Control of Anticoagulation
June 30 and July 1 2006
Oslo Kongressenter, Norway
Chairman: Sam Schulman, Canada
Co-Chairs: M. Greaves, UK; J. Harenberg, Germany; C. Kearon, Canada; M. Laposata, USA;
J. Olson, USA; G. Palareti, Italy; A.M.H.P. van den Besselaar, The Netherlands
Chairmen: Job Harenberg and Anthon van den Besselaar
S. Schulman opened the meeting of the committee and briefed on the activities ovr the past year
Point-of-Care tests for prothrombin time
W Plesch presented the new CoaguChek XS System, which is being introduced on the market now and demonstrated the confirmation of calibration and results of performance testing. The mean ISI of the thromboplastin is 1.01 (CV was 1.1 – 2.0%), and it is an amperometric method. Calibration was according to WHO recommendation at 4 sites, master lot against IRP, production lots against master lot. Performance evaluation was against Innovin and RecombiPlastin with capillary or venous samples on the test strips, using 33 instruments. Accuracy 95% of data withn +/- 14% of the data. Trueness rel bias –0.4%. Close agreement up to an INR of 8. ISO 17593 assassment was also done, within ISO limits more than 99%. Repeatability was very high. Master lot will last for 1 to 1.5 years.
L Poller demonstrated (on behalf of P Meijer, C Kluft, FJM van der Meer, M Keown, S Ibrahim , AMHP van den Besselaar, A Tripodi, J Jespersen) the feasibility of large scale quality assessment of CoaguChek point-of-care testing prothrombin time monitors. A system for quality assessment (QA) of the CoaguChek “ Point of Care testing” prothrombin time monitor has been developed by the European Concerted Action on Anticoagulation (ECAA). Sets of five certified ECAA plasmas were tested on 539 CoaguChek monitors by experienced staff at 9 Netherlands Thrombosis Centres and results compared with certified INR. A 15% INR deviation has been classified as “significant deviation” and was found with 20.3% of the monitors with significant differences between test strip lots. One single lot was responsible for most of the variation. A conventional type of external quality assessment analysis was also applied with similar findings. The results validate the use of the single instrument QA procedure for CoaguChek users developed by the ECAA. This is the largest EQA study so far on a POC instrument. (Paper accepted in Journal of Clinical Chemistry).
W. Plesch commented that CoaguChek is intended for whole blood and that it is not sure that the 20.3% showing deviation would be seen in that case.
Miguel A. Cortés Vázquez (on behalf of E Gómez , A Hervás, R Valero, P Muñoz) described how they had performed validation of the computer decision support software TaoCheck to monitor oral anticoagulant therapy. Many clinical trials have demonstrated the utility of computer-based dosage programs to monitor oral anticoagulant therapy in outpatients. However some of them are not already validated. Therefore they carried out a prospective, randomized trial to validate the efficacy of the computer decision application TaoCheck. 56 were randomized to TaoCheck aided dosing, 62 to experienced physician dosing. No differences seen between groups re Time in Therapeutic Range (50% per group). Number of appointments was lower in control group.
Factor Xa inhibition and the monitoring thereofFrank Misselwitz (Bayer Health Care) reported on the results of the clinical trials in the prophylaxis and treatment of medical and surgical VTE with a Xa inhibitor. Rivaroxaban (BAY 59-7939) is a novel, direct, oral factor Xa inhibitor, which selectively inhibits factor Xa (Ki 0.4 nM) but not other serine proteases. It is a small molecule which fits directly into the active site of FXa;and has a competitive, reversible mechanism of action. Rivaroxaban dose-dependently prolongs PT and aPTT, HepTest, and factor Xa-inhibition is strictly correlated to its plasma concentration (correlation between PK and PD 0.96 for PT) Rivaroxaban dose-dependently prolongs Prothrombinase-induced clotting time (PICT), and inhibits ETP (reases the lag time and decreases the peak in thrombin generation curves) both via the intrinsic and extrinsic pathways (collagen and tissue factor as stimuli). No need for routine monitoring, but good monitorability using PT with different PT reagents, which all highly correlate to PK but result in different slopes (ISI 1.01 with Innovin, 1.3 with Neoplastin, 2.0 with Simplastin, 2.4-2.69 for Thromboplastin C). The "steepness" of the correlation does not depend on ISI of the PT reagent. There is no effect on TT, Ecarin time or on platelet aggregation. There is not enough data to make a recommendation for a therapeutic PT range. PICT is prolonged with the drug and may also be a candidate for monitoring.
Job Harenberg gave a n update on the clinical laboratory validation of PICT test (Pentapharm, Basel ) in the monitoring and potency evaluation of heparins and pentasaccharides. Three hospitals participated. Normal range is approx 20-34 s.RVV activaes F V. PICT seems to be sensitive to low-dose LMWH. PICT versus ECT in Hirudin monitoring gave an r-value of 0.94. PICT is prolonged by antithrombin-dependent anticoagulants but also by direct Xa and IIa-inhibitors. Vitamin K antagonists do not prolong PICT, and therefore PICT is useful during concomitant therapy. Dr Harenberg proposed to perform a larger standardization study with various anticoagulant agents.
Job Harenberg proposed an international study with collection of anti-F Xa levels from patients on therapeutic dose of fondaparinux. Patients on prophylactic or therapeutic doses will have samples collected at trough and peak. Local analysis with an anti-FXa method and/or Hepest. He aims for 200 patients with low dose and 200 with high dose treatment. A fondaparinux standard has to be used. Information on local routines for the analysis will be collected. F Misselwitz commented that both samples after first dose and after 5 th dose or later is desirable to get steady state. It was decided to partly collect retrospective data, partly to go prospectively and then also save plasma for central analysis.
Soumaya Elrouby ((ITC) reported on monitoring of low molecular weight heparin using whole blood clot based Hemonox™ method in comparison to ACT and anti-Xa assays. Hemonox™ is a new point of care test (POCT) for monitoring the anticoagulant effect of LMWHs. It is a one step whole blood coagulation test that uses the HEMOCHRON ® Jr. Signature+ Whole Blood Microcoagulation System with software version 2.4 or higher. The HEMONOX test has been evaluated in several clinical trials. The test showed a clear distinction between untreated and treated patients. Normal HEMONOX baseline is in the range of 70 seconds, usually below 85 seconds. Normal range is 50-76 s in volunteers, 59-81 in obese donors. A baseline value equal or exceeding 100 seconds indicates prior treatment with LMWH or heparin. Peak HEMONOX results are quickly achieved 10-15 minutes post intravenous (IV) bolus; a clotting time >150 seconds corresponds to therapeutic anti-Factor Xa levels ( >0.5U/ml) in PCI patients. The relation between HEMONOX results and anti-Factor Xa is not affected by the presence of GpIIb-IIIa antagonists. The HEMONOX response profile is patient specific and reproducible. In the non-invasive setting using subcutaneous administration, the peak is 4-6 hours post injection. The test may not exhibit the level of sensitivity to the anticoagulant effects of enoxaparin as observed using IV dosing, however the HEMONOX results may provide the clinicians with a diagnostic tool to interpret the anticoagulant response in ACS patients who progress to PCI. The HEMONOX POCT is more sensitive than the ACT test in measuring the progressive effect of LMWH in ACS and PCI patients. The HEMONOX test may also indicate a hypercoagulable state in some obese patients receiving small doses of LMWH for gastric bypass procedure.
HEMONOX vs LMWH conc gave an R=0.966 for dalteparin. HEMONOX vs anti-Xa was R=0.85-0.90 in PCI patients and 0.87 vs Heptest. Coefficient of variation usually 4.5%.
W. Jeske pointed out differences in the 1st and 2 nd LMWH standards.Soon after the introduction of LMWHs, a standard to cross-reference their potency was developed. The 1st LMWH standard (85/600) was produced by nitrous acid digestion of porcine mucosal heparin and exhibited characteristics comparable to dalteparin. Because of the widespread use of the first standard, a new standard material has been introduced, which is also produced by nitrous acid digestion but whose characteristics were different. The MW profile and the biologic activities, including serpin affinity, of these two standards are distinct. As many generic versions of LMWHs are currently being introduced, a systematic study was carried out to cross-reference several commercially available generic products from India and South America against the first and second LMWH standards using amidolytic anti-Xa and anti-IIa assays. Differences were observed in the slopes of the concentration-response curves for these standards in both the anti-Xa and anti-IIa assays. The potencies of the generic LMWHs ranged from 76 to 119 U/mg in relation to the first standard and from 88 to 122 U/mg in relation to the 2 nd standard. On average, the anti-Xa potency was 8% higher when calculated with the 2 nd standard. The differences in anti-IIa potency were further amplified and exhibited wider variation (range 15-31 U/mg with the 1st standard and 19-41 U/mg with the 2 nd standard). On average, the anti-IIa potency was 28% higher when calculated with the 2 nd standard. When used to cross-reference commercially available LMWHs such as enoxaparin, the new standard consistently overestimates the anti-Xa potency by 5-10% in comparison to the first standard. Such an overestimation may lead to a reduction in dosage and have a significant impact on the clinical dosing of LMWH in various indications. W Jeske suggested that if the 2 nd standard is used, it should first be cross-referenced against the first standard in order to properly assign the anti-Xa and anti-IIa activities. He also called for developing individual standards for each of the branded products. Such standards should be defined in terms of not only anti-Xa and anti-IIa activity, but also in terms of anticoagulant activity using the USP assay, activated clotting time (ACT), aPTT and Heptest.
Jawed Fareed pointed out that several generic versions of the LMWHs enoxaparin and dalteparin have been introduced in some Asian and South American countries. In addition, numerous generic suppliers have applied for the approval to sell the generic versions of enoxaparin and dalteparin in the US and European union. Neither the US FDA and EMEA have any guidelines for the generic interchange of branded products. As a result, several substandard products have been introduced and withdrawn. The current pharmacopoeial guidelines are inadequate to accept the generic version of the branded products as these apply the older guidelines that are applicable to the conventional generic drugs. The LMWHs represent a hybrid of the biologic and chemical manufacturing processes. Moreover, these are derived from unfractionated heparin of porcine origin. There are not even proper guidelines to control the raw material, porcine mucosal heparin which is the starting material for the manufacturing of LMWHs. The European Pharmacopeial description of each of the individual LMWHs is incomplete and US Pharmacopeial doesn’t even have a description statement regarding these LMWHs. Few of the professional societies have made any comments on this issue. As the ISTH/SSC has played a role in the biologic standardization and characterization of LMWHs, Dr Fareed proposed that ISTH/SSC on the Control on Anticoagulation, consider the development of specific guidelines for the requirements to accept a generic version of the branded product. The ISTH membership should be highly qualified to undertake this mission. Such guidelines should be based on the newly available analytical methods and characterization of the biological effects of heparins.
Discussion
E. Gray commented that in other studies that she had performed the comparability in e.g. a large cohort study she did not see that kind of comparability, except for one single preparation. Dr Harenberg thought that it might be necessary to have a local correction factor for the individual lab when switching to the second standard. Dr Barrowcliffe informed that the generic LMWHs were not available when NIBSC performed the study. It was therefore decided to continue the discussion afterwards whether to have a working group on the issue of how to assess the generic LMWHs and if the 2 nd standard is useful.
Joint session with Subcommitte on Plasma Coagulation Inhibitors
The minutes from the following part will appear in the minutes of Subcommitte on Plasma Coagulation Inhibitors
Report on the international collaborative study on fluorogenic methods for Thrombin generation test. E. Gray, A. Lawrie
Pre-clinical validation of the Calibrated Automated Thrombogram using platelet rich plasma and the effect of thrombomodulin. H. Spronk
International multi-centre assessment of the calibrated automated thrombogram thrombin generation assay. Y. Dargaud
A new global assay with small amounts of recombinant tissue factor and tissue- plasminogen activator providing novel parameters to determine the overall hemostatic potential. S. He, M. Blombäck
Methods recording dynamics of fibrin formation. B. Sørensen, J. Ingerslev
Discussion
Sam Schulman and Anna Falanga launched an international registry on recurrent VTE in patients with cancer. The registry will thus be under both the committee of Malignancy and Hemostasis and the committee of Control of Anticoagulation. It is anticipated that 200 cases will be registered over 2 years and thus results reported in Vienna at the 54 th Annual SSC Meeting 2008. The objective is to register current regimens used for these problematic patients and to get a rough estimate of the efficacy and safety over 3 months of follow-up. The data collected may be used for guidelines with a low level of evidence and also form the basis for the design of appropriate trials in this field. The registry is supported by an unrestricted educational grant from Leo and deposited at ISTH. A symbolic honorarium is paid for each complete case submitted. Information on the registry, case report forms, instructions and a template for submission to the IRB/REB/ethics committee can now be found under REGISTRIES at the ISTH website.
1 July, 2006
Chairman: Clive Kearon
H.C. Hemker ( Synapse BV ) discussed the need for continuous individual calibration in fluorogenic measurement of thrombin generation .
Fluorogenic thrombin generation (TG) measurements are based on the thrombin-catalyzed production of a fluorescent substance. From the velocity of increase of the fluorescence signal (dF/dt) the concentration of thrombin has to be calculated. Traditionally this is done via measuring the dF/dt of a known amount of thrombin. If a concentration [T] of thrombin is added and a velocity of V is measured then the calibration factor (Cf) equals T/ V (Cf = T/V). Under the conditions of a thrombin generation experiment Cf is, however, not a constant. At the end of the experiment it may be half of that at the beginning. (due to: substrate consumption and non-linearity of the fluorescent signal with the concentration of fluorescent product). Also Cf differs considerably between plasmas. In icteric- or haemolytic plasma it may be half of that in normal plasma. For this reason and also because the error brought about by non-linearity is more important at high thrombin generation than at low (more substrate split, more fluorophore produced), standardization via normal plasma does not work. The only possible solution is continuous individual determination of Cf in a parallel sample. This technique is used in the Thrombinoscope method but not in the alternative Technoclone method. We therefore compared the two methods. From six volunteers we drew 12 samples in the course of one day and determined the six intra-individual coefficients of variation. In the table the CVs are rendered of the Thrombinoscope method with Continuous Individual Calibration (CIC) ON and OFF and of the Technoclone in the two forms in which it is presently marketed. With CIC-ON, the CV of peak-height and initial slope were 4% and 8% resp. In the three methods without CIC the CV varied between 20% and 38%. (ETP cannot be calculated in the Technoclone method).
The presentation was followed by criticism that other manufacturers of kits for measurement of thrombin generation were not invited. S.Schulman responded that the subcommittee was approached by Prof Hemker with data that were found to be pertinent for presentation at the subcommittee but we neither have resources or sufficient time in the program to invite all competitors to present.
A. van Hylckama Vlieg had been invited to present studies on the predictive value of ETP measurement for VTE; first episode and recurrence. However, she had sent apologies, that she was unable to attend.
CliveKearon introduced a session on the possible links between arterial and venous thromboembolism. He mentioned the acute situation when pulmonary embolism with increased cardiac load can unmask ischemic heart disease.
Giancarlo Agnelli Gave the first presentation on “Venous thromboembolism and atherosclerotic disease – common denominators or different diseases?” After reviewing the literature, he went into detail with the study of Prandoni et al (NEJM 2003) showing an association between VTE and carotid plaques. Agnelli’s group has shown a reduced endothelial function in patients with idiopathic DVT. Becattini in his group found increase in the rate of MI and stroke in the long-term course after unprovoked pulmonary embolism. Prandoni also showed in a ohort of almost 2000 patients that those with unprovoked VTE had a HR of 1.9 (adjusted 1.6; 1.2-2.0) for cardiovascular events. He also mentioned Young’s study that residual thrombosis on ultrasound is associated with worse survival, but this is partly due to concomitant cancer.
Sam Schulman gave the second presentation on this topic and also presented some data from the 10-year follow-up of the DURAC 1-trial. Here it was shown that patients with venous thromboembolism have a higher death rate than expected, both from all cause deaths, from cancer and from myocardial infarction/ischemic stroke. This pattern is more pronounced in patients with unprovoked venous thromboembolism and perhaps also in those with proximal DVT or pulmonary embolism (versus distal DVT). A common denominator is difficult to identify from this trial, although cardiolipin antibodies may play a role.
A.M.H.P. van den Besselaar gave a progress report on the revision of the 1999 WHO Guidelines for thromboplastins and plasma to control oral anticoagulant therapy. Several new developments in the control of oral anticoagulant therapy occurred after the publication of the 1999 WHO Guidelines. These should be addressed in the revision of the document.
According to the 1999 WHO Guidelines, it is recommended that patient’ samples with INR values in the range 1.5 – 4.5 should be selected. Samples with INRs outside the 1.5 – 4.5 range shall be excluded. The Guidelines include an example in which the exclusion of samples is based on measurements with the reference thromboplastin (in this case RBT/90).
The above procedure of sample exclusion can lead to a bias in the slope of the calibration line from which the ISI is calculated. Planimetric considerations predict an underestimation of the slope of the line if the exclusion of the samples is performed with one of the two thromboplastin systems only. An alternative procedure for sample exclusion is proposed which avoids the bias in the slope. The alternative procedure takes into account the measurements performed with both thromboplastins. The data of three multicenter calibration studies have been used to estimate the magnitude of the bias induced by the current WHO recommended exclusion procedure. The average bias in one calibration step is 0.5%. This is not of clinical importance but it should be realized that the bias is cumulated in each subsequent calibration step. If the average bias in one step is 0.5%, the total bias after 10 steps may be as high as 5%. The bias should be avoided by using both the reference and the new thromboplastin in the calculations.
In the discussion the issue was brought up whether for a new reference thromboplastin a correction could be made for the error incurred by the type of calculation used in the past. It was agreed that the results should be published as a report from the Subcommittee.
A.M.H.P. van den Besselaar (also on behalf of A. Tripodi) discussed the replacement – or discontinuation – of bovine reference thromboplastin. The second international reference preparation of thromboplastin, bovine, combined (coded OBT/79) was established in 1983 with an assigned ISI of 1.0. OBT/79 was derived from bovine brain and combined with factor V and fibrinogen and should be used to calibrate thromboplastin materials of bovine origin and combined thromboplastins of whatever origin. This recommendation was based on the experience that the calibration of a given thromboplastin is more precise if performed against an international reference preparation of similar composition and from the same species.
The stocks of OBT/79 have been exhausted recently. The question arises whether OBT/79 should be replaced by a similar material. In an ideal world with unlimited resources, the answer would be: YES. It should be realized that worldwide, human and rabbit-based thromboplastins are used on a much larger scale than bovine reagents. As far as the discussants know, only two bovine thromboplastin reagents are prepared commercially. Apart from the question whether the replacement is feasible and affordable, Dr van den Besselaar investigated if bovine thromboplastin may be calibrated against a rabbit thromboplastin with an acceptable level of precision. Two commercial thromboplastins were calibrated by their laboratory using both OBT/79 and RBT/90 (international reference preparation for thromboplastin, rabbit, plain). There was good agreement between the ISI obtained with the two reference prepations. The CV of the calibration slope with OBT/79 was 0.9 and 1.2% for the two commercial reagents, respectively. The CV of the slope with RBT/90 was 2.0 and 2.0%, respectively. Although the CV with RBT/90 is greater than with OBT/79, it is still lower than the CV which is considered by the WHO as the upper limit of acceptable imprecision, i.e. 3%. It was concluded that bovine thromboplastins can be calibrated against rabbit reference thromboplastin with acceptable precision. The calibration against OBT/79 is not needed if a CV of 3% or less can be obtained with a rabbit reference thromboplastin. There is not an absolute requirement to replace OBT/79.
Armando Tripodi (also on behalf ofA.M.H.P. van den Besselaar) proposed to start the process which hopefully will be leading to the replacement of rTF/95 (the human WHO International Standard for thromboplastin). The stocks of the latter will be depleted in a few years. As in the past he proposed to ask manufacturers to submit candidates by December 2006. A maximum of three candidates will be provisionally accepted by the Subcommittee if they fulfil the requirements set in advance. The candidates could be recombinant, placenta or TF derived from cultured endothelial cells (but not brain) and the ISI should be 0.90 – 1.3 with manual technique. The three candidates will then be calibrated in an international collaborative study (involving 20 labs) against rTF/95 and RBT/05, and then the most suitable candidate will be chosen on the basis of pre-defined criteria. The entire process will end on the occasion of the SSC Meeting in 2008 where the new standard will be approved by the Subcommittee and will be submitted to the WHO Expert Committee on Biological Standardization.
A question was raised by W Plesch if we could have a back-up reference to avoid running out of material and then having only a singl reference preparation. T Barrowcliffe responded that this is a luxury that we don’t have for other references. It may also be difficult to have that approved by WHO.
There were no objections to the proposal by A Tripodi and it was approved to go ahead.
Leon Poller thereafter argued for the importance of the continuation of the WHO thromboplastin IRP (combined). The WHO Prothrombin Time (PT) Standardisation Scheme is based on calibration of local PT test systems against the relevant species reference thromboplastin (human, rabbit or bovine). The provision of the replacement bovine IRP to WHO is necessary for various reasons:
The WHO thromboplastin IRP bovine combined should therefore be replaced to maintain the present hierachical structure of the IRP; to provide more reliable IRP stability monitoring; to provide a “combined” IRP for ISI calibration of “combined” thromboplastins; to provide more precise ISI calibration with bovine reagents and to remove species specificity problems in PT testing.
Discussions
A bovine thromboplastin will most likely not be approved by the WHO due to the BSE risk. It was therefore concluded that the main issue to consider and preferably present at the next SSC Annual meeting is a new “combined” thromboplastin, presumably of rabbit origin. Drs van den Besselaar and Tripodi will consider this.
Sam Schulman, chair.