Tests of global haemostasis and their applications in bleeding disorders

Authors


Alok Srivastava, Professor of Medicine, Department of Haematology, Christian Medical College, Vellore – 632004, Tamil Nadu, India
Tel.: +91 416 2282352; fax: +91 416 2226449;
e-mail: aloks@cmcvellore.ac.in

Abstract

Summary.  There is a potential for significant paradigm shift in the assessment of haemostasis from the conventional plasma recalcification times, such as prothrombin time (PT) and activated partial thromboplastin time (APTT), which correspond to artificially created compartments of haemostasis to tests that assess the entire process in a more physiological and holistic manner. These include the thrombin generation test, thromboelastogram and the clot wave form analysis. While these tests have been described many years ago, there is renewed interest in their use with modified technology for assessing normal haemostasis and its disorders. Although early data suggest that they can provide much greater information regarding the overall haemostasis process and its disorders, many challenges remain. Some of them are possible only on instruments that are proprietary technology, expensive and are not widely available. Furthermore, these tests need to be standardized with regard to their reagents, methodology and interpretation, and finally, much more data need to be collected regarding clinical correlations with the parameters measured.

Introduction

Haemostasis and its abnormalities have been traditionally assessed by plasma clotting times, such as the prothrombin, activated partial thromboplastin and thrombin times[1]. These times depend on the thrombin dependent conversion of fibrinogen to fibrin, but note only the initiation of this process and not its speed or total extent. Factor assays based on these tests have defined the different coagulation disorders including haemophilia[2,3]. While tests have been useful for these purposes, they also have several limitations. These include the fact that all of them are performed under conditions that are far from physiological, and they split the process of coagulation into artificial segments thus not assessing the potential impact of other components of the haemostatic system. Factor assays performed using these tests are limited by their sensitivity at very low levels [4]. Factor levels below 1.0% (0.01 IU/mL) have therefore not been traditionally quantified. In many patients with coagulation disorders, factor assays alone do not correlate well with clinical symptoms. It has been shown that plasma from some patients with severe haemophilia A (HA) has the ability to generate thrombin [5]. The exact basis for this phenomenon is not well understood, but may be related to the balance of levels of different procoagulant and anticoagulant proteins in the blood [6]. It is possible that tests that assess global haemostasis may be better reflective of the clinical features.

Currently, there are no widely available and standardized tests that can quantitatively assess the overall haemostastic potential of blood. The process of thrombin generation and fibrin clot formation can be captured with greater sensitivity and completeness by tests that measure global haemostasis. These include the thrombin generation tests/assay (TGT/TGA) [5,7], thromboelastography(TEG) [8] and the activated partial thromboplastin time (APTT) waveform analysis (WA) [9] using different instrument systems. These tests have not only helped in more complete assessment of the process of normal haemostasis but have also provided newer insights into the evaluation of disorders of haemostasis. However, several issues remain to be resolved with regard to standardization of methodology and interpretation of these tests. This study will describe some of these issues with particular reference to hereditary coagulation disorders.

The thrombin generation test

Concept and methodology

Thrombin is the final product and the key enzyme of the coagulation system. Thrombin generation measurement would be therefore able to reflect the overall coagulating capacity of each individual, taking into account the effect of all parameters influencing the coagulation system. In addition, to TGT that measure the overall potential of plasma to form thrombin, there is a second type of assay that measures whether more than normal amounts of thrombin are formed in vivo, i.e. the measurement of molecules that result from thrombin formation and thrombin action. This group includes D-dimers that indicate that fibrin has been formed, F1 + 2 that indicate that prothrombin has been split and thrombin-antithrombin complex (TAT) that indicates that active thrombin has been present. These products do not represent overall coagulation capacity but are markers of ongoing coagulation activation, while TGT is an activity assay representing an individual’s potential to generate thrombin, should coagulation triggering circumstances arise. For this reason, it can be surmised that the outcome of the TGT is directly related to risk, whereas the other tests show whether the risk has actually worked out. Over the last 15 years, a technique has been developed at the University of Maastricht by Hemker et al.[10] in which a specific slow reacting fluorescent substrate of thrombin is added to either platelet-poor and platelet-rich plasma samples without defibrination, and the course of thrombin formation is monitored in real time. These technical developments of the TGT make it potentially applicable to clinical laboratories.

Clinical applications

Correlations between the TGT parameters and clinically observed bleeding in patients with haemophilia and other inherited bleeding disorders have been published[5,11]. This correlation between clinical bleeding risk and thrombin generating capacity led to the evaluation of this technology as a surrogate marker for by-passing therapy in haemophilia patients with anti-FVIII/FIX alloantibodies[12]. It has also been shown that TGT is sensitive to hypercoagulability [13]. Endogenous thrombin potential (ETP) representing the enzymatic activity of the generated thrombin during its life-time is the parameter that correlates with clinical bleeding or thrombosis [5,11–13]. Moreover, there are several studies on the standardization of the TGT making its use suitable in clinical trials, in comparison with other global haemostasis assays, e.g. thromboelastography that are not yet as well standardized[14,15]. However, wide acceptance of TGT as a clinical tool to evaluate individual clinical phenotype of patients with coagulation disorders also depends on the accuracy and precision of the test. The ability to reproduce reliable thrombin generation measurements should be facilitated by the use of standardized preanalytical and analytical procedures.

Preanalytical and analytical variables

It has been shown that inappropriate phlebotomy and sampling materials may produce significant activation of coagulation and may be responsible for erroneous results[16]. Thrombin generation measured in platelet-rich plasma (PRP) samples obtained from Vacutainer tubes® (Becton Dickinson, Meylan, France) with a negative air pressure inside overestimates the coagulation capacity in comparison with that obtained from Monovette S® tubes (Sarstedt, Orsay, France) that have a piston allowing a slow aspiration of blood and therefore limiting platelet damage[17]. The addition of a contact factor inhibitor into the collection tubes, i.e. corn trypsin inhibitor (CTI) may significantly reduce imprecision of TGT results obtained in the presence of low tissue factor concentrations ≤1 pM [14]. It has been recently suggested that contact activation was particularly high with ‘butterfly’ needles equipped with tubing, which are widely used in hospitals [18].

One of the most critical steps among the preanalytical variables is the preparation of plasma samples. Previous data have showed that traces of platelets and/or white cells present in platelet-poor plasma (PPP) could dramatically change the TGT results in haemophiliacs and lead to over-estimations of the thrombin generation capacity with a shorter lag time and falsely increased peak and ETP values [17]. This could be explained by the presence of additional tissue factor (TF) bared on leukocytes and phospholipids (PL) coming from platelets and cell surfaces, as thrombin generation is closely dependent on the TF and PL concentrations present in the medium. The use of double-centrifuged PPP samples is therefore required for TGT.

Another important preanalytical parameter of interest is the dilution of plasma in the test medium. It has been demonstrated that over dilution of plasma sample has a negative effect on the participation of the antihaemophilic factors VIII and IX in thrombin generation driven by low TF concentration. TGT does not discriminate between low and high FVIII plasma levels at dilutions >1:6[19]. This would be explained by the fact that plasma dilution alters procoagulant and anticoagulant pathways differently and slows down the inhibitory activity of tissue factor pathway inhibitor (TFPI) to a greater extent when compared with the down regulation by diluting procoagulant factors[19].

The temperature can also influence the reliability of thrombin generation results. Commercially available softwares start reading fluorescent signal at 37°C as soon as the starting mixture containing thrombin substrate and calcium is dispensed into the measurement wells. For this reason, it is recommended to warm up the plate at 37°C during 5–10 min before the measurement. When a plate at room temperature is used for thrombin generation measurement, the results are overestimated, which can lead to miss the diagnosis and underestimate the bleeding risk in patients with mild haemophilia (Fig. 1). Thrombin generation measured in PPP samples from five patients with mild HA showed 32% overestimated thrombin generating capacity when the plate was not preheated at 37°C in comparison with a plate warmed up at 37°C during 10 min before thrombin generation measurement (ETP = 1045 ± 47 nM min vs. 707 ± 33 nM min and thrombin peak = 91 ± 3 nM vs. 57 ± 5 nM).

Figure 1.

 Representative thrombin generation curves demonstrating the importance of temperature for thrombin generation measurement. Thrombin generating capacity of a patient with mild haemophilia A (HA) (FVIII = 30 IU/dL) was measured in platelet-poor plasma (PPP) with tissue factor (TF) 1 pM and phospholipids (PL) 4 μM (final concentrations). The black curve shows thrombin generation of the patient when a plate at ambient temperature was placed into the fluorometer for thrombin generation measurement. The grey curve corresponds to the real thrombin generating capacity of the same patient measured after preheating the plate at 37°C during 5 min. In both cases, the temperature of the fluorometer was 37°C.

There are several promising preliminary results reporting that TGT can detect hypocoagulability and hypercoagulability related to coagulation disorders, but the real challenge is to predict the likelihood of clinical outcome and to identify the individual bleeding or thrombosis risk of each patient to individually tailor their management. The ability of TGT to predict the clinical outcome with high accuracy and low imprecision should be assessed in prospective multicenter clinical trials that can be conducted once the test is standardized regarding the preanalytical and analytical conditions that directly influence the reliability of results.

Thromboelastography

Concept and methodology

The evaluation of coagulation has come a long way from assays such as the PT and APTT, and global assays such as Thromboelastography are now regaining an interest as they lend the ability to assess the cumulative effects of the plasma factors and platelets, leucocytes and red cells on coagulation. Thromboelastography was developed by Dr. Hellmut Hartert in 1948, and the term was used to describe the trace produced from measuring the viscoelastic changes seen with fibrin polymerization. Thromboelastography allows the evaluation of clot formation from initialization into formation and stability. The two instruments currently available are: TEG® Hemostasis Analyzer (Haemonetics Corp, Braintree, MA, USA) and the ROTEM® (Tem International GmbH, Munich, Germany).

The basic principle of thromboelastography involves incubation of whole blood or PRP in a heated sample cup into which is suspended a pin. The pin and the cup or the cup alone oscillates, and as the blood clots, the motion of the cup is transmitted to the pin which is recorded via a computer. There are minor mechanical differences between the two instruments, and the activators used differ with respect to potency. In the TEG®, a sensor (pin) is connected with a torsion wire, and clot formation generates a physical connection between the cup and sensor, which is recorded via a mechanical-electrical transducer. In the ROTEM®, the pin (sensor) is fixed on the tip of a rotating shaft, whereas the sample cup is stationary and the position of the axis is detected by reflection of light on a small mirror on the axis. Therefore, the results differ from each other and are not comparable between instruments.

Clinical applications

Even though the method has not yet been fully standardized, there are more than 300 publications in the field of thromboelastography and bleeding disorders. Its initial utility was to decrease the transfusion requirements in patients with complicated surgical procedures, but it has now expanded to include bleeding as well as thrombotic disorders. Thromboelastography has defined the phenotypic variation seen in patients with severe haemophilia both in adults and children[20,21]. Using this test, we were able to demonstrate a decrease in fibrin polymerization in haemophilia patients with increased clinical bleeding compared with those who had mild bleeding symptoms, although the factor VIII levels were <1% in all cases[21]. This information can then be used to rationally tailor the treatment in each individual. Management of bleeding in haemophilia patients with inhibitors remains a great challenge. The greatest impact of thromboelastography in the field of haemophilia has been in the monitoring of by-passing agents, such as activated recombinant FVII (rFVIIa) and activated prothrombin complex concentrate (APCC’S). With thromboelastography, clot formation can be assessed to determine efficacy of by-passing agents, instead of measuring individual clotting factor activity. This has brought about the ability to assess haemostasis with product replacement therapy in such patients[22,23]. Recent studies using thromboelastography have demonstrated the changes in clot growth and disintegration with the use of epsilon amino caproic acid (EACA) [24] in vitro, whereas others have shown the clinical efficacy of antifibrinolytics during surgery to prevent or control bleeding, using thromboelastography[25]. The benefits of using antifibrinolytics in haemophilia with inhibitors and decreased use of APCC’S have also been shown using thromboelastography[26]. This is of major consequence in the developing world. In the field of platelet disorders, the thromboelastograph has demonstrated laboratory evidence of response to treatment with agents such as rFVIIa[27]. Other coagulation factor deficiencies such as FXIII deficiency are difficult to diagnose and monitor due to lack of universally available laboratory tests, and thromboelastography has been shown to be extremely useful in this situation[28]. Modified thromboelastography (TEG Platelet Mapping®) has also been extensively used to monitor antiplatelet agents and prevent bleeding complications[29]. In the field of haemostasis during major surgical procedures and trauma, thromboelastography has proven to be beneficial to predict transfusion requirements and guided transfusional therapy[30]. In neonates, the small volume of blood required for thromboelastography testing makes it a valuable tool in assessing for coagulopathy and needs to be investigated. Dysfibrinogenemias are a unique group of disorders with a potential for either bleeding or thrombosis. Until the patient develops symptoms, there are no assays to define this risk. Thromboelastography may provide an insight as it provides an opportunity to evaluate fibrinogen clot formation. A three-year-old child and her mother with severe bleeding symptoms were shown to have a novel mutation, AGT to CGT at codon 313 in the fibrinogen gene (which was named Fibrinogen Detroit II), with resultant prolonged lag time to clot formation as demonstrated by thromboelastography (Fig. 2) (Letter to the editor, Thrombosis and Hemostasis, 103.2/2010). The unique ability to study clot formation from initiation to breakdown makes this a valuable tool for assessment of every aspect of coagulation and fibrinolysis.

Figure 2.

 Low maximum amplitude (MA) indicating poor fibrinogen clot formation as demonstrated by Thromboelastography Platelet Mapping Assay® in patient with hypo-dysfibrinogenemia.

Preanalytical and analytical variables

While thromboelastography has generated considerable interest, its potential for increased clinical utility in assessing haemostasis and monitoring therapy will depend on the test being standardized with demonstration of clinical reliability and essential elements of any good laboratory assay.

A critical preanalytical issue is the collection of blood without contact activation. The same precautions mentioned above for sample collection using CTI in the collection tubes also apply to samples collected for thromboelastography[31]. The source and amount of CTI can also be an issue when very low concentrations of tissue factor are used for the test. If thromboelastography is to be performed on plasma, double centrifugation as well as standardization of centrifugation protocols may also help get more reproducible results. For whole blood samples, storage temperatures prior to testing can also affect results, and it is best to store at 37°C.

APTT waveform analysis

Concept and methodology

As a screening test for haemophilia, a prolonged APTT time conventionally measured reflects the initiation of the fibrin formation process. At the same time, the most advanced automated coagulometers, especially those with photo-optical mode of detection, continue to measure the entire process of rate of fibrin formation over time in addition to clocking the APTT in seconds. This is recorded as the change in the optical output of the incident light of photo-optical coagulometers during the formation of the precipitating fibrin clot and is recorded as a clot curve that can be displayed on the monitor of the coagulometer. Further analysis of the configuration of this curve provides information that correlates with the velocity and amount of fibrin formation as reflected in the height and shape of the curve. Computing of the 1st and 2nd derivatives of the clot curve was first reported by Braun et al.[9] as the APTT waveform analysis (APTT WA) on MDA coagulometer (Organon Teknika, Durham, North Carolina, USA). These parameters therefore quantify the velocity and acceleration, respectively, reflecting the rate at which fibrinogen is being converted to fibrin which is in turn dependent on the kinetics of coagulation factors generating thrombin. Shima et al. demonstrated the utility of APTT waveform analysis (APTT WA) to define qualitative and quantitative differences at levels of FVIII:C less than 0.01 IU/mL, raising the possibility that the correlation observed, between the laboratory definition of severity and the clinical phenotype, could be improved by this approach[32].

We have developed an alternative way of measuring the same phenomenon on the ACL 10 000 (IL, Milan, Italy) and called it the APTT Clot Curve Analysis (CCA). It assesses the thrombin generation and fibrin clot formation based on the analysis of the photo-optical data, light scatter (LS), from the ACL 10 000[33,34]. The light scatter is harnessed and exported and processed offline on Microsoft Excel. The first and second derivatives were calculated from the clot curve data. The first derivative (dLS/dt) is the change in light scatter signal detected over unit time and is therefore indicative of the velocity of the clotting process taking place in the plasma after recalcification. The second derivative (d2LS/dt2) is the degree of change in light scatter with respect to time, and the maximal change quantifies coagulation acceleration. The degree of change in light scatter is the maximum where the curve ascends when clotting is initiated and corresponds to the maximum value of the second derivative (Max2) (Fig. 3). The maximum of 2nd derivative was used to measure the degree of acceleration of clot formation. APTT, on carefully selected FVIII deficient clinical sample spiked with rFVIII concentrate to obtain random concentrations ranging from 0.01 to 0.3 IU/mL, showed progressively increasing values of Max2 on the CCA with very good correlation between the Max2 and FVIII:C (r > 0.99).

Figure 3.

 Activated partial thromboplastin time (APTT) clot curve and its derivatives. The APTT curve above shows the changes in light scatter (LS) against time (t). The graph below shows both the first derivative (dLS/dt) and in dotted line the second derivative (d2LS/dt2). From the derivative, it is easy to identify the initiation of coagulation, the velocity and the point of acceleration.

Clinical relevance

We have evaluated the APTT-CCA in a cohort of patients with HA (severe – 70, moderate - 18, mild - 8 and healthy controls - 20). Median Max2 could not only discriminate haemophilia A from healthy controls but also between mild HA, moderate HA and Severe HA (Fig. 4), where the median Max2 clearly showed a decline as the FVIII:C decreased. However, Max2 in the 70 severe HA samples with the FVIII levels obtained on ACL 10 000 showed wide variations unlike spiked samples even when FVIII levels were same (less than 0.01 IU/mL), implying that FVIII is not the only determinant for the clot acceleration in clinical samples. Although a wide range was noted, they need to be correlated with the clinical profile to assess the usefulness of these tests in identifying phenotypic heterogeneity. This phenomenon has also been reported by Shima et al.[32]. Effectiveness of Factor VIII infusions in haemophilia A patients with high responding inhibitors reflected changes in APTT WA seen even 24 h after FVIII infusion and even when FVIII:C was less than 1.0 IU/dL[35]. It has also been used to monitor the use of bypassing agents such as APCC or rFVIIa in patients with haemophilia and inhibitors[36]. There are not many reports on the use of APTT WA in patients with other bleeding disorders, but its potential use in monitoring patients with disseminated intravascular coagulation (DIC) and sepsis[37] and in evaluating lupus anticoagulants[38] has been described.

Figure 4.

 Max2 on subtypes of Haemophilia A (HA) compared with Healthy controls (HC). Max2 values on clinical samples obtained from 70 Severe HA, 18 moderate HA and 8 mild HA patients compared with Max2 on plasma from 20 healthy controls. The median of Max2 significantly discriminates between each group.

Preanalytical and analytical variables

Preanalytical issues for APTT WA, a test initiated by contact activation, are just like preparing plasma in any APTT test that has to be free of platelets and contamination by tissue factor. In that sense, this test may be easier to standardize in more laboratories around the world. It may also be possible to modify this test to conditions where the plasma is activated at physiological concentrations of tissue factor for different applications. In those situations, other preanalytical issues such as avoiding contact activation by addition of CTI and taking steps to avoid activation by platelets and microparticles in the plasma will also be issues.

In conclusion, tests that assess global haemostasis have great potential for allowing a new look at the process of haemostasis. Much more work needs to be carried out to standardize their methodology, applications and clinical correlations with the measured parameters. This is being carried out through several independent groups working in this area as well as by working parties established by the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis. As efforts in this area advance, it is possible that there could be major paradigm shifts in the way in which we can assess haemostasis and evaluate its disorders.

Ancillary