs with abelacimab, due to its relatively long half-life, while oscillations over time in the extent of FXI inhibition might occur with small molecules, due to their shorter half-life. Certainly, the early termination of OCEANIC AF has raised many doubts and uncertainties for the future of asundexian in the field of atrial fibrillation.

Anyway, full data will further be analyzed and published to explain why asundexian did not meet the desired effect. OCEANIC-AFINA, a third randomized, placebo-controlled, double-blind phase III trial has been planned, in which asundexian should be compared with placebo in AF patients at high risk for stroke or systemic embolism who were deemed ineligible for oral anticoagulant therapy due to bleeding concerns. This trial has not yet started recruiting patients. Nevertheless, the study design will now be re-evaluated in light of the OCEANIC-AF interruption [36].

As for other small molecules, milvexian (BMS-986177/JNJ-70033093) is a direct-acting, high-affinity small molecule acting as inhibitor of FXIa. The LIBREXIA-AF (NCT05757869) is an ongoing phase III trial that evaluates the efficacy and safety of milvexian vs. apixaban in 15,500 patients from U.S, Canada, Australia, Europe, South Africa, with AF and CHA2DS2-VASc ≥ 2. The purpose of this study is to assess if milvexian is at least as effective as apixaban for reducing the risk of the composite stroke and non-central nervous system (CNS) systemic embolism (Table 2). The primary outcome will be the time to first occurrence of the composite endpoint of stroke and non-CNS systemic embolism, with up to 4 years of follow-up. This trial will end in May 2027.

In a recent metanalysis, Galli et al. [37] pooled the results coming from eight phase II randomized controlled trials comparing FXI inhibitors vs. other anticoagulants (enoxaparin or DOACs) or vs. placebo on top of antiplatelet therapy, including a total of 9216 patients. Two of these studies enrolling a total of 935 patients compared the FXI inhibitors with DOACs, and reported that FXI inhibitors were associated with a trend to reduce the endpoint “any bleeding” (RR 0.66, 95% CI 0.31–1.38) without any difference in major bleeding (1.03, 0.22–4.78) or in trial-defined efficacy endpoint (1.23, 0.88–1.70).