First generation Bruton's tyrosine kinase (Btk) inhibitors were associated with both decreased thrombosis and increased bleeding. However, newer more selective Btk inhibitors may not cause bleeding. In this issue of Blood Advances, Smith et al1 assessed the potential of these antagonists as antithrombotics in thromboinflammation.
Following the discovery in 1993 that Btk is critical for B-cell function, small molecules targeting this TEC family kinase were developed for treatment of B-cell malignancies.2,3 Ibrutinib was the first to be approved.4,5 Although it was effective, ibrutinib was associated with bleeding. Because patients who lack Btk (as in X-linked agammaglobulinemia [XLA]) do not have a bleeding diathesis, the bleeding diathesis was thought to be secondary to off-target effects, the most important of which was found to be inhibition of Src.6 Continued refinement of Btk inhibitors led to the development of rilzabrutanib (PRN1008) and atuzabrutinib (PRN473). In phase 2 clinical trials in patients with previously treated immune thrombocytopenic purpura (ITP), PRN1008 did not appear to be associated with bleeding.7 Importantly, several immunoreceptor tyrosine–based activation motif (ITAM) receptors including CLEC-2, glycoprotein VI (GPVI), and FcgRIIA signal through Btk in platelets and can contribute to thrombus formation.8 Preclinical and clinical studies have indicated that inhibition of ITAM receptors such as GPVI can inhibit thrombus formation without promoting bleeding. These observations raise the question of whether PRN1008 and PRN473 spare Src and whether they can inhibit thrombosis without eliciting bleeding.
Smith et al now address these issues by carefully evaluating the specificity of PRN1008 and PRN473 for Btk and determining whether they inhibit thrombus formation in vivo. Both inhibitors blocked signaling through CLEC-2 or GPVI. Downstream signaling through phospholipase Cγ2 was blocked, indicating on-target effectiveness, yet upstream signaling targets were not affected, providing some indication of selectively. Importantly, no effect on Src phosphorylation was observed. Btk also failed to inhibit activation stimulated by thrombin or U46619, indicating a lack of effect on G protein coupled receptors such as protease-activated receptors (PARs) or the TP receptor, respectively. The authors also tested PRN1008 in platelets from patients with XLA, which provided a targetless mutant well-suited for testing specificity. XLA platelets showed reduced responses to 10 μg/μL collagen and were unaffected by PRN1008, consistent with the premise that it is specific. However, Btk did affect the response of XLA platelets to 3 μg/μL collagen, raising the possibility of some off-target effect.
PRN1008 and PRN473 were tested in both in vitro and in vivo models of thrombus formation. Compounds decreased adhesion and aggregate size when added to whole blood flowed over the CLEC-2 ligand podoplanin at venous shear or over collagen at arterial shear. PRN1008 and PRN473 also inhibited mouse platelets, enabling studies of thromboinflammation in murine models. The authors had previously shown that CLEC-2 functions in Salmonella infection–driven liver thrombosis9 and in an inferior vena cava model of deep vein thrombosis.10 They evaluated the effect of dietary administration of PRN473 in the Salmonella–induced thrombus formation model. Mice were fed 6.66 g/kg PRN473 for 7 days. With this regimen, spleens showed 100% kinase occupancy by the inhibitor. In platelets purified from these mice, signaling through platelet CLEC-2 and GPVI were inhibited in ex vivo studies, confirming efficacy of the dosing regimen. Furthermore, platelet responses to a PAR4 agonist were intact, demonstrating PRN473 selectively. Reduced platelet adhesion when blood was flowed over collagen or fibrinogen was also observed. Seven days following intraperitoneal injection of attenuated Salmonella, livers were removed and evaluated for thrombosis. Although there was no effect of PRN473 on thrombus area, there was a trend toward decreased thrombus number in the treated group and a significant reduction in thrombi forming in podoplanin-expressing vessels. Similarly, there was a trend toward reduced platelet consumption. For testing in the inferior vena cava stenosis model, mice were treated with PRN473 via oral gavage (80 mg/kg) once daily. This regimen yielded an occupancy of 60.5% and an average plasma drug level of 14.5 ng/mL. Again, platelets showed reduced responses to CLEC-2- and GPVI–mediated platelet responses, but normal activation in response to a PAR4 agonist. In this model, the prevalence of thrombi in control mice was nearly 60%, whereas that in PRN473-treated mice was ∼20%. Furthermore, thrombi were smaller in mice receiving PRN473. Of note, plasma concentrations of drug varied considerably and thrombi only formed in mice whose plasma concentrations of PRN473 were <8.5 ng/mL.
These studies underscore the possibility of Btk as a target in the context of thrombinflammation. A large part of this appeal is the possibility of inhibiting thrombosis without promoting bleeding. Genetic deficiency of Btk is not associated with bleeding. Clinical trials of PRN1008 in patients with ITP have not demonstrated bleeding complications even though these patients are prone to bleed. Targeting Btk enables inhibition of signaling mediated via CLEC-2 and GPVI without directly affecting the functions of these receptors in adhesion. Limitations of the study include the fact that while PRN1008 is the compound found to be free of bleeding complications in clinical studies of ITP, PRN473 is the compound evaluated in the thromboinflammatory models. Nonetheless, the fact that both PRN1008 and PRN473 have similar selectivity profiles and in particular, do not inhibit Src, enhance the likelihood that they will behave similarly with regard to thrombotic and hemorrhagic phenotypes. Furthermore, several lines of evidence support the contention that GPVI and CLEC-2 pathways could be targeted for inhibition of thrombosis without promoting bleeding.8 These Btk antagonists could represent an orally available option for targeting these pathways.
Conflict-of-interest disclosure: R.F. is founder of and has equity in PlateletDiagnostics; is a consultant for Xap Therapeutics and Tessera Therapeutics; and is on the scientific advisory board of Function Therapeutics and Porosome Therapeutics.
