Thrombin generation measured by two platforms in patients with a bleeding tendency: Comment

Abstract

Dear EditorsinChief, The recent paper by Cornette et al (2021), where patients with undiagnosed bleeding tendencies were investigated with two thrombin generation platforms, is of interest to haematologists because this is a common problem faced in bleeding disorders clinics.2 There is growing recognition and interest in this patient group which has recently been reviewed in detail, and that are rather variably referred to as bleeding of unknown cause (BUC) and unclassified bleeding disorder (UBD).1 For the purposes of this discussion in this manuscript BUC/UBD is defined as a clear bleeding tendency with normal conventional tests of haemostasis and where other medical causes of bleeding (e.g. scurvy) have been excluded. Thrombin generation is an intellectually attractive technique for investigating these patients. Itis a global assessment of haemostasis and reflects not only procoagulant factors but also potentially increased coagulation inhibitor activity such as thrombomodulin (i.e. thrombomodulin coagulopathy) or tissue factor pathway inhibitor (i.e. the factor V short disorders). In this present study the authors found that 62.7 and 69.5% of undiagnosed patients (of which there were 59 in their collection) showed abnormal thrombin generation (in platelet poor plasma) using the ST Genesia (Diagnostica Stago, AsnièressurSeine, France) and calibrated automated thrombogram (Diagnostica Stago) respectively.1 Specifically there were increased lagtime or reduced endogenous thrombin potential. Whilst the authors did not explicitly state the results of coagulation factor testing they followed a previously published algorithm and so we can assume that mild coagulation factor deficiencies would not account for the thrombin generation abnormalities reported. It is useful to put this current study into the context of other publications over the last decade which have evaluated thrombin generation in BUC/UBD in platelet poor plasma. These studies are detailed in Table 1 and the study by Cornette et al (2021) has also been added to the table for comparison.1,3– 8 In total 7 studies are summarised; 4 of these studies had negative findings (either no difference between BUC/UBD and controls or abnormalities seen in <5% BUC/UBD patients3,4,6,8) and 3 of the studies were able to demonstrate substantial abnormalities between BUC/UBD patients and healthy controls or a reference range.1,5,7 The table demonstrates the heterogenous nature of the study methodologies in terms of thrombin generation assays, tissue factor concentrations (which could effect sensitivity to coagulation defects) and patient cohorts (e.g. the range of bleeding scores and bleeding score utility). One consistent and striking finding is the preponderance of women with BUC/UB (>75% cases). Recently Obaji et al (2020) published in abstract form a study of 20 BUC/UBD patients investigated using platelet rich plasma thrombin generation. The authors found median peak thrombin generation of 65.9 (interquartile range 56.9– 87.7) for BUC/UBD vs. 86.2 (68– 121) nM for healthy controls, (p = .03).9 Similarly platelet derived microparticles supported lower thrombin generation than in healthy controls. There was no difference between BUC/UBD patients and healthy volunteers in terms of externalised phosphatidylserine. One of the severe limitations of trying to compare these studies is that there are no standardised criteria for including BUC/ UBD. It is likely that these studies have a heterogeneous mix of patients included in, many of which may actually not have a clinically significant bleeding tendency (i.e. have had several haemostatic challenges without bleeding) as many of the studies (detailed information on the bleeding score is in Table 1) included patients with nonsignificant bleeding scores, which usually would not suggest a bleeding tendency. It is premature to add thrombin generation to the routine clinical diagnostic repertoire. Lack of reproducibility of findings across studies including diagnostic criteria continues. Interpretation of abnormal thrombin generation findings in patients with BUC/UBD is hampered by an absence of a recognised standardisation of the technique. Thrombin Generation remains restricted to specialist centres and proven cost effectiveness remains to be shown. Further research in large multicentre cohorts is required with a standardised research protocol. Despite these limitations, the results presented here by Cornette et al (2021) raise the intriguing question as to the cause of the thrombin generation abnormalities seen. Recently, coagulation inhibitors (including tissue factor pathway inhibitor) have been found to be increased in BUC/UBD and their impact on thrombin generation parameters is reviewed in detail.10 For the sake of patient care and future research standardisation is required for both clinical diagnosis of BUC/UBD and also standardised data sets (e.g. for inclusion criteria, clinical features and minimum conventional laboratory tests) for research projects. Received: 30 July 2021 | Accepted: 6 August 2021 DOI: 10.1111/jth.15524