Not in the beginning Free

In this issue of Blood, Sang et al use complementary, independent approaches to demonstrate that plasmin(ogen) has no effect on thrombogenesis.¹ 

Thrombogenesis in venous thromboembolism (VTE) is a complex pathology in which an activated coagulation cascade and immune/inflammatory processes collide. Resolution of the resulting fibrin-rich thrombus involves the fibrinolytic system, immune cells, and vascular remodeling processes. Treating VTE with anticoagulation inhibits further thrombogenesis while fibrinolysis proceeds. Although thrombogenesis and fibrinolysis progress at contrasting time scales (hours/days vs weeks/months, respectively), the role, if any, of the fibrinolytic system in thrombogenesis has been a long-standing question.

In the fibrinolytic system, plasminogen is activated by proteolysis to generate plasmin, the primary enzyme that degrades fibrin. Genetic ablation of plasminogen in mice resulted in fibrin deposition in multiple organs and a markedly reduced capacity to lyse experimental emboli.² Conversely, deficiency of plasminogen activator inhibitor-1 (ie, functionally enhanced plasmin activity) elevates bleeding risks in patients³ and results in reduced thrombogenesis as well as accelerated thrombus resolution in mice.⁴ Together, these observations pose the question of whether plasmin may partly regulate thrombogenesis, giving pause to the use of the plasmin inhibitor, tranexamic acid, in managing bleeding diatheses due to safety implications.

Uncertainty about elevated thrombotic risks has persisted due to conflicting interpretations of the results from clinical trials of tranexamic acid use in different bleeding conditions. The HALT-IT trial found no survival benefit and a significant risk for VTE associated with the use of tranexamic acid in patients with acute gastrointestinal bleeding.⁵ However, the POISE-3 trial met the primary efficacy outcome of reduced bleeding and found no difference in VTE risks between tranexamic acid and placebo in patients who underwent general or nongeneral (eg, arthroplasty) surgery.⁶ The WOMAN-2 trial found no clinical improvement and no significant risk for VTE with tranexamic acid compared with placebo in managing postpartum hemorrhage.⁷ Multiple reasons including differences in study populations and design, comorbidities, bleeding pathologies, etc may be involved in the discordant conclusions on VTE risk with fibrinolytic inhibition by tranexamic acid. But without further study, the reasons for the discordance remain speculative.

Seeking to clarify the relationship between inhibition of fibrinolysis and VTE, Sang et al first revisit the question on the role of plasmin(ogen) in thrombogenesis via population studies.¹ Reanalysis of an international cohort of inherited plasminogen deficiency, Hypoplasminogenemia: An International Retrospective and Prospective Cohort Study (HISTORY), found no reported history of VTE. Using genomes from the International Network Against Venous Thrombosis (INVENT) consortia (a combined cross-ancestry genome-wide association study on VTE), variant analyses on established VTE-associated genes and candidate fibrinolysis-related genes indicated that plasminogen was not associated with VTE. Proteomic analysis of a large biorepository (UK Biobank) corroborated the genomic findings: plasminogen levels failed to predict VTE incident.

Sang et al validated their association studies in a murine model of VTE in which the inferior vena cava is completely stenosed to cause stasis, a thrombotic risk factor in Virchow’s triad. Thrombi developed no differently between wild-type and plasminogen knockout mice. Consistent with genetic deletion of plasminogen, plasmin inhibition with tranexamic acid in this mouse model of VTE resulted in thrombi with morphologies, masses, and content comparable to placebo. Altogether, the population studies and murine models confirm a lack of a role for plasminogen in the initial phases of VTE in uncomplicated settings.

How closely pharmacologic inhibition of plasmin phenocopies genetic deficiency of plasminogen in more complicated settings associated with major bleeding (eg, trauma, intracranial hemorrhage, postpartum hemorrhage) remains an open question. VTE complications with tranexamic acid may reflect different, yet-to-be-discovered pathologies that coincide with hemorrhage. A retrospective study of patients with combat casualties who require massive transfusions (defined as ≥10 units of packed red blood cells within 24 hours) found tranexamic acid and injury severity to associate with VTE, though independence between these 2 variables is unclear.⁸ Expanding mechanistic studies to include more complex models may help reveal which settings plasminogen would have a role in thrombogenesis and when plasmin inhibition with tranexamic acid may pose a VTE risk.

Nonetheless, Sang et al’s findings viewed in the context of the reported association between tranexamic acid and VTE pose another question: does tranexamic acid do anything else besides inhibit plasmin? Tranexamic acid was recently shown to bind near the active site of the endonuclease domain in open reading frame 2 protein (ORF2p) of the retrotransposon, long interspersed nuclear element-1 (LINE-1), with unknown consequences.⁹ Intriguingly, inhibition of the reverse transcriptase domain in ORF2p of LINE-1 increased platelet activation.¹⁰ Whether an interaction between these 2 domains in ORF2p has an impact on VTE and more generally, hemostasis, remains to be determined.

As thrombi may vary in structure and composition, future studies outlining how different pathological contexts affect VTE initiation, progression, and resolution may help distinguish bleeding conditions that may improve with plasmin inhibition vs complex situations in which tranexamic acid may bring on thrombotic complications. Toward this goal, Sang et al provide a foundation upon which mechanisms of thrombogenesis and thrombus resolution may be evaluated to measure VTE risks in different conditions.

Conflict-of-interest disclosure: A.Y. declares no competing financial interests.