Anthony R. Hubbard

Inhibition of the tissue factor-factor vii complex: involvement of factor xa and lipoproteins

Inhibition of the procoagulant activity of a tissue factor-Factor VII (TF-FVII) complex by Al(OH)3-adsorbed plasma (AP) was found to require the presence of Factor Xa (FXa). Inhibitory activity seems to be generated through the interaction of FXa with a component in AP rather than with the TF-FVII complex. Quantitation of inhibitor activity was carried out using an amidolytic assay for TF-FVII activity. Incubation of AP with various antisera demonstrated that the inhibition was mainly associated with the presence of apolipoprotein B (apo B) rather than alpha 2-macroglobulin or antithrombin III. Purified lipoprotein-rich fractions prepared from AP, using density gradient ultracentrifugation, all contained some inhibitory activity. Incubation with anti-apo B greatly reduced the inhibitor in the very low density lipoprotein (VLDL)- and low density lipoprotein (LDL)-rich fractions but had essentially no effect on inhibition by the high density lipoprotein (HDL) fraction, which was rich in apo A. The inhibitory activity of AP was 60% that of normal plasma and this correlated well with the relative apo A and apo B concentrations. It is proposed that inhibition requires the interaction of FXa with plasma lipoproteins or associated components and that the product of this interaction is then able to bind to and inhibit the TF-FVII complex.

A survey of one-stage and chromogenic potencies in therapeutic factor VIII concentrates.

Methods-based potency discrepancies in factor VIII (FVIII) concentrates can lead to confusion when the method used for product labelling differs from the method used for regulatory batch release or the testing of post-infusion plasma samples. A previous survey reported substantial discrepancies (‡ 15%) between one-stage and two-stage clotting potencies, in 7 out of 13 concentrates tested (Barrowcliffe et al, 1990). The present survey was undertaken as there have been changes in the products, standards and assay technology over the last 10 years. The use of recombinant products has increased considerably although there is continued use of both intermediateand high-purity plasma-derived products. Recently, the choice of recombinant FVIII concentrates has widened with the introduction of the B-domain-deleted product (Sandberg et al, 2001). The methodology for potency estimation has also evolved with the publication of recommendations for the assay of highpurity FVIII concentrates (Barrowcliffe, 1993) and the replacement of the two-stage clotting method with the recommended chromogenic method (European Pharmacopoeia (EP), 2002). In the present survey we tested several batches of 10 different products licensed for use in the UK using both the one-stage clotting and chromogenic methods (Table I). Seven of the products were plasma-derived (codes 1–7) and three were recombinant (codes 8, 9, 10). One-stage clotting assays were carried out using the Instrumentation Laboratory activated partial thromboplastin time (APTT) reagent (APTT-SP liquid) and artificially depleted FVIIIdeficient plasma containing normal levels of von Willebrand Factor (Organon Teknika Ltd, Cambridge, UK). Chromogenic assays were carried out using the Chromogenix Coatest FVIII:C ⁄ 4 kit. All potencies were estimated relative to the EP Human Coagulation Factor VIII concentrate standard biological reference preparation (batch 1), which consists of intermediate purity FVIII concentrate and has good agreement between the one-stage clotting potency (mean 6Æ30 IU per vial) and the chromogenic potency (mean 6Æ22 IU per vial) (Barrowcliffe, 1996). In all assays both the standard and test concentrates were prediluted in FVIII-deficient plasma (Organon Teknika Ltd) and further diluted in buffer containing human albumin (10 mg per ml), in accordance with the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis (SSC) recommendations (Barrowcliffe, 1993). Potencies between methods differed by £ 7% for 7 out of the 10 products, although there were significant differences (paired t-test; P < 0Æ05) for 3 of these products (coded 2, 4, 7). The largest discrepancies were found with two immunopurified products (coded 5 and 6), manufactured using Method M technology (Addiego et al, 1992), in which the mean one-stage potency exceeded the mean chromogenic potency by 33% and 24% respectively. A large discrepancy was also found with the B-domain-deleted recombinant product (coded 10) in which the mean one-stage potency was 22% lower than the mean chromogenic potency. In comparison with the previous survey (Barrowcliffe et al, 1990), the present survey found generally closer agreement between the different assay methods for the majority of products. This was most striking with the results

Recommendations on the potency labelling of factor VIII and factor IX concentrates

A. R . HUBBARD,* J . DODT ,† T . LEE ,‡ K . MERTENS ,§ R . SE ITZ ,† A. SR IVASTAVA, ¶ M. WEINSTE IN‡ and ON BEHALF OF THE FACTOR VI I I AND FACTOR IX SUBCOMMITTEE OF THE SC IENT I F IC AND STANDARDISAT ION COMMITTEE OF THE INTERNAT IONAL SOCIETY ON THROMBOS IS AND HAEMOSTAS I S *National Institute for Biological Standards and Control, Potters Bar, UK; †Paul-Ehrlich-Institut, Langen, Germany; ‡Center for Biologics Evaluation and Research/Food and Drug Administration, Rockville, MD, USA; §Sanquin Blood Supply Foundation, Amsterdam, the Netherlands; and ¶Christian Medical College, Vellore, India

A multi-centre collaborative study on the potency estimation of ReFacto

Seven laboratories estimated factor VIII coagulant activity in recombinant B-domain-deleted (ReFacto) and plasma-derived FVIII concentrates (Octonativ-M) using chromogenic methods relative to the WHO 6th International Standard FVIII Concentrate (WHO 6th IS), European Pharmacopoeia BRP#2 (EP#2) and the ReFacto Laboratory Standard (RLS). Significantly higher estimates were obtained for all batches of product when calculated relative to the RLS in comparison with estimates vs WHO 6th IS and EP#2. Mean estimates for two batches of ReFacto product vs the RLS were within 10% of the labelled potency whereas estimates vs WHO 6th IS and EP#2 ranged from 21 to 31% lower than the label. Conversely, mean estimates for Octonativ-M relative to WHO 6th IS and EP#2 were within 10% of the label whereas the mean estimate vs RLS was 117% of label. Mean estimates for the ReFacto product, vs the WHO 6th IS and EP#2, varied considerably between the different chromogenic kits whereas estimates vs the RLS showed good agreement between kits. Mean estimates for the RLS vs the WHO 6th IS (8.10 IU/vial) and the EP#2 (7.66 IU/vial) were lower than the assigned value of 9.4 IU/vial. The results are consistent with ReFacto and full-length FVIII responding differently to variations in assay methodology and also indicate that the assigned value on the RLS may be too high. Since this study the unitage on the RLS has been adjusted to effectively increase the amount of ReFacto in the product by 20%.

Activation profiles of factor VIII in concentrates reflect one-stage/chromogenic potency discrepancies

Summary.  We have investigated the possibility that differences in the profile of factor VIII (FVIII) activation, by thrombin, may help to explain the one‐stage/chromogenic potency discrepancies in two therapeutic concentrates. A Method M concentrate and a recombinant B‐domain‐deleted (B‐DD) concentrate were found to have one‐stage/chromogenic ratios of approximately 1·15 and 0·70, respectively, relative to the World Health Organization (WHO) 6th International Standard (IS) FVIII concentrate, whether pre‐diluted in FVIII‐deficient plasma or buffer (± von Willebrand factor, VWF). The activation of FVIII, by thrombin, was followed in a buffer medium (± VWF) and all three concentrates showed similar times to reach peak FVIII coagulation (FVIII:C) activity. However, despite the use of equivalent amounts of FVIII:C for all three concentrates, the B‐DD concentrate reached a higher peak level and maintained higher FVIII:C compared with the WHO 6th IS throughout the incubation period, whereas the Method M concentrate reached a lower peak level and maintained lower FVIII:C throughout the incubation period. We propose that the higher levels of FVIII:C found with the B‐DD concentrate and the lower levels with the Method M concentrate, following activation, may be reflected in the potencies obtained by the chromogenic method and may be consistent with one‐stage/chromogenic ratios of < 1·0 and > 1·0 respectively.

International Normalized Ratio determination using calibrated reference plasmas

We have compared the conventional method of International Normalized Ratio (INR) determination with an alternative method involving extrapolation from a calibration curve using freeze‐dried ‘reference’ plasmas. The latter approach does not require the determination of a mean normal prothrombin time (MNPT) or local system International Sensitivity Index (ISI). Calibration curves were constructed by plotting local prothrombin time (PT) against assigned INR values for a normal plasma and either two plasma pools from patients on oral anticoagulants or two artificially depleted plasmas. Six laboratories determined the INR of a freeze‐dried test plasma and frozen patient plasma samples using the conventional method and by extrapolation. Similarities in the results with the freeze‐dried test plasma and the frozen plasmas were encouraging for the projected use with fresh plasma samples. INR values by the conventional method for the test plasma gave an overall mean of 2.73 and inter‐laboratory variability (gcv%) of 8.92%, whereas estimates by extrapolation against the normal and patient plasmas or the normal and artificially depleted plasmas gave identical overall mean INR values of 2.70 with inter‐laboratory variability (gcv%) of 3.44% and 4.92% respectively. The results indicate that INR determination by extrapolation is associated with reduced inter‐laboratory variability.

A collaborative study to establish the 7th International Standard for Factor VIII Concentrate

Summary.  A candidate concentrate, preparation N (99/678), was assayed and calibrated, as a potential replacement, against four established factor (F) VIII concentrate standards: the current WHO 6th International Standard (IS) (97/616), the previous 5th IS (88/640), the Mega 1 standard and Ph. Eur. BRP Batch 2 standard, in a collaborative study involving 38 laboratories. All laboratories were instructed to use the ISTH/SSC recommendations, including predilution of concentrates in FVIII‐deficient plasma. Several laboratories performed more than one assay method and altogether there were 27 sets of assays with the one‐stage method, 31 with the chromogenic method, and 18 with both methods. There was good agreement between laboratories using each of the two methods for comparison of preparation N against the four established standards, with overall potencies by one‐stage and chromogenic methods differing only by less than 2%. However, there were significant differences in potencies relative to the different standards, ranging from 10.1 IU per ampoule against the Ph. Eur.BRP2 to 11.4 against the WHO 6th IS. Accelerated degradation studies showed that the proposed standard is very stable, with a predicted loss of activity per year of less than 0.001% at the recommended storage temperature of −20 °C. Various options for potency of preparation N were considered by the participants and by members of the ISTH/SSC FVIII/FIX Subcommittee. In November 2003, preparation N (NIBSC 99/678) was proposed to and accepted by the Expert Committee on Biological Standardization of the World Health Organization to be the 7th International Standard for Factor VIII Concentrate with an assigned potency of 11.0 IU per ampoule.

International Collaborative Study for the Calibration of a Proposed International Standard for thromboplastin, human, plain: International standard for thromboplastin

Summary. Background: A preparation of rabbit brain thromboplastin, provisionally coded 04/162, is proposed as a candidate for the World Health Organization (WHO) International Standard (IS) for thromboplastin (rabbit, plain), meant to replace the IS coded RBT/90 (rabbit, plain), stocks of which are now exhausted. Results: The preparation was calibrated in an international collaborative study involving 21 laboratories from 13 countries and the calibration was performed against the existing WHO-IS (i.e. rTF/95 and OBT/79) and other Certified Reference Materials from the Institute for Reference Materials and Measurements of the European Commission (i.e. CRM149 S) and from the European Action on Anticoagulation (i.e. EUTHR-01). An additional candidate rabbit brain thromboplastin coded as 04/106 was also included in the study. On the basis of predefined criteria (the within- and between-laboratory precision of the calibration and the conformity to the calibration model), 04/162 was the preferred candidate. Conclusions: The assigned International Sensitivity Index value was 1.15 and the inter-laboratory SD and coefficient of variation were 0.057% and 4.9%, respectively.

Establishment of the WHO 1st International Standard ADAMTS13, plasma (12/252): communication from the SSC of the ISTH

ADAMTS13 (A Disintegrin And Metalloprotease with ThromboSpondin type 1 motifs 13), also known as ‘von Willebrand factor (VWF) cleaving protease’ is responsible for modulating the size of VWF multimers in the circulation. Acquired or congenital deficiency of ADAMTS13 is associated with the circulation of ultra-large multimers of VWF, which can lead to thrombotic thrombocytopoenic purpura (TTP) characterized by disseminated platelet aggregation and thrombosis in the microcirculation, severe platelet deficiency, red cell hemolysis and organ damage [1,2]. Measurement of ADAMTS13 activity in plasma is an important component in the diagnosis and treatment of TTP and numerous methods, both commercial and ‘in house’, are available for the estimation of ADAMTS13 activity and antigen [3]. However, there is currently no internationally accepted unitage to support harmonization of measurement between laboratories, hence the development of the World Health Organisation 1st International Standard (WHO IS) for ADAMTS13 in plasma. The candidate WHO IS (coded 12/252) was prepared from pooled donations from 38 normal healthy donors and 1-mL aliquots were dispensed into approximately 10 000 glass ampoules prior to freeze-drying and sealing. A collaborative study involving 32 laboratories from 14 countries was undertaken to assign values for activity and antigen to the candidate WHO IS based on assays relative to local pooled normal plasma preparations. The candidate WHO IS was included in the study as coded duplicates (samples A and B). Most laboratories used VWF A2 domain peptide substrate assays either in a Fluorescence Resonance Energy Transfer (FRET) assay (n = 18) or an activity ELISA (n = 9) to measure ADAMTS13 activity and all laboratories used ELISA for antigen measurement. Comparison of the candidate WHO IS with the local normal pools was associated with a high degree of validity in terms of parallelism of the dose-response relationships, with only 9/117 activity assays and 8/58 antigen assays excluded because of non-parallelism. There was no significant difference between the coded duplicates of the candidate WHO IS for estimates of activity or antigen, nor between estimates of activity using the FRET and the activity ELISA methods. Combination of all results for activity gave an overall mean of 0.91 units mL 1 for the candidate WHO IS with inter-laboratory variability (geometric coefficient of variation, GCV) of 12.4%. For estimates of ADAMTS13 antigen the combination of all results gave an overall mean of 0.92 units mL 1 with inter-laboratory variability (GCV) of 16.3%. Two plasma exchange waste bag samples (coded C and D) from a patient with acquired ADAMTS13 deficiency due to an inhibitory autoantibody, treated with plasma exchange, were also included in the study. The level of ADAMTS13 in sample C was below the limit of detection for assays of activity in many cases (21/32) and calculated estimates were only possible in 11 laboratories. However, 31/32 results were consistent with a severe deficiency below 0.1 units mL . Patient sample D contained a higher level of ADAMTS13 and 24/32 datasets from the activity assays could be quantified. The overall mean estiCorrespondence: Anthony R. Hubbard, Haemostasis Section, Biotherapeutics Group, National Institute for Biological Standards and Control, Blanche Lane, South Mimms, Potters Bar, Hertfordshire EN6 3QG, UK. Tel.: +44 1707 641318; fax: +44 1707 641050. E-mail: anthony.hubbard@nibsc.org

International reference standards in coagulation

Measurement of coagulation factor activity using absolute physico-chemical techniques is not possible and estimation therefore relies on comparative bioassay relative to a reference standard with a known or assigned potency. However the inherent variability of locally prepared and calibrated reference standards can give rise to poor agreement between laboratories and methods. Harmonisation of measurement between laboratories at the international level relies on the availability of a common source of calibration for local reference standards and this is provided by the World Health Organization (WHO) International Standards which define the International Unit for the analyte. This article describes the principles, practices and problems of biological standardisation and the development and use of reference standards for assays of coagulation factors, with particular emphasis on WHO International Standards for both concentrates and plasma.