Asundexian FXIa inhibitor: SAR and selectivity analysis

Why Factor XIa Is the New Frontier in Anticoagulation

Factor XIa (FXIa) sits at a pivotal junction in the intrinsic coagulation cascade, where it amplifies thrombus propagation while contributing relatively little to physiological haemostasis—the primary driver of which is the extrinsic (tissue factor) pathway. This mechanistic asymmetry is the scientific rationale behind Bayer’s decision to target FXIa with asundexian (BAY 2433334): selective inhibition of this enzyme is hypothesised to uncouple antithrombotic efficacy from bleeding risk, the defining limitation of all approved direct oral anticoagulants (DOACs).

The clinical significance of this target choice is substantial. Current DOACs—Factor Xa inhibitors such as rivaroxaban and apixaban, and the thrombin inhibitor dabigatran—are associated with dose-dependent bleeding risk that limits their use in patients with atrial fibrillation, acute coronary syndrome, or recent stroke who are already considered high-risk. According to WHO data, cardiovascular disease remains the leading cause of mortality globally, and safer anticoagulation represents one of the most consequential unmet needs in medicine. By positioning FXIa as a target that is more dispensable to normal haemostasis than FXa or thrombin, Bayer’s programme—and the broader field validated by organisations such as WIPO-tracked patent filings—represents a strategic inflection point in anticoagulant drug discovery.

The hypothesis that FXIa inhibition can be antithrombotic without increasing bleeding has now been tested in preclinical rabbit models. Asundexian demonstrated efficacy in reducing thrombus weight in both arterial (ferric chloride injury) and venous (arterio-venous shunt) models, without increasing bleeding times or blood loss—even when co-administered with antiplatelet agents aspirin and ticagrelor. This preclinical dissociation of antithrombotic effect from haemostatic impairment is the central value proposition of the entire programme.

The Substituted Oxopyridine Scaffold: Structural Identity of the Lead Series

The lead chemical series of asundexian is defined in Bayer’s core patent as substituted oxopyridine derivatives—a heterocyclic scaffold centred on a pyridine ring bearing a ketone (oxo) substituent, with additional substitution patterns optimised for FXIa active-site engagement. This scaffold identity is the foundational IP claim that distinguishes asundexian’s chemical series from all other anticoagulant programmes.

Structure-based drug design (SBDD) is the methodological framework through which the oxopyridine scaffold was optimised. SBDD leverages three-dimensional structural information about the target enzyme’s active site to guide iterative chemical modifications, enabling medicinal chemists to maximise binding interactions while simultaneously minimising off-target engagement. For serine proteases like FXIa, the active site contains a catalytic triad (Ser, His, Asp) and a specificity-determining S1 pocket whose geometry differs meaningfully from those of FXa and thrombin—differences that the oxopyridine scaffold exploits to achieve exceptional selectivity.

The iterative optimisation process involves cycles of chemical synthesis, enzymatic assay (measuring Ki and IC50 against FXIa and selectivity panels), and pharmacokinetic profiling. Analytical techniques including high-performance liquid chromatography (HPLC) and mass spectrometry (MS) are standard tools for compound characterisation and metabolic stability assessment throughout this workflow.

Patent-Disclosed SAR and Selectivity Data Across the Serine Protease Panel

The patent-disclosed selectivity profile of asundexian is among the most striking in the anticoagulant literature: a Ki of 0.26 nmol/L for human FXIa against Ki values exceeding 10,000 nmol/L for Factor Xa, thrombin, Factor IXa, Factor VIIa/TF, tPA, uPA, and APC. This represents a minimum selectivity ratio of greater than 38,000-fold over the two most clinically relevant off-targets—FXa and thrombin—and confirms that the oxopyridine scaffold achieves near-complete functional orthogonality to these enzymes.

“Asundexian’s Ki values for Factor Xa and thrombin both exceed 10,000 nmol/L—a greater than 38,000-fold selectivity margin over its primary target FXIa—placing it in a distinct mechanistic class from all approved direct oral anticoagulants.”

The SAR implications of this selectivity data are significant for medicinal chemists working in the serine protease inhibitor space. The oxopyridine scaffold achieves near-absolute selectivity over FXa and thrombin while retaining meaningful (though far weaker) activity against plasma kallikrein (Ki 110 nM) and trypsin (Ki 1,100 nM). This selectivity gradient—spanning four orders of magnitude from FXIa to FXa—reflects the scaffold’s exploitation of structural differences in the S1 and adjacent subsites of these enzymes. The activated partial thromboplastin time (aPTT) assay, which measures the activity of the intrinsic and common coagulation pathways, is used as a functional readout of FXIa inhibition and serves as a key pharmacodynamic biomarker in both preclinical and clinical settings.

Oral Bioavailability and Preclinical Pharmacology

Asundexian is described as orally bioavailable in the patent literature, a property that is essential for its intended clinical use as a long-term antithrombotic agent. The path to oral bioavailability in the oxopyridine series involved iterative pharmacokinetic optimisation—cycles of chemical synthesis followed by testing for oral absorption, metabolic stability, and systemic exposure—before asundexian was selected as the clinical candidate.

A critical species-specificity finding disclosed in the preclinical programme is that asundexian is not effective in rodents. This limitation necessitates the use of alternative animal models—principally rabbits—for preclinical evaluation of both efficacy and safety. The preclinical model battery employed includes the ferric chloride (FeCl₂) carotid and mesenteric artery injury models, the arterio-venous (AV) shunt model for venous thrombosis, and dedicated bleeding models using ear, gum, and liver injury endpoints. The dissociation of antithrombotic efficacy from haemostatic impairment observed across these models in rabbits is the primary preclinical evidence supporting the therapeutic hypothesis.

The oral bioavailability optimisation of FXIa inhibitor series is a recognised challenge in the field, as highlighted by other programmes that have also focused on achieving moderate to good oral bioavailability for this target class. Computational approaches including molecular docking (FlexX), CoMFA, and HQSAR have been applied to related anticoagulant series to predict binding modes, activity, and oral bioavailability parameters, and represent tools that are consistent with the SBDD methodology used for asundexian’s development. Research published through bodies such as NIH and indexed in Nature journals has extensively characterised the structural biology of FXIa that underpins these computational approaches.

Scaffold Differentiation from Approved DOAC Chemical Series

The structural differentiation of asundexian from approved DOACs operates at two levels: target selectivity and chemical scaffold identity. At the target level, all approved DOACs inhibit either Factor Xa (rivaroxaban, apixaban, edoxaban) or thrombin (dabigatran)—both downstream in the coagulation cascade relative to FXIa. Asundexian’s upstream positioning at FXIa is therefore a fundamentally different mechanistic strategy, not merely a chemical variation on existing themes.

At the chemical scaffold level, the distinction is equally clear. Rivaroxaban—also developed by Bayer, making this an instructive intra-company comparison—is built on an oxazolidinone core, a five-membered heterocyclic ring containing both oxygen and nitrogen. Asundexian’s oxopyridine scaffold is a six-membered pyridine ring with an oxo substituent—a fundamentally different ring system with different three-dimensional geometry, hydrogen-bonding capacity, and electronic character. These differences translate directly into the divergent selectivity profiles: the oxazolidinone series is optimised for the FXa active site, while the oxopyridine series is optimised for the FXIa active site.

Within the FXIa inhibitor landscape itself, other experimental programmes have explored pyridazine and pyridazinone scaffolds—six-membered rings containing two adjacent nitrogen atoms. These are also chemically distinct from the oxopyridine series, reflecting the multiple structural approaches being pursued to address the FXIa active site. The patent landscape tracked by EPO confirms that Bayer’s oxopyridine series represents a differentiated IP position within the broader FXIa inhibitor patent space. For R&D teams conducting freedom-to-operate or competitive landscape analyses, this scaffold differentiation has direct implications for IP strategy and compound design.

For drug development scientists, the validation of the substituted oxopyridine scaffold as a viable path to sub-nanomolar FXIa inhibition with exceptional selectivity and oral bioavailability provides a concrete benchmark for future anticoagulant design. The development of asundexian underscores that pathway-selective inhibition—targeting FXIa rather than the downstream FXa or thrombin—can achieve the long-sought goal of separating antithrombotic efficacy from haemostatic impairment, and that the chemical series enabling this separation is now well-characterised in the patent record.