Elevated Microclots with Low D-Dimer As an Indicator of Impaired Plasmin Generation in Patients with Viral Infections

Background: D-dimer, a proteolytic fragment of clots, is a commonly used plasma marker of clot formation. Alternatively, assays to directly measure microclots have been recently developed, utilizing the fibrin-binding dye thioflavin T. However, a discrepancy between microclots and D-dimer exists, as plasma samples with high D-dimer (>5-times normal) have few microclots, while those with high microclots have low D-dimer. As D-dimer requires both clot formation and lysis, we hypothesized that differences in fibrinolytic activity explain this discrepancy. Here, we measured ex vivo plasmin generation in well-characterized plasma samples to test this hypothesis and to determine which components of the coagulation and fibrinolytic systems influence plasmin generation.

Methods: All studies were approved by the Institutional Review Boards of the University of Kentucky and Medical College of Wisconsin. Ex vivo plasmin generation was measured in citrated plasma from two independent cohorts of COVID-19 patients, as well as normal pooled plasma and immunodepleted plasmas: (1) Cohort 1 (Kentucky, average age 55 yrs, 52.8% male) consisted of 23 SARS-CoV-2+ inpatients with severe symptoms, 53 SARS-CoV-2+ outpatients with mild symptoms, and 57 SARS-CoV-2- controls with mild symptoms; (2) Cohort 2 (Wisconsin, average age 60 yrs, 41% male) consisted of 13 SARS-CoV-2+ inpatients with samples collected pre- and post-recovery, 7 SARS-CoV-2+ inpatients who did not recover, and 7 SARS-CoV-2- inpatients.

Results: In Cohort 1, plasmin generation was significantly lower in patients than controls, and did not vary between outpatients and inpatients, whether measured in the presence or absence of thrombomodulin. Thrombomodulin inhibited plasmin generation to a greater extent in outpatients than either inpatients or controls. Ex vivo plasmin generation correlated negatively with D-dimer, though samples with very high D-dimer (>10μg/mL) had high plasmin generation. Conversely, plasmin generation correlated positively with microclots, but samples with very high microclots (>5x control) had low plasmin generation.

Plasmin generation positively correlated with plasma tissue factor activity and with the anticoagulants protein C (PC) and free protein S (PS). Multiple lines of evidence indicate that this correlation is independent of PC/PS anticoagulation activity: (1) the effect of thrombomodulin on plasmin generation was independent of PC or PS concentration and was not reduced in normal pooled plasma and plasma immunodepleted of either PC or PS; and (2) there was no correlation between the effect of thrombomodulin on plasmin vs. thrombin generation. Finally, plasmin generation was independent of myeloperoxidase (a marker of neutrophil activation) and von Willebrand Factor, a marker of endothelial activation which can incorporate into a fibrin clot and alter its structure.

Analyses of Cohort 2 confirmed and extended the results from Cohort 1, as measurements of tissue plasminogen activator (tPA) and plasminogen activator inhibitor-1 (PAI-1) were available for Cohort 2. Plasmin generation was negatively related to tPA activity but independent of PAI-1. Finally, the effect of thrombomodulin on plasmin generation was not influenced by either tPA or PAI-1.

Conclusions: Our data support the hypothesis that ex vivo plasmin generation is predominantly regulated by plasminogen concentration. In some patients, elevated tPA activity leads to activation of plasminogen and plasmin-mediated degradation of fibrin clots, resulting in higher D-dimer and lower microclot measurements. However, this process likely also reduces the amount of plasminogen available, yielding decreased ex vivo plasmin generation. Thus, our data suggest that reduced plasmin generation may reflect overactivation of the fibrinolytic system and depletion of its components, akin to disseminated intravascular coagulation.