Blood group o and a lower endogenous thrombin potential are associated with increased risk for future bleeding after hemostatic challenges in bleeding disorder of unknown cause

Abstract Background. Bleeding disorder of unknown cause (BDUC) is diagnosed in patients who present with clinically significant bleeding symptoms despite normal results on detailed hemostatic testing. Although a substantial proportion of these patients report bleeding complications following hemostatic challenges prior to diagnosis, prospective data on risk factors for surgery- or tooth-extraction-related bleeding in BDUC patients remain scarce. The aim of this study was to identify routine laboratory parameters and experimental coagulation assays associated with bleeding complications following medical interventions in patients with BDUC. Methods. A total of 105/443 (24%) BDUC patients from the Vienna Bleeding Biobank, who had at least one surgery and/or tooth extraction during prospective follow-up, were included in this analysis. In this study, we investigated the association between prolonged bleeding or bleeding-related complications after surgery or dental extractions and relevant coagulation-related laboratory parameters (hemoglobin, platelet count, fibrinogen, von Willebrand factor antigen and activity, C-reactive protein, thrombin and plasmin generation assays (all parameters) and native rotational thromboelastometry (clotting time, clot formation time, mean clot firmness, maximum lysis and area under the curve, AUC), measured at study inclusion, by calculating univariate logistic regression models adjusted for age, sex, blood group O and the use of bleeding prophylaxis, such as desmopressin or tranexamic acid. Results. Of the 105 patients, 64 had surgery, 22 had tooth extractions and 19 underwent both procedures during the follow-up period (median 4.3 years, interquartile range 2.6-6.7 years). Hemostatic prophylaxis was used in 13 surgeries (16%) and in 2 tooth extractions (5%). Among all these medical procedures, bleeding complications were reported in 22/83 surgeries (27%) and in 10/41 (24%) tooth extractions. In patients who received hemostatic prophylaxis for their surgery, only 1/13 reported bleeding (7%), while bleeding was non-significantly more common in patients without prophylaxis, among whom 21/70 (30%) had bleeding, pfisher=0.168. Furthermore, hemostatic prophylaxis did not reduce the rate of bleeding complications after tooth extractions, where 2/3 patients (67%) had bleeding despite prophylaxis compared to 8/38 (21%) patients who bled without prophylaxis, pfisher=0.142. None of the aforementioned routine laboratory parameters were associated with the occurrence of any bleeding event during the follow-up period. However, blood group O was consistently associated with a higher risk of bleeding after hemostatic challenges, even after the adjustment for bleeding prophylaxis and the von Willebrand factor (adjusted odds ratio 3.18, 95% CI 1.17-9.82). Despite the limited sample size, thrombin generation assay results, available in 90/105 BDUC patients and in 27/30 patients with bleeding events, demonstrated a statistically significant non-linear association between reduced endogenous thrombin potential (represented by the AUC) and bleeding risk after surgery or dental extractions. Specifically, a 50% decrease in thrombin-generation AUC was independently associated with a markedly increased risk of bleeding after any hemostatic challenge (adjusted odds ratio 5.26, 95% CI 1.25–27.0), based on a logistic regression model using log₂-transformed AUC values. However, none of the parameters of the plasmin generation assay or the rotational thromboelastometry were associated with bleeding after hemostatic challenges. Conclusions. Stratifying the individual risk of bleeding after medical interventions, such as surgery or dental extractions, in BDUC patients has remained a great challenge for researchers and physicians. In this study, we have identified blood group O as the only parameter routinely assessed during hemostatic investigations associated with the occurrence of these bleeding complications. Importantly, we could show an association between the endogenous thrombin generation potential, measured using a commercially available thrombin generation assay, and bleeding after hemostatic challenges in patients after the BDUC diagnosis. Lastly, our data highlights the importance of hemostatic prophylaxis in patients with clinically relevant bleeding symptoms within a personalized treatment approach, regardless of whether they received an established diagnosis of a bleeding disorder or were classified as having BDUC.