SCIENTIFIC SUBCOMMITTEE SESSION
6 July 2007 Palexpo, Geneva, Switzerland  

Plasma Coagulation Inhibitors

Chair: E. Gray (UK)
Co-Chairs F. Bernardi (Italy), S. Kitchen (UK), H. Whinna (USA)

WHO international Standards

Proposed international standard for protein C, concentrate. E Gray (UK)

In 2006, a collaborative study was carried out to value assigned a replacement international standard for protein C, plasma and a new international standard for protein C, concentrate. The candidate plasma preparation, 02/342 was subsequently established by the Expert Committee on Biological Standardisation (ECBS) of the World Health Organisation (WHO) in October 2006 as the 2 nd International Standard for Protein C, Plasma. It was value assigned against the 1st International Standard (IS) for Protein C, plasma (86/622) and the labelled values are 0.85 and 0.84 IU/ampoule for function and antigen respectively.

For the proposed concentrate standard, 04/252, all participants and the SSC subcommittee approved the proposed assigned antigenic value of 14.3 IU/ampoule against the 1st plasma IS. However, based on their in-house experience, one participant did not agree with assigning the candidate with an overall functional potency ie combining the values obtained for chromogenic and clotting assays. As this has implications for the comparability of the proposed candidate with clinical products, further consideration and study was required to clarify this issue and so the recommendation to establish the 1st IS for Protein C, Concentrate was deferred until more discussion and/or data are available.

NIBSC and the participant who raised this concern have now carried out a joint study involving assays of a number of their production batches by chromogenic and clotting methods and have found that there is a distinct discrepancy between the potencies obtained by chromogenic and clotting assays. Taking into consideration that 02/342 has been established as the 2nd IS, the following new proposals and options were presented to the participants of the study and a panel of SSC experts: the candidate concentrate standard, 04/252 to be assigned with potencies relative to the 2nd International Standard for Protein C, Plasma, 02/342 with the two options. The first option is to label with chromogenic and antigenic value and the second option is to label with functional chromogenic and clotting potency and antigen value.

All participants agreed with the proposed assigned value for antigen. All participants agreed with the proposed assigned value for functional chromogenic activity. Five out of the 20 participants also agreed with labelling with functional clotting potency. All SSC experts agreed with the assigned values for antigen and functional chromogenic activity only. Therefore it will be recommended to the ECBS of the WHO to establish 04/252 with an antigenic and a functional chromogenic value of 14.5 and 15.0 IU/ampoule respectively.

Replacement of the 2 nd international standard for antithrombin, concentrate. E Gray (UK)

Twenty-one laboratories participated in a collaborative study to establish a replacement for the 2nd International Standard for Antithrombin, Concentrate (96/520). There was excellent agreement between laboratories, as indicated by low inter-laboratory % GCV for the 3 candidate materials which consisted of one recombinant and two plasma derived clinical concentrates. In terms of performance, stability profile and physical characteristics of the candidates, all 3 materials are very similar. However, there is a larger number of ampoules of sample C, 06/166 available. It is therefore proposed that sample C, 06/166 is considered as the 3rd International Standard for Antithrombin, Concentrate, with labelled potencies for both functional (4.4 IU/ampoule) and antigenic (4.5 IU/ampoule) activities. All participants have agreed with this proposal. All participants and SSC experts agreed with this proposal. Therefore it will recommended to the ECBS of the WHO that 06/166 to be established as the 3rd iS for Antithrombin, Concentrate, Human.

Protein S and protein C

Protein C and protein S asay discrepancies experience from UK NEQAS for Blood Coagulation. I Jenning (UK)

Accurate diagnosis of PC and PS deficiency depends on precise and reliable laboratory methods. UK NEQAS for Blood Coagulation exercises have identified discrepancies in results for PC and PS assays between users of different sources of commercial kit. For PC assays, marked differences (>10%) have been observed over the last 2 years between the two kits most widely used by participants in the programme, both employed by >60 laboratories.

In-house investigations used the SSC secondary plasma standard lot #3 as a calibrator for the two kits, and results for both lyophilised and frozen plasma demonstrated good agreement when this calibrator was used. Data indicated that the potency for one commercial reference plasma was incorrectly assigned by approximately 6%, which contributed to the discrepancy in the NEQAS exercises. A further UK NEQAS exercise in which a patient sample and the SSC secondary plasma standard were distributed demonstrated improved agreement in the patient sample results when results for both methods were cross-calibrated against the SSC plasma standard.

Discrepant results continue to be observed between PS kits, and the importance of locally-determined reference ranges is demonstrated by a >10% error rate in diagnosis of PS deficiency if a local range is not utilised.

Protein C and protein S assay discrepancies - North Amercia experience. R Marlar (USA)

Often there is a lack of correlation between thrombophilia diagnostic test result and correct clinical phenotype. This could be due to numerous interfering substances in the tests or other conditions that may affect the test and interpretations of results. For example, genetic deficiency of PC does not correlate with clinical phenotype (venous thrombosis). Plasma level of protein C or genetic mutation do not always predict clinical phenotype. Age was not found to be a factor for thrombosis and major life events do not affect phenotype. Therefore other non-protein C factors must influence the development of thrombosis . I nterfering substances or conditions can cause inaccurate test results. For example, heparin therapy can cause decrease of factors utilizing or influenced by heparin and warfarin causes decrease of vitamin K-dependent protein levels (and increased Antithrombin levels). The interpretation of protein C and protein S test results should take into account that 1. expression of clinical phenotype is not consistent with genotype, 2. thrombophilia phenotype is a multi-mechanism disorder based on multiple genes & acquired RF, 3. inconsistencies of test results can be due to patient, pre-analytical variables, interfering substances and/or assay problems and 4. cost effective protocol of diagnostic tests should be ordered based on prevalence and understanding.

Global coagulation/haemostatic tests

Progress on the activities of the working party on thrombin generation tests.
E Gray (UK) on behalf of the Working Party on Thrombin Generation Tests

The Working Party (WP) on Thrombin Generation Tests (TGT) was set up in 2004 under the auspice of the Plasma Coagulation Inhibitors subcommittee. The main remit is to investigate, standardise and validate methodologies for the quantitation of results to facilitate good within and between laboratory agreement. In 2004 a survey on current TGT methods was carried out and the WP published a mini review (on line publication www.bloodmed.com The Thrombin Generation Test (TGT) by Lawrie et al.). In 2005 the survey results and a pilot study results within WP on chromogenic non-sub-sampling methods were presented at the SSC. In 2006 a study on “Fluorogenic Methods for Thrombin Generation Tests” was initiated and completed. It is the intention of the WP to publish the data on the findings of the fluorogenic method study in 2007 and also to initiate and complete a study on the feasibility of establishing a reference plasma for thrombin generation tests. The WP also has plans to investigate the application of thrombin generation tests in the study of haemophilic plasma.

Report on the international collaborative study on thrombin generation tests.
E Gray (UK) on behalf of the Working Party on Thrombin Generation Tests

The main aim of this study is to investigate the feasibility of establishing a reference plasma for thrombin generation tests. One hundred and ten labs returned results and in total there was 128 sets of data available for analysis. Six coded freeze-dried samples including 3 candidate normal pooled platelet poor plasmas, the SSC Lot#3 (included for comparative purpose only) and 2 abnormal plasmas were provided. Four commercial methods (CAT, Technothrombin, Dade-Behring-ETP and In-TDT) were used by the participants and results were also obtained for 4 in-house methods. This study confirmed results from the previous study that a reference plasma would improve intra- and inter-laboratory variability. The Working Party therefore recommends the establishment of a reference plasma for Thrombin Generation Tests. Further discussion will be required to determine how the reference plasma should be used. The Working Party will also make a proposal to the SSC to establish the reference plasma as a SSC reference material.

Inter-laboratory evalutation of the TGA Assay. P Meijer (NL)

The first survey on the Technothrombin TGA concluded that the inter-laboratory variation observed depends upon the read-out variable and the level of thrombin generation. There is a difference in quantification of thrombin generation between laboratories and that this difference is affected by the type of instrument used. This second survey aimed to investigate if there is improvement in the inter-lab variation, are there still quantification differences, does the effect of instrument still exist and could harmonization improve result comparability. Forty-one participants were involved and a set of 8 plasmas including normal and abnormal samples were sent to each lab. The inter-laboratory variability for the measurement of thrombin generation in a normal pooled plasma is comparable in both surveys. The quantification differences between laboratories still exist and that these differences are more systematically higher the higher the thrombin generation. There are small but not statistically significant differences between fluorimeters. There are obvious differences in the inter-laboratory variability between fluorimeters. Harmonisation is only possible if there is a good correlation between two samples. Samples with low to very low thrombin generation did not show a good correlation with a normal sample. Harmonisation of absolute thrombin generation did not improve the inter-laboratory variation. Harmonisation by expressing readouts relative to a refrence plasma shows a small improvement of the inter-laboratory variation.

Report on the ECAT Workshop on Thrombin Generation Tests. K Kluft (NL)

Four manufacturers of Thrombin Generation Tests (TGT) Kits took part in a wet workshop to demonstrate their details of their techniques on 9 plasma samples provided by ECAT. The 4 kits were CAT (Thrombinoscope), Technothrombin, TGA (Technoclone), Dade-Behring-ETP (Dade Behring) and In-TDT (Pentapharm). Two variants of two of the kits were also included, so in total there was 6 different methods.

The samples included plain pooled plasma, pooled plasma spiked with hemolysed cell material, alpha-2-macrogloblin (alpha-2-M), argatroban, unfractionated heparin, an ultracentrifuged pooled plasma with low level of microparticle, protein S congenital deficient plasma, a factor VIII congenital deficient patient and a plasma with lupus antibodies.

The absolute data could only be compared when expressed relative to the pooled plasma. It was observed that the 9 plasmas showed quite different TGT profiles and read-out. Method specific effects were also noted. Some of these effects on a single sample were further investigated by ECAT, others will also be follow-up to fully define the differences between the methods which are clearly present. It was identified that ultracentifugation of a plasma highly reduced thrombin generation in some methods. This was not corrected by exogenous lipids, but only by reconstitution with microparticles (MPs). MPs are a major determinant of some methods. It was observed that increase in alpha-2- M had a strong effect on some read-outs of some methods and further investigation is carried out on this observation. It was observed that argatroban (a direct thrombin inhibitor) reduced the read-out on thrombin activity, but did not properly identified what happened with prothrombin conversion when compared with F1+2 generation. It is recommended that each method should be considered as different. This will provide opportunities for specific applications, with refinement by selecting specific read-outs.

Thrombin generation induced by cancer cells. G Gerotziafas (FR) on behalf of GT Gerotziafas, C Prengel, E Verdy, I Elalamy, J-F Bernaudin

Several lines of evidence show that thrombotic risk is different in patients suffering from different histological types of cancer. Experimental studies have shown that cells from some histological types of cancer express tissue factor (TF) which is implicated to their metastatic and angiogenetic potential. However, the influence of cancer cells on blood coagulation has not been adequately studied. The procoagulant potential of pancreatic and breast cancer cells (BXPC3 and MCF7 cell lines respectively) when they are in contact with human platelet-poor plasma (PPP) were evaluated. In addition the procoagulant activity of cancer cells using a specific anti-TF antibody was titrated. The contact of cancer cells with recalcified PPP resulted in acceleration of TG as compared to the control. This effect was manifested by a significant decrease of the lag-time, and time to Peak of thrombin (ttPeak) and by significant increase of the mean rate index (MRI) of the propagation phase of TG as compared to the control experiment. Cancer cells induced a slight increase of thrombins’ Peak but they did not significantly influence the endogenous thrombin potential (ETP). Both cell lines when issued from cultures with 40% confluence showed higher procoagulant activity as compared to that manifested by cells from cultures with 90% confluence. BXPC3 had significantly more potent procoagulant activity compared to MCF7 cells. BXPC3 manifested maximum procoagulant activity at the concentration of 2 cells/ml whereas MCF7 manifested maximum effect at the concentration of 200 cells/ml. The incubation of cancer cells with an anti-TF antibody resulted in a concentration dependent inhibition of their procoagulant effect mainly on the lag-time of TG. Significantly higher concentration of the anti-TF antibody was required for 50% inhibition of the effect of BXPC3 on thrombin generation as compared to that required for 50% inhibition of MCF7 procoagulant activity. In conclusion, pancreatic cancer cells (BXPC3) and breast cancer cells (MCF7) accelerate thrombin generation of human plasma in a TF dependent manner. BXPC3 have a significantly more potent procoagulant activity than MCF7 probably due to increased TF expression. The number of cells suspended in plasma and their proliferative status according to the level of confluence, are important determinants for the procoagulant potential of the studied cancer cell lines. Chronometric parameters of thrombogram (lag-time and ttPeak) and the mean rate index of the propagation phase of TG seem to be more sensitive than ETP and thrombin’s peak to detect the TF dependent procoagulant potential of cancer cells. A specific anti-TF antibody might serve as calibrator for the evaluation of the TF-dependent procoagulant ppotential of cancer cells from different histological types of tumors.

Behaviour of different anticoagulants in thrombin generation tests. M Samama (FR)

Three different patterns of thrombograms measured by the CAT were reported. Type 1 was typical of irreversible thrombin inhibitor, hirudin. Hirudin prolonged lag time, but had no effect on peak and ETP. There was an artifact at low dose which was due to mathematical inadequacy in the software to resolve the thrombin generation curve produced by hirudin. Type 2 was typical of reversible thrombin inhibitors such as dabigatran, argatroban, melagatran. These inhibitors all prolonged lag time, decreased velocity and lower peak thrombin and ETP. Type 3 was typical of danaparoid and fondaparinux. These inhibitors had minor influence on lag time, but dramatic decrease in velocity and lowering of peak thrombin. Dermatan sulphate had almost no effect on lag time but decreased peak thrombin and ETP.

With regards to low molecular weight heparins, the importance of antithrombin activity on the thrombograms has been presented. Full detail on the effects of low molecular weight heparins on thrombogram has been published by Gerotziafas et al (JTH, 2007; 5, 955-962)

Thrombin generation in patients with arterial and venous thrombosis.
H Spronk (NL) on behalf of HMH Spronk, AWJH Dielis, AJ ten Cate – Hoek, M Marchetti, R van Oerle, MH Prins, A Falanga, K Hamulyák and H. ten Cate

Thrombin generation (TG) has been shown useful to detect a hypercoagulable state in individuals at risk of venous thrombosis. In the current study we investigated the applicability of TG by means of the Calibrated Automated Thrombogram (CAT) under conditions of hypercoagubility in plasmas from patients suffering from one of the following conditions: acute myocardial infarction (AMI), deep vein thrombosis (DVT), or a chronic myeloproliferative disorder (MPD, such as essential thrombocytosis (ET)).

For the AMI group, mean age was 62 years, 74% was male (n=100). TG was increased and shortened at 0d (lag time (LT) 0.86, ETP 1.07, peak height (PH) 1.19), but decreased and prolonged at 4d (LT 1.43*, ETP 0.99, PH 0.80*) when all patients received heparin/LMWH. Patients with heparin levels >0.05 U/mL at 4d had prolonged LT (1.43), and decreased ETP (0.81) and PH (0.64) compared to patients with levels below 0.05 U/mL (LT 1.04*, ETP 1.26*, PH 1.51*). Heparin concentration correlated with LT, ETP and PH (R=0.37*, -0.82* and -0.87*). Between 4d-3m, TG changed to levels comparable with 0d (no differences between 3m and 6m).

The DVT cohort consisted of 72 males (46%) and the total group had a mean age of 56.1 years (17.5-82.6), with no difference between males and females. TG slightly increased with age. At all visits lag time, ETP and peak height at 1 pM TF were significantly increased in patients compared to healthy subjects. In patients, TG measured at 1 pM TF changed between V1 and V3 with a shortened lag time (-26.1%), increased ETP (+11.2%) and peak height (+13.7%). TG measured at 5 pM showed the same changes (lag time -20.4%, ETP +9.2%, peak height +13.7%). Healthy subjects showed no changes over time. Addition of TM to the 1 pM TF assay reduced ETP in all patients at all visits (between -8.8% and -19.50%) - but considerably less than in normal pool plasma (-48%).

ET-patients had shorter LT (2.10±0.48 vs 2.39±0.33 min*) and time to peak (4.08±0.73 vs 4.50±0.43 min*) compared to controls for TG with 5 pM TF, while no differences were observed in ETP and PH. Similarly, at 1 pM TF, patients showed shorter LT (3.69±0.88 vs 4.56±0.68 min*) and time to peak (6.5±1.4 vs 8.2±1.3 min*). In addition, PH (279±55 vs 244±62 nM*) and slope (106±36 vs 75±35 nM/min*) were increased in MPD. Among MPD, ET patients had higher ETP (1521±219 vs 1290±369 nM.min*), PH (290±49 vs 245±63 nM*), and steeper slope (112±37 vs 87±32 nM/min*) compared to PV patients.

In conclusions, as compared to healthy individuals, the TG shows alterations in time that may indicate systemic hypercoagulability in patients after AMI, DVT, or with ET. The sensitivity of the assay differs per patient population, such that test conditions may have to be adjusted per indication (eg. for DVT, TG assessed with 1 pM TF is more sensitive than 5 pM TF). Finally, TG is very sensitive to the effects of anticoagulant treatment. Prospective follow up of these patient cohorts should establish the predictive values of TG for recurrent arterial and venous thrombosis.