New approaches for measuring coagulation

Haemophilia (2006), 12, (Suppl. 3), 76 81 New approaches for measuring coagulation T. W. BARROWCLIFFE, M. CATTANEO,* G. M. PODDA,* P. BUCCIARELLI, F. LUSSANA,* A. LECCHI, C. H. TOH, à H. C. HEMKER, S. BÉGUIN, J. INGERSLEV and B. SØRENSEN *Unità di Ematologia e Trombosi, Ospedale San Paolo, Università di Milano, Italy; A. Bianchi Bonomi Hemophilia and Thrombosis Center IRCCS Ospedale Maggiore, Milano, Italy; à Royal Liverpool University Hospital, Liverpool, UK; Cardiovascular Research Institute Maastricht (CARIM), Universiteit van Maastricht, Maastricht, The Netherlands; and University Hospital Skejby, Aarhus, Denmark Summary. Although specific assays of coagulation factors are essential for diagnostic purposes they only give partial information about an individual s haemostatic state. This can be better assessed by various global tests, and recent developments and evaluations of five such tests are described in this symposium: the PFA-100; waveform analysis; thrombin generation; overall haemostasis potential; thrombelastography. Each test has advantages in various applications, but the thrombin generation test and waveform analysis have been found most useful in haemophilia, whilst the PFA-100 is helpful in von Willebrand s disease. Keywords: coagulation, global tests, bleeding disorders Introduction As knowledge of the coagulation system developed in the 1950s onwards, specific assays were devised for most of the main clotting factors, and these largely replaced most of the non-specific global tests such as thrombin generation and thrombelastography. However, in recent years there has been a recognition that specific assays may not be sufficiently informative to assess the patient s overall haemostatic state and that more global tests may contribute useful additional information. Another reason for the revival of global assays is the development of technical innovations and more sophisticated methods of analysis, which have improved their reliability and reproducibility and led to more widespread use. In this symposium, the clinical utility of three such global assays is described. Dr Marco Cattaneo assesses the PFA-100 as an in vitro substitute for the bleeding time; Dr H. Coenraad Hemker describes the use of thrombin potential, as determined by the thrombin generation test, in haemophilia; Dr Cheng Hock Toh discusses the application of waveform analysis in bleeding disorders. Finally, Dr Jorgen Ingerslev compares the use of waveform, thrombin Correspondence: Dr Trevor W. Barrowcliffe, 35 Oakhurst Avenue, East Barnet, Herts EN4 8DL, UK. Tel.: +44 208 368 2035; e-mail: twbarrowcliffe@yahoo.co.uk generation and traditional bioassays in haemophilia patients. The PFA-100Ò versus the bleeding time in the diagnosis of bleeding disorders We compared the performance of PFA-100Ò closure time (CT) and skin bleeding time (BT) in the diagnostic work-up of 128 consecutive patients referred to our centre to be screened for bleeding disorders. The correlation with the severity of the bleeding symptoms and the sensitivity of BT and PFA-100 CT for known defects of primary haemostasis were evaluated. All patients underwent (i) a careful medical interview and were assigned a Ôbleeding scoreõ, based on the number, type, frequency and severity of bleeding symptoms; (ii) firstline screening [prothrombin time, activated partial thromboplastin time (APTT), PFA-100 CT both the collagen-adp (C-ADP) and the collagen-epinephrine (C-EPI), cartridges and BT]. The search for von Willebrand s disease (VWD), platelet function disorders (PFD), clotting factor defects and abnormalities of fibrinolysis was performed according to the results of the first-line screening tests and the severity and type of bleeding history. Seven (6%) patients had type 1 VWD, 12 (9%) PFD, 29 (23%) defects of clotting factors, 18 (14%) prolongations of the APTT as a result of the abnormalities that are not associated with bleeding (factor XII deficiency and 76 Journal compilation Ó 2006 Blackwell Publishing Ltd

NEW APPROACHES FOR MEASURING COAGULATION 77 lupus anticoagulant), while in 63 (49%) all tests gave normal results. The sensitivity of PFA-100 for VWD was 86% (both cartridges), for PFD 75% (C-EPI) and 8% (C-ADP). The sensitivity of BT for VWD was 29%, for PFD 33%. After dividing the patient population into four quartiles of distribution, according to the severity of the bleeding score, only the CT of the PFA-100 CT of C-EPI showed a progressive and significant prolongation from the first to the fourth quartile (P ¼ 0.04). In conclusion, (i) CT of the PFA-100 C-EPI was significantly associated with the severity of the bleeding history; (ii) PFA-100 CT showed a better sensitivity than BT for VWD and PFD; (iii) at variance with the BT, PFA-100 could help discriminate between VWD (prolongation of CT of both cartridges) and PFD (prolongation of CT of the C-EPI cartridge only). The application of waveform analysis in coagulation Waveform analysis is the optical profile of clot formation on simple assays of coagulation, such as the activated partial thromboplastin time (aptt) [1]. Optical profiles can be characterized using a set of parameters describing onset and completion of coagulation, magnitude of signal change, rate of coagulation and other properties. Work from our group has shown that this can be atypical in patients with haemostatic dysfunction [2]. The biphasic waveform (BPW) was initially described when a decrease in plasma light transmittance prior to clot formation on the MDA 180Ò automated coagulation analyzer (biomérieux, Marcy, France) was shown to correlate with disseminated intravascular coagulation (DIC) in critically ill patients [3]. In contrast to the normal sigmoidal waveform pattern that is characterized by an initial 100% light transmittance phase prior to clot formation, patients with a biphasic pattern had an immediate, progressive fall in light transmittance that occurred even in the preclotting phase [4]. The magnitude of this BPW often varied with sequential samples taken from individual patients and appeared early in samples from patients who were later diagnosed with DIC by more conventional criteria [5]. The utility of this for the forewarning of DIC, the diagnosis of sepsis and the risk of mortality have been validated by others both in the intensive care and in the emergency room setting [6 8]. In contrast to the significance of the initial slope (slope 1), work by others has also demonstrated the utility of the second derivative of the waveform. The group at Nara University has shown that this aspect of aptt waveform analysis can be utilized in the assessment of factor VIII (FVIII) levels below 1.0 IU dl )1. Whereas conventional one-stage clotting assays are insufficiently sensitive at this level, Shima et al. [9] have shown that FVIII levels of 0.2 IU dl )1 can be measured reproducibly and that such accurate discrimination has phenotypic relevance and clinical significance. Furthermore, this can be utilized to monitor the effectiveness of FVIII infusions in haemophilia A patients with high responding inhibitors. In the presence of type 1 inhibitors quantified at 6 14 BU ml )1, 0.9 IU dl )1 FVIII could be detected 24 h after FVIII infusion [10]. This suggests that FVIII infusion may be continued with clinical benefit in some haemophiliacs with high responding inhibitors and that the haemostatic response can be monitored accurately and efficiently by waveform analysis. For those that do not respond to FVIII and require recombinant factor VIIa, waveform analysis can also indicate dosedependent improvements in clot formation [11]. Thrombin potential in haemophilia Haemophilia is a deficiency of factor VIII (FVIII) or IX (FIX) that leads to a defect in the thrombingenerating system of the blood. Although in haemophiliacs the deficient factor limits the haemostatic function, it is not its sole determinant. Its residual activity remains dependent upon the ensemble of clotting factors, platelets and the vessel wall. This is best illustrated by the observation that substitution with factor VIIa (FVIIa) instead of the lacking FVIII or FIX can restore haemostasis. In addition, the common clinical observation that haemophiliacs with identical residual levels of the missing factor can show large variations in bleeding tendency stresses this point. It therefore seems logical to try and measure the affected function rather than the missing factor, at least that part of the function, which is accessible because it is located in the blood, i.e. thrombin generation. Indeed, it can be shown that thrombin generation in haemophiliac blood is dependent on other factors than the deficient one. It is, for example, proportional to the prothrombin level and inversely proportional to the antithrombin level and augmented by inhibition of the activated protein C system. The most important practical reason for measuring thrombin generation in haemophiliacs is in fact its sensitivity to FVIII bypassing therapy with FVIIa-containing preparations. A typical thrombin generation curve (Fig. 1) shows a lag time, a steep rising slope and a slower declining Journal compilation Ó 2006 Blackwell Publishing Ltd Haemophilia (2006), 12, (Suppl. 3), 76 81

78 T. W. BARROWCLIFFE et al. Thrombin (nm) 400 300 200 100 0 0 5 10 15 20 25 30 35 40 Time (min) Fig. 1. Effect of factor VIII infusion on thrombin generation in a haemophilic patient. 50 IU kg )1 was given to a haemophilic patient. Upper fat line: normal plasma; lower fat line: patient before injection; thin lines: patient after injection. From top to bottom at the following times after injection: 15 min, 1, 3, 5, 24, 48, 30, 60 h. The thrombin generation curves are obtained with the CAT method using 2 pm tissue factor as a trigger. slope. Important parameters are the peak of thrombin activity attained and the area under the curve or endogenous thrombin potential (ETP). The importance of the latter parameter derives from the fact that it indicates how much substrate thrombin can convert during its lifetime in plasma. This parameter can only be found if the entire course of thrombin generation is recorded. The ETP and the peak value together determine the gradient of thrombin, which forms around a local trigger and hence the extent of the recruitment of all haemostatic mechanisms (on platelets and vessel wall) that are triggered by thrombin. It also determines the amount of thrombin-activatable fibrinolysis inhibitor (TAFI) formed and thus the resistance of the haemostatic plug to fibrinolysis. The fibrinolytic character of much haemophiliac bleeding suggests a direct link between late bleeding tendency and ETP via this mechanism. This may explain the fibrinolysis often observed in thrombelastography when thrombin generation is low. The thrombin generation curve can be obtained by either subsampling or conversion of a slow chromogenic or fluorogenic substrate. In the last few years, we developed calibrated automated thrombinography, a method that can record the concentration of thrombin in clotting platelet-poor plasma (PPP) or platelet-rich plasma (PRP) continuously in up to 24 samples. It is based on the conversion of a slowreacting small fluorogenic substrate that does not interfere with the normal mechanism of prothrombin conversion. It slows down thrombin inactivation by a precisely known percentage (50%), which depends upon the binding constant of the substrate and the amount added. When this is corrected for, the resulting curves superimpose with those obtained by subsampling under identical conditions. With this method, the restoration of thrombin generation in haemophiliacs by infusion of FVIII preparations (Fig. 1) or by preparations containing activated factor VII is readily seen. An approximate impression of thrombin generation can also be obtained by surrogate methods, i.e. methods that do not render thrombin concentration (in nm) as a function of time. Measuring the first derivative of substrate conversion without proper calibration gives curves that to the non-initiated looks deceivingly like a thrombogram but that gives a distorted view and introduces important errors. The first derivative being a function of both thrombin concentration and substrate concentration, variation in the latter needs to be corrected for by continuous calibration. If not, low thrombin generation, as in haemophilia, is overestimated because more fluorogenic substrate is available towards the end of the reaction than in the normal thrombin generation. Other surrogate methods are based on the conversion of fibrinogen into fibrin, measuring either its turbidity or its tensile strength. Through mathematical manipulation, curves can be obtained, which look like thrombin generation (TG) curves but are not. Because fibrinogen is converted into fibrin long before all prothrombin is converted into thrombin, essentially only the lag time of the thrombelastogram is comparable to one of the parameters of the actual thrombin generation curve. After clotting occurs, the signal results from the changing properties of the clot rather than from the amount of thrombin present. The method has the advantage that it can be carried out in whole blood and that fibrinolysis is readily observed. The latter, via TAFI, may be a secondary effect of a low ETP (see above). Surrogate methods have the additional disadvantage that the signals are influenced by variables outside the thrombin-generating system, such as fibrinogen content, and the Ca ++ -dependent interaction of very low-density lipoproteins and C-reactive protein. In contrast to the classical view, it becomes more and more obvious that platelets are not just responsible for that phase of haemostasis that occurs before thrombin is formed. Recent experiments show clearly that thrombin starts to form within seconds after a lesion occurs. Platelets influence the amount and the effect of thrombin formed in at least three ways: by forming an aggregate from which the formed thrombin cannot be washed away, by providing the required procoagulant phospholipids and by releasing factor V. Thrombin generation in Haemophilia (2006), 12, (Suppl. 3), 76 81 Journal compilation Ó 2006 Blackwell Publishing Ltd

NEW APPROACHES FOR MEASURING COAGULATION 79 platelet-rich plasma, therefore, is one step nearer again to complete physiological function testing than TG in PPP. This is especially important in von Willebrand s disease where impaired thrombin generation can be demonstrated, that is not, or only partly, due to the lack of FVIII but must be attributed to impaired platelet activation via the fibrin von Willebrand factor (VWF) GPIb platelet activation axis [12]. By comparison of the thrombin generation test in PRP to that in PPP with added phospholipids, this phenomenon allows to differentiate between the effect on thrombin generation of FVIII and of VWF. Experiences in comparison of waveform, thrombin generation test and bioassay in haemophilia patients Historically, laboratory practises adopted in the management of haemophilia in recent times have been based on the experiments conducted in the early 1950s when the activated partial thromboplastin time (APTT) method was first devised [13]. Today, a prolonged APTT test result found in a male person with an abnormal tendency to bleeding should always raise suspicion of haemophilia, and the relevant investigation will include a bioassay recording the level of factor VIII and/or factor IX (FIX) based on the APTT methods. Should the patient suffer from haemophilia, APTT assays are also employed to monitor treatment. Thus, common laboratory routines around haemophilia are closely linked with the earliest small signal of fibrin formation. However, one must be aware that the APTT method only records the beginning of the overall clotting process, leaving >95% of the reaction untold. For this reason, methods were developed and published providing more detail of the entire course of the clotting process in haemophilia as early as 55 years ago. In principle, two avenues were followed: one type of assay recorded the thrombin activity generated during the coagulation process [14], while the other assay illustrated the formation of fibrin during clotting [15]. In the original versions, these methods were quite crude and imprecise and were not well suited for routine use and therefore were not generally adopted in clinical work. Thrombin generation methods Recent progress has refined these assays to a large extent, and the thrombin generation principle has been used more generally in haemostasis research [16,17]. As thrombin formation occurring during plasma coagulation is quite explosive by nature, slow-reacting substrates have been developed and introduced. For practical reasons, in particular when the thrombin generation test is used in the study of plasma with platelets, thrombin-induced peptide hydrolysis in a chromogenic reaction has been superseded by fluorogenic substrates [18]. In order to study continuous thrombin generation in plasma, either fibrinogen must be removed or its polymerization into fibrin must be inhibited. The continuous curve of thrombin generation is usually differentiated and derived parameters such as the lag time, the peak value and the area under the curve are employed in the interpretation of the signal. Most commonly, the activator is tissue factor in small amounts, but a variety of concentrations of tissue factor have been used, making direct comparisons quite difficult. A major clinical and laboratory hallmark of haemophilic plasma clotting the prolonged initiation phase seems to be lost if too much tissue factor is used for activation [19]. Alternatively, for the study in haemophilia A, the use of activated FIX as an activator has been more successful in this regard [20]. Fibrin formation methods The systems developed for recording fibrin formation, as assessed by the change in the viscoelastic properties of blood or plasma during coagulation, had to come quite some way until velocity characteristics were introduced to substitute for the classical Ôbrandy-glassÕ shape signature. Two systems are available today and quite comparable in performance: the thrombelastograph and the rotating thrombelastometer. Compared with Hartert s original model, the new equipments employ single-use reaction cells and the raw continuous signal may be converted into velocity by differentiation. With the viscoelastic recorders, the optimal activator for study in haemophilia is tissue factor in low concentrations at 0.35 pm or less [21]. These assays appear to mirror the haemophilic bleeding phenotype quite well, as the initiation of clotting is severely prolonged in severe haemophilia, and the fibrin formation signal lacks acceleration and is weakly represented [22]. Overall haemostasis potential In addition, a slightly different system has been communicated, in which fibrin formation dynamics in plasma is recorded by photometry following activation with a small amount of thrombin. This method denoted the overall haemostasis; haemo- Journal compilation Ó 2006 Blackwell Publishing Ltd Haemophilia (2006), 12, (Suppl. 3), 76 81

80 T. W. BARROWCLIFFE et al. static potential also has the convenience of measuring the fibrinolytic resistance of the clot. The most recent report demonstrates the feasibility of the method in hypocoagulable states such as haemophilia and other single factor deficiencies [23]. A study was recently reported comparing thrombin generation with fibrin formation in whole blood showing a quite convincing correlation between the two principles of recording whole blood coagulation in a continuous manner. As fibrin formation is the consequence of thrombin generation, this finding is by no means surprising [24]. Conclusions The PFA-100 would appear to be useful in the diagnosis of different types of von Willebrand s disease (VWD) and in distinguishing VWD from platelet function. As well as showing a better correlation with the various bleeding disorders, the PFA-100 has the clear advantage over the bleeding time of practical convenience. Analysis of the waveform of the clotting process can be carried out in a number of ways with appropriate instrumentation. The biphasic waveform is atypical in disseminated intravascular coagulation and has been found useful in diagnosing this coagulation abnormality. In haemophilia, analysis of the second derivative has allowed quantitative measurements at Factor VIII (FVIII) levels below I IU dl )1 and has proved useful in monitoring the treatment of inhibitor patients. The thrombin generation test can be used to monitor treatment in haemophilia and gives more information than factor assays, as it takes into account non-fviii influences on clotting, such as thrombin-activatable fibrinolysis inhibitor. However, the concentration of tissue factor is critical for the application of thrombin generation test (TGT) in haemophilia, and an alternative activator that has been used successfully is Factor IXa. Two other tests that may be of clinical benefit are thrombelastography and the overall haemostasis potential, although these have not been studied as intensively in haemophilia as TGT and waveform analysis. References 1 Braun PJ, Givens TB, Stead AG et al. Properties of optical data from activated partial thromboplastin time and prothrombin time assays. Thromb Haemost 1997; 3: 1079 87. 2 Downey C, Kazmi R, Toh CH. Novel and diagnostically applicable information from optical waveform analysis of blood coagulation in disseminated intravascular coagulation. Br J Haematol 1997; 98: 68 73. 3 Downey C, Kazmi R, Toh CH. Early identification and prognostic implications in disseminated intravascular coagulation through transmittance waveform analysis. Thromb Haemost 1998; 80: 65 9. 4 Toh CH, Samis J, Downey C et al. The biphasic transmittance waveform in the aptt coagulation assay is due to the formation of a Ca ++ -induced complex of C-reactive protein with very low density lipoprotein and is a novel marker of impending disseminated intravascular coagulation. Blood 2002; 100: 2522 9. 5 Toh CH, Ticknor LO, Downey C, Giles AR, Paton R, Wenstone R. Early identification of sepsis and mortality risks through simple, rapid clot-waveform analysis. Intensive Care Med 2003; 29: 55 61. 6 Dempfle CE, Lorenz S, Smolinski M et al. Utility of activated partial thromboplastin time waveform analysis for identification of sepsis and overt disseminated intravascular coagulation in patients admitted to a surgical intensive care unit. Crit Care Med 2004; 32: 520 4. 7 Smith EY, Charles LA, Van Cott EM. Biphasic activated partial thromboplastin time waveform and adverse events in non-intensive care unit patients. Am J Clin Pathol 2004; 121: 138 41. 8 Fernandes B, Giles AR. An abnormal activated partial thromboplastin time clotting waveform is associated with high mortality and a procoagulant state. Lab Hematol 2003; 9: 138 42. 9 Shima M, Matsumoto T, Fukuda K et al. The utility of activated partial thromboplastin time clot waveform analysis in the investigation of hemophilia A patients with very low levels of factor VIII activity. Thromb Haemost 2002; 87: 436 41. 10 Kasuda S, Tanaka I, Shima M et al. Effectiveness of factor VIII infusions in haemophilia A patients with high responding inhibitors. Haemophilia 2004; 10: 341 6. 11 Sima M. Understanding the hemostatic effects of recombinant factor VII a by clot waveform analysis. Semin Hematol 2004; 41(Suppl. 1): 125 31. 12 Béguin S, Kumar R, Keularts I, Seligsohn U, Coller BS, Hemker HC. Fibrin-dependent platelet procoagulant activity requires GP1b receptors and von Willebrand factor. Blood 1999; 93: 564 70. 13 Langdell RD, Wagner RH, Brinkhous KM. Effect of anti-hemophilic factor on one-stage clotting tests: a presumptive test for hemophilia and a simple one-stage anti-hemophilia assay procedure. J Lab Clin Med 1953; 41: 637 47. 14 Biggs R, MacFarlane RG. The reaction of haemophilic plasma to thromboplastin. J Clin Path 1951; 4: 445 59. 15 Hartert H. Thrombelastography, a method for physical analysis of blood coagulation. Z Gesamte Exp Med 1951; 117: 189 203. 16 AlDieri R, Peyvandi F, Santagostino E et al. The thrombogram in rare inherited coagulation disorders: its relation to clinical bleeding. Thromb Haemost 2002; 88: 576 82. Haemophilia (2006), 12, (Suppl. 3), 76 81 Journal compilation Ó 2006 Blackwell Publishing Ltd

NEW APPROACHES FOR MEASURING COAGULATION 81 17 Hemker HC, Giesen P, AlDieri R et al. The calibrated automated thrombogram (CAT): a universal routine test for hyper- and hypocoagulability. Pathophysiol Haemost Thromb 2002; 32: 249 53. 18 Hemker HC, Beguin S. Phenotyping the clotting system. Thromb Haemost 2000; 84: 747 51. 19 Dargaud Y, Beguin S, Lienhart A et al. Evaluation of thrombin generating capacity in plasma from patients with haemophilia A and B. Thromb Haemost 2005; 93: 475 80. 20 McIntosh JH, Owens D, Lee CA, Raut S, Barrowcliffe TW. A modified thrombin generation test for the measurement of factor VIII concentrates. J Thromb Haemost 2003; 1: 1005 11. 21 Sørensen B, Johansen P, Christiansen K, Woelke M, Ingerslev J. Whole blood coagulation thrombelastographic profiles employing minimal tissue factor activation. J Thromb Haemost 2003; 1: 551 58. 22 Ingerslev J, Poulsen LH, Sørensen B. Potential role of the dynamic properties of whole blood coagulation in assessment of dosage requirements in haemophilia. Haemophilia 2003; 9: 348 52. 23 Antovic A, Blombäck M, Sten-Linder M, Petrini P, Holmström M, He S. Identifying hypocoagulable states with a modified global assay of overall haemostasis potential in plasma. Blood Coagul Fibrinolysis 2005; 16: 585 96. 24 Rivard GE, Brummel-Ziedins KE, Mann KG, Fan L, Hofer A, Cohen E. Evaluation of the profile of thrombin generation during the process of whole blood clotting as assessed by thrombelastography. J Thromb Haemost 2005; 3: 2039 43. Journal compilation Ó 2006 Blackwell Publishing Ltd Haemophilia (2006), 12, (Suppl. 3), 76 81