AbstractBleeding disorder of unknown cause (BDUC) is a diagnosis of exclusion after evaluation of plasma coagulation and platelet function. Patients with BDUC (n = 375) recorded in the Vienna Bleeding Biobank were analyzed in comparison with healthy controls (HCs; n = 100) in this case-control study. Plasmin generation (PG) parameters were analyzed using calibrated fluorescence detection in citrated plasma. Turbidimetric plasma clot formation/lysis of 293 (78%) patients with BDUC and confocal microscopy of clots from representative patients with BDUC (n = 6) and HCs (n = 9) were assessed. In the PG analysis, patients with BDUC exhibited lower velocity and peak plasmin levels but a higher endogenous plasmin potential than HCs. Peak plasmin levels correlated with maximum clot absorbance but not with clot lysis time. Clot absorbance is an indicator of clot fiber density. Confocal microscopy analysis revealed a tendency towards thicker fibers in clots of patients with BDUC, which negatively correlated with peak plasmin (r = −0.561; P = .030). Peak plasmin correlated weakly with factor XIII, but not with other fibrinolytic factors (alpha2-antiplasmin, thrombin activatable fibrinolysis inhibitor, or plasminogen activator inhibitor 1) or bleeding severity. A model comprising fibrinogen and parameters of PG yielded high predictive power in discriminating between patients with BDUC and HCs across a fivefold stratified cross validation (80% of data; mean area under the curve [AUC], 0.847). The model generalized well to unseen data (20% of data; AUC, 0.856). Overall, patients with BDUC counterintuitively exhibited reduced peak plasmin levels, potentially related to altered clot structure.
Read full abstract