EXPERT COMMITTEE ON BIOLOGICAL STANDARDIZATION Geneva, 15 to 19 October 2012

Transcription

1 ENGLISH ONLY EXPERT COMMITTEE ON BIOLOGICAL STANDARDIZATION Geneva, 15 to 19 October 2012 Value Assignment to the WHO 2 nd International Standard for Fibrinogen Concentrate (09/242) Sanj Raut, Sarah Daniels, Michelle Hamill* and Alan Heath* Haemostasis Section and Biostatistics Section*, National Institute for Biological Standards and Control, Potters Bar, Hertfordshire, EN6 3QG, UK Principal investigator and coordinator of project This document has been prepared for the purpose of inviting comments and suggestions on the proposals contained therein, which will then be considered by the Expert Committee on Biological Standardization (ECBS). Comments MUST be received by 01 October 2012 and should be submitted electronically to the Responsible Officer: Dr Ana Padilla at padillaa@who.int, with a copy to Dr David Wood at woodd@who.int. World Health Organization 2012 All rights reserved. Publications of the World Health Organization are available on the WHO web site ( or can be purchased from WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: ; fax: ; bookorders@who.int). Requests for permission to reproduce or translate WHO publications whether for sale or for noncommercial distribution should be addressed to WHO Press through the WHO web site ( The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. All reasonable precautions have been taken by the World Health Organization to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall the World Health Organization be liable for damages arising from its use. The named authors alone are responsible for the views expressed in this publication.

2 Page 2 Summary Two candidate materials, (samples B and C), have been evaluated as potential replacements of the WHO 1 st International Standard (IS) for Fibrinogen Concentrate, by assays relative to the current primary WHO 1 st IS for Fibrinogen Concentrate (98/614), in international collaborative studies involving 27 laboratories. All laboratories were instructed to use their routine methods for assessment of both Clottable and Total Protein. For Clottable Protein estimates, laboratories used either Clauss assays or subsequent protein determination assay methods following clot removal (CLOTr) e.g. Kjeldahl, Biuret, absorbance assays. Several laboratories performed more than one assay method. ELISAs were also carried to determine Clottable Protein estimates, however, data from this method were not included in the combined overall means as they do not represent fibrinogen function. For Total Protein estimates, laboratories used protein assays such as colorimetric assays e.g. Bicinchonic Acid (BCA) protein assay, Biuret assay, BioRad protein assay or N 2 determination - Kjeldahl assay or absorbance assays. Intra-laboratory variability, for both candidates, as measured by geometric coefficient of variation (GCV), was found to range from % for Clottable Protein and from % for Total Protein, with GCVs from the majority of the labs below 8%. For Clottable Protein estimates, there was good agreement in potencies between laboratories using the CLOTr methods when assayed relative to the WHO 1 st IS, with the lowest interlaboratory variability for candidate sample B, where a GCV value of 3.6% and a mean potency of mg/ml (n=10) was obtained; whilst for sample C, a GCV value of 6.0% and a mean potency of mg/ml by CLOTr (n=5) was obtained. There was poorer agreement in potencies between laboratories using the Clauss assays when assayed relative to the WHO 1 st IS, with a GCV value of 17.2% and mean potency of mg/ml (n=24) for sample B; and a GCV value of 26.8% and mean potency of mg/ml (n=17) for sample C. A combined mean potency value of mg/ml was obtained for sample B (GCV=16.7%; n=34), whilst a combined mean potency value of mg/ml was obtained for sample C (n=22; GCV=38.9%). Furthermore, there was a significant difference between Clauss assays and CLOTr methods, with the Clauss assay giving 18% (p< ) & 81% (p<0.001) higher potencies for samples B and C respectively, compared to the CLOTr method. For Total Protein estimates, there was good agreement in potencies between laboratories for both candidates, with a GCV value of 4.4% and a mean potency of mg/ml (n=9) for sample B and a GCV value of 7.7% and a mean potency of mg/ml (n=10) for sample C. Proposal Overall, it is proposed that sample B is the favoured candidate for the following reasons: - Better agreement in mean values obtained for Clottable Protein (Clauss assay and CLOTr method) and for Total Protein content for sample B compared to sample C. - Lowest overall inter-laboratory variability for combined estimates (GCV=16.7% Clottable Protein; GCV=4.4% Total Protein) for sample B compared to sample C (GCV=38.9% Clottable Protein; GCV=7.7% Total Protein).

3 Page 3 However, due to the large discrepancy observed between the methods for Clottable Protein estimates, it would mean that combination of method potencies for value assignment for Clottable protein would cause a large discontinuity in the International Unit, which is unacceptable, particularly in relation to product labeling and testing by regulatory authorities. Furthermore, the high variability observed when using the Clauss assay for sample B (GCV=17.2%) compared to the CLOTr method (GCV=3.6%) makes the data from Clauss assay unsuitable for value assignment to the proposed 2 nd IS. It is therefore proposed that candidate B (NIBSC code 09/242) be accepted as the WHO 2 nd International Standard for Fibrinogen Concentrate with a Clottable Protein potency of 10.9 mg/ampoule (based on the CLOTr method) and a Total Protein potency of 15.0 mg/ampoule. Introduction The current World Health Organization (WHO) 1 st International Standard (IS) for Fibrinogen Concentrate (98/614) (Sands et al, 2000) is used in potency estimation of fibrinogen in plasmaderived therapeutic concentrates, which are primarily used in the treatment of bleeding in patients with congenital or acquired fibrinogen deficiencies (Fenger-Eriksen et al, 2008; Farriols Danés et al, 2008), or are components of Fibrin Sealants routinely used as haemostatic agents (Spotnitz et al, 2008; Achneck et al, 2010). Stocks of this standard are running low and are expected to be exhausted by end of The National Institute for Biological Standards and Control, NIBSC (a WHO Collaborating Centre) has undertaken a project to replace this standard. Following trial-fill studies on a number of therapeutic fibrinogen concentrates, two candidate materials were selected based on fill characteristics and stability of fibrinogen potency. Ten thousand (10,000) ampoules of each of the 2 candidate materials have been prepared. This project was reviewed and endorsed by the WHO Expert Committee on Biological Standardisation (ECBS) in October 2009 and by the ISTH/SSC Sub-Committee on Factor XIII and Fibrinogen at the 55 th SSC meeting of the ISTH in July 2009 (Boston). Objectives of the study To assign values for Clottable Protein and Total protein to the replacement for the WHO 1 st IS for Fibrinogen concentrate. Participants Twenty seven (27) laboratories (from 12 countries) participated in the value assignment exercise and returned data for analysis. They are listed in Appendix I. The participants included 14 manufacturers, 11 academic institutes and 2 national control authorities. Laboratories were coded for the study and the order of listing in Appendix I does not necessarily correspond with the numerical codes. All raw data returned by the participants were analysed at NIBSC. Materials Proposed 2 nd IS Candidate, Sample B (09/242): the raw material for this candidate was a plasma-derived fibrinogen concentrate, prepared using precipitation and chromatography methods. All donations used to prepare this product were tested and found negative for HBsAg, anti-hiv-1 and -2, anti-hcv, HCV-RNA (plasma pools). Manufacturing of this product also included 2 viral inactivation steps, solvent detergent treatment and heat treatment at 80 C for 72

4 Page 4 hours. After reconstitution of the product, the concentrate material was pooled and formulated in the following buffer: 40mM Tri-Sodium Citrate.2H 2 O, 20mM Tris-HCl, 3% sucrose (w/v), 4mM L-arginine-HCl (ph 7.3). The formulated material was filled and freeze-dried in sealed glass ampoules at NIBSC, under conditions required for International Standards (Campbell, 1974). One ml of this material was dispensed into each of approximately 10,000 ampoules. The mean filling weight was g (range g to g) and the coefficient of variation (CV) was 0.27% based on 468 check-weight samples. Mean residual moisture after freeze-drying was 0.39% (CV 19.3%, n=12) and mean oxygen headspace was 0.11% (CV 55.2%, n=12). Secondary Candidate, Sample C (10/100): the raw material for this candidate was a plasma-derived fibrinogen concentrate bulk. All donations used to prepare this product were tested and found negative for HBsAg, anti-hiv-1 and -2, anti-hcv, HCV-RNA (plasma pools). The bulk concentrate powder was dissolved in the following formulation: 90mM NaCl, 20mM Tri-Sodium Citrate.2H 2 O, 67mM Glycine, 83mM Glucose (ph 7.0). The formulated material was filled and freeze-dried in sealed glass ampoules at NIBSC, under conditions required for International Standards (Campbell, 1974). One ml of this material was dispensed into each of approximately 10,000 ampoules. The mean filling weight was g (range g to g) and the coefficient of variation (CV) was 0.20% based on 489 check-weight samples. Mean residual moisture after freeze-drying was 1.22% (CV 18.1%, n=6) and mean oxygen headspace was 0.38% (CV 32.1%, n=12). Sample 1 st IS - WHO 1 st IS for Fibrinogen Concentrate (98/614): This primary standard was supplied by NIBSC and has the following assigned potencies: Clottable Protein: Total Protein: 10.4 mg/ampoule 15.0 mg/ampoule Materials dispatched for study Ten ampoules of each of the above materials were dispatched by NIBSC. Participants were asked not to use any additional test materials in the assays, such as any internal (in-house) standards. Assay Methods and Study Design Participants were requested to carry out 4 independent assays for each analyte/potency parameter (i.e. 4 assays for Clottable Protein determination and 4 assays for Total Protein determination), using their normal routine methodologies, e.g. Clauss assay or clot removal method (CLOTr - where subsequent protein determination assay methods were carried out following clot removal e.g. Kjeldahl N 2, Biuret, absorbance assays) for Clottable Protein determination or protein assays for Total Protein determination e.g. colorimetric, absorbance or Kjeldahl N 2 assays), preferably assays by more than one method. They were requested to carry out assays relative to the current WHO 1 st IS Fibrinogen Concentrate (98/614), on separate days using fresh ampoules of samples B, C and 1 st IS in each assay, and according to a balanced assay design recommended in the study protocol (Appendix II) for those laboratories carrying out Clauss assays for Clottable Protein determination. A separate ampoule of each material was provided for each assay. Laboratories that used more than one method were requested to use material from the same ampoules for each method, provided this could be done within 2 hours of reconstitution. Instructions and assay design were provided in the study protocol (Appendix II). The details of the methods, instruments and reagents used by the participants are listed in Appendix III.

5 Page 5 Statistical Analysis All assays were analysed as multiple parallel line bioassays comparing log response to log concentration (Finney, 1978). All assay data were plotted, and in some cases individual data points were excluded from extreme ends of the dose-response curve to improve the fit of the parallel line model to the data. The statistical validity of the assays was assessed by the usual ANOVA tests for parallelism and linearity and by visual inspection of the plotted dose-response curves. For each assay the potency estimates of samples B & C were calculated relative to the concurrently tested 1 st IS Fibrinogen Concentrate (98/614). Combined potency estimates for each laboratory were obtained by taking unweighted geometric means of results from all assays. Overall combined estimates were obtained by taking unweighted geometric means of the mean results from the different laboratories. Where a laboratory performed more than one assay method, the results for each method were analysed as if from separate laboratories. Intra- and inter-laboratory variability is expressed as the percentage geometric coefficient of variation (%GCV) (Kirkwood, 1979). The mean potency estimates calculated by the participating laboratories are also presented (Appendix IV). Grubbs test (Grubbs, 1969) was used to assess any outlying laboratory mean potency estimates. For the laboratories measuring clottable protein, differences in potency estimates between assay methods (Clauss assay or CLOTr methods) were assessed by 2-sample student s t-tests (Altman, 1991) (where the distribution of potency estimates appeared to be normally distributed) or by Mann-Whitney tests (Altman, 1991). Results Data received Clottable Protein: Results were received from 26 laboratories and comprised a total of 34 data sets for sample B (25 data sets from Main Study; 10 data sets from Field Study) & 22 data sets (Main Study) for sample C. Laboratories used either a Clauss assay (24 data sets for sample B; 17 data sets for sample C), or a CLOTr method (10 data sets for sample B; 5 data sets for sample C). In addition 2 laboratories submitted data from ELISAs (2 data sets) for samples B and C. Several laboratories performed more than one assay method (laboratories 3, 10, 14, 20 & 26) or more than one variation of the same method (laboratories 3, 7, & 18). Laboratories 7 and 18 each performed two different Clauss assays (see Appendix III), with different kits and different thrombin concentrations respectively. They were analysed separately and coded: 7A (Multifibrin U), 7B (rfibrinogen in-house assay); 18A (STA-Fib 2) and 18B (STA-Fibrinogen). Laboratory 3 performed two different versions of the CLOTr method (see Appendix III). Total Protein: Data were received from 9 laboratories with a total of 10 datasets for both candidates (see Appendix III). Laboratories used protein assays such as colorimetric assays e.g. Bicinchonic Acid (BCA) protein assay (1 data set), Biuret assay (4 data sets), BioRad protein assay (1 data set), N 2 determination - Kjeldahl assay (2 data sets) or absorbance assays (2 data

6 Page 6 sets). Laboratory 26 provided two different methods and were analysed separately as 26A (Kjeldahl) and 26B (Biuret). Assay Validity All assays were found to be valid for linearity and parallelism except the exclusions detailed below. For laboratory 9, clotting times appeared to be very low for all dilutions and so the data were excluded. Data for laboratory 19 showed a lack of dose-response and the data were excluded from subsequent analyses. Some data from individual samples were also excluded. Sample B was excluded from laboratory 7B and sample C from laboratory 23 due to non-parallelism. Sample C was also excluded from laboratory 20A (clottable protein) due to a lack of dose-response. Clottable Protein Estimates Relative to the 1 st IS Fibrinogen Concentrate (98/614) The laboratory geometric mean estimates of potency, the intra-laboratory %GCVs, together with the overall geometric mean and inter-laboratory %GCVs, relative to the current 1 st IS, for the Clauss assays, CLOTr methods & ELISAs are shown in Table 1. The table also shows the combined overall means and the inter-laboratory %GCVs. Data from ELISAs, although analysed in the study, were not included in the combined overall mean values as they do not represent fibrinogen function. The potency estimates are also shown in the form of stacking histograms (Figures 1-4). Each box represents the laboratory mean, expressed as a percentage of the overall geometric mean (excluding ELISA data) calculated from results from Clauss assays and CLOTr methods combined. The boxes are labelled with the laboratory code number. The potency estimates calculated by the participating laboratories are shown in Appendix IV. Proposed 2 nd IS, Sample B: Estimates of Clottable Protein for the proposed WHO 2 nd IS, sample B (relative to the 1 st IS) are shown in Table 1. From the histogram (Figure 1) and Table 1, it can be seen that, there is poor agreement between assay methods. Combination of estimates from all assays (CLOTr & Clauss) gave an inter-laboratory GCV of 16.7%, with an overall combined geometric mean potency of mg/ml (n=34) (excluding ELISA data). The Clauss assays gave higher results compared to the CLOTr methods, with the overall means being mg/ml (GCV=17.2%; n=24) and mg/ml (GCV=3.6%; n=10), for Clauss assays and CLOTr methods, respectively. This difference in potency (by 18%) between Clauss assays and CLOTr methods was found to be statistically significant (p= ). The geometric mean potency for the ELISA data was mg/ml (GCV=15.1%; n=2). No individual laboratories were found to be outliers. Intra-laboratory GCVs ranged from % (n=24) for the Clauss assays and % (n=10) for the CLOTr methods and from % (n=2) for the ELISAs. Sample C: From the histogram (Figure 2) and Table 1, it can be seen that for sample C, when potency is expressed relative to the current 1 st IS, there is even poorer agreement between assay methods. Furthermore, when potency is expressed relative to the current 1 st IS, agreement between laboratories for sample C is poorer than for sample B. The overall inter-laboratory

7 Page 7 GCV is 38.9% with a combined geometric mean potency of mg/ml (n=22) (excluding ELISA data). Once again Clauss assays gave higher potencies compared to CLOTr methods with geometric mean potencies of mg/ml (GCV=26.8%; n=17) and mg/ml (GCV=6.0%; n=5) for Clauss assays and CLOTr methods, respectively. This difference in potency (by 81%) between Clauss assays and CLOTr methods was found to be statistically significant (p=0.001). The geometric mean potency for the ELISA data was mg/ml (GCV=12.7%; n=2). No individual laboratory was found to be an outlier. Intra-laboratory GCVs ranged from % (n=17) for the Clauss assays, % (n=5) for the CLOTr methods and from % (n=2) for the ELISAs. Clottable Protein Potency Estimates for the proposed WHO 2 nd IS (Sample B) using Absolute methods (Kjeldahl and Weighing methods) Three laboratories also carried out absolute measurements [Kjeldahl and Weighing (Ingram, 1952) methods] for Clottable Protein (Table 2). Mean individual laboratory estimates for proposed 2 nd IS (sample B) of mg/ml (intra-laboratory GCV=1.2%) and mg/ml (intra-laboratory GCV=0.3%) were obtained from Laboratories 15 and 30 respectively for the Kjeldahl method and estimate of mg/ml (intra-laboratory GCV=2.9%) obtained from Laboratory 26 for the Weighing method (Ingram, 1952), giving an overall mean estimate of mg/ml (inter-laboratory GCV=5.9%; n=3). Total Protein Estimates Relative to the 1 st IS Fibrinogen Concentrate The laboratory geometric mean estimates of potency, relative to the current 1 st IS, the intralaboratory %GCVs, the overall geometric mean and inter-laboratory %GCVs, are shown in Table 3. The potency estimates are also shown in the form of stacking histograms (Figures 3-4). Each box represents the laboratory mean, expressed as a percentage of the overall geometric mean. The potency estimates calculated by the participating laboratories are shown in Appendix IV. Proposed 2 nd IS, Sample B: From the histogram (Figure 3) and Table 3, it can be seen that there is good agreement between laboratories for Total Protein determination, except for the result from laboratory 26A, which was found to be an outlier using the Grubbs test. The overall inter-laboratory GCV is 4.4% with an overall geometric mean potency of mg/ml (excluding outlier laboratory 26; n=9). Intra-laboratory GCVs ranged from % (n=9) for sample B. Sample C: From the histogram (Figure 4) and Table 3, it can be seen that, when potency is expressed relative to the current 1 st IS, there is slightly poorer agreement between laboratories (although still good) for sample C than for sample B. The overall inter-laboratory GCV is 7.7% and the overall geometric mean potency is mg/ml (n=10). No individual laboratory was found to be an outlier. Intra-laboratory GCVs ranged from % (n=10) for sample C. Stability Studies Accelerated degradation study Investigations to assess the long-term stability of the candidate preparations (samples B & C) were carried out through accelerated degradation studies which allow the prediction of

8 Page 8 degradation rates for ampoules stored at low temperatures (e.g. -20 C) based on the observed loss in samples stored at elevated temperatures (e.g. +4, +20, +37, +45 C) (Kirkwood and Tydeman, 1984). This is an indirect method used to determine rate of loss based on the relationship between reaction rates and temperature given by the Arrhenius equation and where a first order reaction rate is assumed (Kirkwood, 1977). Ampoules of the candidate preparations (samples B & C) were placed into elevated temperature storage in May They were assayed concurrently with samples stored at the usual storage temperature of -20 C. Three laboratories assayed samples that had been stored for 12 months. Each laboratory carried out 3 repeat assays for both Clottable Protein and Total Protein. Results of all assays from individual laboratories were combined, to give the figures shown in Tables 4 & 5, for Clottable Protein and Total Protein respectively, where the residual potencies of samples stored at elevated temperature are expressed as a % relative to the -20 C sample. Clottable Protein: There was no observed drop in potency when sample C was stored for 12 months at +4 C & +20 C relative to the baseline -20 C sample. The Arrhenius model for accelerated degradation was applied to obtain predictions of the expected loss in potency over time at different temperatures. Results in Table 4 show that the sample C is stable for long periods at -20 C with a mean predicted % loss per year of It also appears to be stable for shorter periods at higher temperatures with a predicted % loss per month of 0.005, 0.09 and 1.3 when stored at +4 C, +20 C and +37 C respectively. For candidate sample B, there was no observed loss in potency after storage for 12 months at any of the higher temperatures relative to the baseline -20 C sample. It was therefore not possible to obtain a predicted % loss per year from the data, although the observed potencies show that the preparation is stable for long term storage at -20 C and short periods at +20 C and +37 C. Furthermore, both samples B & C stored at -20 C, +20 C, +37 C and +45 C were also assessed after a 10 month period at NIBSC, using the CLOTr method, where results were obtained as direct OD ratios of the higher temperatures relative to -20 C sample (see Table 4). Mean potency values (relative to the -20 C sample) of 97%, 100% & 101% for sample B and of 100%, 100% & 106% for sample C, were obtained when stored at +20 C, +37 C and +45 C respectively. Total Protein: For candidate samples B & C the residual relative potencies for samples stored at elevated temperatures for 10 months or 12 months differed by no more than 5% from samples stored at -20 C. There was insufficient degradation to allow calculations of predicted loss (see Table 5). Overall, these figures represent very good stability particularly for the candidate sample B, and indicate that sample B would be suitable to serve as a WHO International Standard. However, the accelerated degradation studies should be continued and the stability confirmed with further assessments after 3 and 5 years storage. Furthermore, as with all WHO International Standards, the proposed IS will undergo real-time stability monitoring throughout the life time of the standard. Stability of proposed 2 nd IS (sample B - 09/242) after reconstitution Although the Instructions for Use will recommend that assays are performed as soon as possible after reconstitution it is useful to indicate a suitable period of use. In common with previous WHO Concentrate Standards for blood coagulation factors it is recommended that the standard is transferred, after reconstitution, to a plastic tube. Recommendations for the storage after reconstitution have been limited to the period of storage on melting ice since local ambient

9 Page 9 temperature can vary considerably. The mean Clottable Protein results from three separate tests performed at NIBSC, indicated that 96%, 94% and 104% of the starting potency of Fibrinogen was retained for the freshly reconstituted proposed standard when stored on melting ice in plastic tubes, for 2, 4 and 6 hours, respectively. Based on this, a conservative estimate of a 3 hour maximum storage period is recommended, which is sufficient for numerous assays to be performed. The use of frozen aliquots of the proposed 2 nd IS for the estimation of Fibrinogen potency is not recommended. Effect of Proposed & Alternative Assigned Values on the Clottable Protein Potency Estimates for Fibrinogen Concentrate Products In order to assess the impact on product potency if value assignment of the WHO 2 nd IS was based on combined data from both Clauss assays and CLOTr methods, a field study involving 8 laboratories (consisting of manufacturers and regulators) was carried out where actual products were assayed for Clottable Protein estimates relative to the proposed 2 nd IS (sample B 09/242) and compared to assays relative to the current 1 st IS (98/614), see Appendix V. Results from assays on different Fibrinogen Concentrate products are summarised in Table 6, where data obtained tended to give 2-9% higher potencies for Clauss Assays when assayed relative to the current 1 st IS compared to assays relative to the proposed 2 nd IS sample B (with an assigned potency based on Combined Method Mean (12.1 mg/ml). For the CLOTr method, between 8-14% lower potencies were obtained for assays relative to the current 1 st IS than when compared to assays relative to the proposed 2 nd IS (with an assigned potency based on Combined Method Mean), see Table 6. This represented between 10-23% shift in potency. However, minimal discrepancy (i.e. minimal shift in potency) was obtained for CLOTr method when assayed relative to the proposed 2 nd IS (sample B) with a proposed assigned value of 10.9 mg/ml (based on CLOTr potency) whilst a greater shift in potency is observed with the Clauss Assay (see Table 6). Discussion The current WHO 1 st IS for Fibrinogen Concentrate (98/614) was established by the WHO/ECBS (Expert Committee on Biological Standardization) in November 2000 with an assigned Clottable Protein potency of 10.4 mg/ampoule and assigned Total Protein potency of 15.0 mg/ampoule (Sands et al, 2000). Stocks of this standard are running low and are expected to be depleted by end of As such, international collaborative studies were undertaken by NIBSC to calibrate the replacement of the above standard. The WHO 1 st IS (98/614) was previously calibrated for Clottable Protein potency using predominantly CLOTr methods (although smaller number of data sets from Clauss assays were also included), against laboratories own in-house fibrinogen standards and following their own routine methodologies (Sands et al, 2000). For the sake of harmonisation and continuity of the unitage, assays on replacement candidate materials in this study were similarly carried out using methods including Clauss assays as well as CLOTr methods but against the current WHO 1 st IS for Fibrinogen Concentrate (98/614) (sample 1 st IS). However, data from ELISAs, which measure protein concentration rather than fibrinogen function, were excluded from the overall combined geometric mean values for the Clottable Protein parameter. The Total Protein content for the WHO 1 st IS (98/614) was previously calibrated using a variety of methods (Sands et al, 2000) e.g. optical density/spectrophotometery measurements, protein N 2

10 Page 10 determinations and dye binding assays. The Total Protein content for the replacement candidates was similarly determined. This report presents data from assays of 2 fibrinogen candidate preparations, (samples B and C) with the emphasis primarily on value assignment to the replacement for the WHO 1 st IS for Fibrinogen Concentrate (98/614). Intra-laboratory variability The variability of assays within laboratories differed considerably for both candidates, with GCV s ranging from % for Clottable Protein and from % for Total Protein, with intra-laboratory GCVs from the majority of the labs below 8% (Tables 1 & 3). There was no obvious trend for one method to give better or worse inter-assay variability; lower intralaboratory GCV s were associated with particular laboratories rather than particular methods, indicating that internal quality control procedures within laboratories are probably the most important factor in determining reproducibility of fibrinogen (Clottable & Total Protein) assays, rather than the method used. Inter-laboratory variability Inter-laboratory variability for Clottable and Total Protein estimates for both samples by the different methods is given in Tables 1 & 3. Clottable Protein: Clauss assays vs CLOTr methods For Clottable Protein estimates relative to the 1 st IS, there was very good agreement between laboratories (for both candidate materials) for the CLOTr methods with GCV values of 3.6% and 6.0% for samples B & C respectively. There was much poorer agreement between laboratories for the Clauss assays with higher GCV values of 17.2% and 26.8% for samples B & C respectively. This resulted in high overall GCV values of 16.7% and 38.9% for samples B & C respectively (see Table 1). These results indicate that Clauss assays are highly variable when measuring fibrinogen concentrates and caution should be exercised when doing so, in particular when assigning Clottable Protein potencies for lot release for fibrinogen concentrate products, for which the Clauss assay should at least be validated against a CLOTr method. Total Protein For Total Protein estimates relative to the 1 st IS, there was very good agreement between laboratories for both candidate samples, with overall method GCVs of 4.4% and 7.7% for candidate samples B and C respectively (see Table 3). Potency estimates of candidate materials Mean potency estimates by the two methods (Clottable Protein & Total Protein) for the 2 candidate samples relative to the current WHO 1 st IS for Fibrinogen Concentrate (98/614) are given in Tables 1 & 3. For Total Protein, overall mean potencies of mg/ampoule and mg/ampoule were obtained for samples B and C respectively.

11 Page 11 For Clottable Protein, when assayed relative to the the current WHO 1 st IS for Fibrinogen Concentrate (98/614), the Clauss assay gave a significantly higher potency compared to the CLOTr methods, by 18% and 81%, for samples B & C respectively. In the study to calibrate the current WHO 1 st IS for Fibrinogen Concentrate (98/614) (Sands et al, 2000), there was no significant method discrepancy between Clauss assay and the CLOTr (clot removal) methods. This may be due to the fact that Clauss assays represented only 5 data sets (25% of total data sets) for the overall combined estimate in the calibration of the current WHO 1 st IS for Fibrinogen Concentrate (98/614) (Sands et al, 2000), whereas in this study Clauss assays represent 24 data sets (71% of total data sets) for sample B and 17 datasets (71% of total data sets) for sample C. The higher variability, observed for the Clauss assays in this study, has therefore had a significant influence on the overall GCV values for samples B & C. It was also interesting to note that for sample C questionable results were obtained by Clauss assay with a mean Clauss potency of mg/ml (Table 1). This value was actually greater than the mean Total Protein content of 12.1 mg/ml for sample C (Table 3), whereas the mean CLOTr value for sample C (10.04 mg/ml) (Table 1) was in the expected range. Overall, taking all the data into consideration, it is proposed that sample B is the favoured candidate over sample C, for the following reasons: - Better agreement in mean (and combined) values obtained for Clottable Protein (Clauss assay and CLOTr method) and for Total Protein content for sample B compared to sample C. - Lowest overall inter-laboratory variability for combined estimates (GCV=16.7% Clottable Protein; GCV=4.4% Total Protein) for sample B compared to sample C (GCV=38.9% Clottable Protein; GCV=7.7% Total Protein). - Lower discrepancy between Clauss Assay and the CLOTr method for the Clottable Protein determination for sample B compared to sample C. - Lower mean residual moisture after freeze-drying for sample B (0.39%) compared to sample C (1.22%) indicating the possibility of a more stable candidate long-term. However, the large discrepancy between CLOTr method and Clauss assay would mean that combination of method potencies for value assignment for Clottable protein would cause a large discontinuity in the International Unit, particularly in relation to product labeling and testing by manufacturers and national control laboratories. Case for Assigning Clottable Protein Potency Using the CLOTr method In the interest of continuity it was proposed that the candidate sample B be assigned the consensus mean value from the CLOTr methods for the following reasons: The value assigned to the WHO 2 nd IS should reflect the methods used by manufacturers & national control laboratories, primarily the CLOTr methods. Fibrinogen Concentrate product field study shows minimal shift in potency when assayed relative to the proposed 2 nd IS with a proposed assigned value based on CLOTr method. Data from laboratories using CLOTr method is less variable (GCV < 4%) compared to data from laboratories using Clauss Assay (GCV~17%).

12 Page 12 CLOTr potency value (10.9 mg/ml) is supported by good agreement with additional data obtained from absolute measurements for Clottable Protein (Kjeldahl & Weighing methods: mean potency=10.96 mg/ml). Best continuity for product labelling by CLOTr method (the 1 st IS was value assigned with data from CLOTr methods primarily). Proposal: It is therefore proposed that candidate preparation sample B (NIBSC code 09/242) be accepted as the 2 nd International Standard for Fibrinogen concentrate with assigned potencies of: Clottable Protein mg/ampoule Total Protein mg/ampoule Comments from participants and the ISTH/SSC Factor XIII and Fibrinogen Subcommittee Responses were received from all the participants (27 laboratories), all of whom agreed that the candidate material Sample B (09/242) is suitable to serve as the WHO 2 nd IS Fibrinogen Concentrate. 27/27 laboratories agreed with assignment for Total Protein potency of 15.0 mg/ampoule. 26/27 laboratories agreed with assignment for Clottable Protein potency of 10.9 mg/ampoule. There were the following comments with respect to the objection from one participant, together with response from NIBSC: Comments: We do not agree to change candidate B (09/242) value to 10.9 mg of fibrinogen obtained by using the CLOT method, since most laboratories use the von Clauss method to determine the Clottable protein potency. We consider that the method of the Eur. Ph. is difficult to perform in the laboratory and we don t know if it is possible its automation. Do you know if it is possible to automate In the first place, we should test CLOT method with our products and, in case to change it, to validate the method for them and check its consistence. Response from NIBSC: The laboratory concerned was informed that the variability for the proposed 2 nd IS (sample B - 09/242) was exceptionally high using the Clauss assay (GCVs= 17.2%) compared to the CLOTr method (GCV< 4%) and together with the large discrepancy (~18%) in Clottable Protein potency between the two methods made value assignment to the proposed 2 nd IS (09/242) based on a combined method mean unacceptable. In order to maintain continuity of the fibrinogen unit, value assignment for Clottable Protein potency was therefore proposed based on CLOTr method, the data from which was also supported by absolute methods such as Kjeldahl method. They were informed that the Clauss assays could still be used for labelling of products as long as it was validated against a CLOTr method. They were also reminded that any issues relating to the European Pharmacopoeia should be addressed to the EMA/EDQM. In addition, responses have also been sought from experts associated with the Factor XIII/Fibrinogen sub-committee of the ISTH/SSC. Responses from 8 experts in the field were received and all were in favour of the proposal.

13 Page 13 The proposal to accept candidate sample B (coded 09/242) as the WHO 2 nd IS Fibrinogen Concentrate with assigned potencies of: Clottable Protein mg/ampoule Total Protein mg/ampoule was presented to the WHO-ISTH Standards Liaison Group and the Factor XIII and Fibrinogen Subcommittee at the ISTH/SSC meeting and the recommendation endorsed at the SSC Annual Business Meeting, held in Liverpool UK, on 30 th June References Achneck HE, Sileshi B, Jamiolkowski RM, Albala DM, Shapiro ML, Lawson JH (2010). A comprehensive review of topical hemostatic agents: efficacy and recommendations for use. Ann Surg 251(2);217. Altman DG (1991), Practical Statistics for Medical Research, 1 st Edition, Chapman and Hall. London. Campbell PJ (1974). International biological standards and reference preparations. 1. Preparation and presentation of materials to serve as standards and reference preparations. J Biol Standardisation 2; Farriols Danés A, Gallur Cuenca L, Rodríguez Bueno S, Mendarte Barrenechea L, Bruno Montoro Ronsano J (2008). Efficacy and tolerability of human fibrinogen concentrate administration to patients with acquired fibrinogen deficiency and active or in high-risk severe bleeding. Vox Sang 94; Fenger-Eriksen C, Lindberg-Larsen M, Christensen AQ, Ingerslev J and Sørensen B (2008). Fibrinogen concentrate substitution therapy in patients with massive haemorrhage and low plasma fibrinogen concentrations. BJA 101; Finney DJ (1978). Statistical methods in biological assay. 3rd edition Charles Griffin. London. Grubbs F (1969). Procedures for detecting outlying observations in samples. Technometrics 11; Ingram GIC (1952). The determination of plasma fibrinogen by the clot method. Biochem J 51(5); Kirkwood, TBL (1977). Predicting the stability of biological standards and products. Biometrics 33 (4); Kirkwood TBL (1979). Geometric means and measures of dispersion. Biometrics 35; Kirkwood TBL, Tydeman MS (1984). Design and analysis of accelerated degradation tests for the stability of biological standards II. A flexible computer program for data analysis. J Biol Standardisation 12;

14 Page 14 Sands D, Whitton C, Barrowcliffe TW (2000). Collaborative Study to establish an international standard for fibrinogen concentrate using clottable and total protein methods. WHO/ECBS Report. WHO BS Document BS/ Spotnitz WD, Burks S (2008). Hemostats, sealants, and adhesives: components of the surgical toolbox. Transfusion 48(7):1502. Acknowledgements The contributions of all the participants in the study are gratefully acknowledged (Appendix I). We are grateful to our colleagues in the Standards Division (Paul Jefferson and his team in SPD and Paul Matejtschuk and his team at TDI) NIBSC, for ampouling and processing the candidate and trial preparations and for the dispatch of collaborative study samples to participants. We are grateful to Baxter Healthcare (Austria) and BPL (UK) for their kind donation of materials for the study. We further like to thank the ISTH/SSC Fibrinogen and Factor XIII Subcommittee (current Chair - Professor Hans-Peter Kohler; previous Chair - Professor Moniek de Maat) for their guidance.

15 Page 15 Table 1. Clottable Protein mean laboratory estimates for the proposed WHO 2 nd IS (sample B) and sample C relative to WHO 1 st IS Fibrinogen Concentrate (sample 1 st IS - Clottable Protein: 10.4 mg/ml), together with estimates of intra-laboratory variability (GCV%) for individual laboratories and inter-laboratory variability for method estimates and combined estimates. Lab No. No of Assays (n) Proposed 2 nd IS, Sample B (09/242) [Data from Main & Field Studies] Geometric Mean (mg/ml) Intra-Lab GCV% No of Assays (n) Sample C (10/100) [Data from Main Study] Geometric Mean (mg/ml) Intra-Lab GCV% Clauss Assay Data from Main Study Data from Field Study A B A A A B A A A B C B Geometric Mean = (n=24) Inter-Lab GCV = 17.2% Geometric Mean = (n=17) Inter-Lab GCV = 26.8% CLOTr Method Data from Main Study Data from Field Study 3A A B B C B C Geometric Mean = (n=10) Inter-Lab GCV = 3.6% Geometric Mean = (n=5) Inter-Lab GCV = 6.0% ELISA Data from Main Study 14B B Geometric Mean = (n=2) Inter-Lab GCV = 15.1% Geometric Mean = (n=2) Inter-Lab GCV = 12.7% Combined Geometric Mean = (n=34) Inter-Lab GCV = 16.7% (Excluding ELISA Data) Combined Geometric Mean = (n=22) Inter-Lab GCV = 38.9% (Excluding ELISA Data)

16 Page 16 Table 2. Mean laboratory estimates for the proposed WHO 2 nd IS (sample B) using absolute measurements (Kjeldahl method; Weighing method) for Clottable Protein, together with estimates of intra-laboratory variability (GCV%) for individual laboratories and inter-laboratory variability for overall mean estimates. Lab No. Proposed 2 nd IS, Sample B (09/242) No of Assays Geometric Intra-Lab Absolute Method (n) Mean (mg/ml) GCV% Kjeldahl Weight Kjeldahl Overall Geometric Mean =10.96 (n=3) Inter-Lab GCV = 5.3% Table 3. Total Protein mean laboratory estimates for the proposed WHO 2 nd IS (sample B) and sample C relative to WHO 1 st IS Fibrinogen Concentrate (sample 1 st IS - assigned Total Protein potency: 15.0 mg/ml), together with estimates of intra-laboratory variability (GCV%) for individual laboratories and inter-laboratory variability for overall mean estimates. Lab No. No of Assays (n) Proposed 2 nd IS, Sample B (09/242) No of Assays (n) Sample C (10/100) Geometric Mean (mg/ml) Intra-Lab GCV% Geometric Mean (mg/ml) A B Intra-Lab GCV% Overall Geometric Mean =15.04 (n=9) Inter-Lab GCV = 4.4% (Excluding outlier Lab 26A) Overall Geometric Mean = (n=10) Inter-Lab GCV = 7.7%

17 Page 17 Table 4. Clottable Protein results from the accelerated degradation study on the candidate materials samples B & C for the proposed WHO 2 nd IS for fibrinogen concentrate. Candidate Materials Lab ID Mean residual potencies after storage (% vs -20 C ampoules) and [95% confidence limits] +4 C +20 C +37 C +45 C Mean predicted % loss per year at -20 C A 101 [ ] 99 [ ] 101 [ ] 103 [ ] Sample B (09/242) B C 101 [ ] 99 [ ] 103 [ ] 101 [ ] 103 [ ] 101 [ ] 106 [ ] 101 [ ] No observed loss in potency after storage up to 12 months B* nt 97 [ ] 100 [ ] 101 [ ] - A 97 [ ] 97 [ ] 87 [ ] 62 [ ] Sample C (10/100) B C 98 [ ] 101 [ ] 96 [ ] 99 [ ] 86 [ ] 90 [ ] 60 [ ] % B* nt 100 [ ] 100 [ ] 106 [ ] - tested after storage for 12 months - results are the mean values from 3 independent Clauss assays; * tested after storage for 10 months - results are the mean values from 3 independent CLOTr method; results are based on limited data and further stability studies are ongoing. nt: not tested

18 Page 18 Table 5. Total Protein results from the accelerated degradation study on the candidate materials samples B & C for the proposed WHO 2 nd IS for fibrinogen concentrate. Candidate Materials Lab ID Mean residual potencies after storage (% vs -20 C ampoules) and [95% confidence limits] Mean predicted % loss per year at - 20 C +4 C +20 C +37 C +45 C Sample B (09/242) B C 100 [ ] 104 [ ] 100 [ ] 103 [ ] 100 [ ] 105 [ ] 100 [ ] 101 [ ] No observed loss in potency after storage up to 12 months B* nt 97 [ ] 100 [ ] 102 [ ] - Sample C (10/100) B C 100 [ ] 97 [ ] 99 [ ] 98 [ ] 101 [ ] 98 [ ] 113 [ ] 99 [ ] No observed loss in potency after storage up to 12 months B* nt 98 [ ] 101 [ ] 112 [ ] - tested after storage for 12 months - results are the mean values from 3 independent assays * tested after storage for 10 months - results are the mean values from 3 independent assays Results are based on limited data and further stability studies are ongoing. nt: not tested

19 Page 19 Table 6. Summary table showing the % shift (change) in potency of Fibrinogen Concentrate Products. Comparison of Clottable Protein potencies relative to the current 1 st IS (98/614 - assigned potency 10.4 mg/ml) and the proposed 2 nd IS, sample B (09/242 - with various assigned potency options). % Change in Clottable Protein Potency of Fibrinogen Products when comparing the current WHO 1st IS vs proposed WHO 2nd IS CLOTr Method (4 Laboratories) Clauss Assay (4 Laboratories) Proposed 2nd IS assigned value based on CLOTr Method: 10.9 mg/ml Alternative 2nd IS assigned value based on Combined Method Mean: 12.1 mg/ml Alternative 2nd IS assigned value based on Clauss Assay: 12.9 mg/ml 2 4% 8 14% 17 19% 14 21% 2 9% 2 4%

20 Page 20 Figure 1. Stacking histogram showing Clottable Protein mean laboratory potency estimates (as a % of Overall Mean) for the proposed 2 nd IS sample B relative to the 1 st IS Fibrinogen Concentrate (98/614). Each box has a laboratory code number and represents the laboratory mean, expressed as a percentage of the overall combined geometric mean. Results are from Clauss Assays, CLOTr methods and ELISAs Clauss Assay Number of Laboratories 6 4 3C 3A 7A 5B 5A B B CLOTr ELISA 10C 29 26A 26C 14B B 20B 18B A 27A A 15 26B 18A A Potencies of Sample B vs 1st IS (98/614) as % of Overall Potency

21 Page 21 Figure 2. Stacking histogram showing Clottable Protein mean laboratory potency estimates (as a % of Overall Mean) for sample C relative to the 1 st IS Fibrinogen Concentrate (98/614). Each box has a laboratory code number and represents the laboratory mean, expressed as a percentage of the overall combined geometric mean. Results are from Clauss Assays, CLOTr methods and ELISAs Number of Laboratories B 10B Clauss Assay 7B 7A 6 26B B 14B 26A 18A A Potencies of Sample C vs 1 st IS (98/614) as % of Overall Mean

22 Page 22 Figure 3. Stacking histogram showing Total Protein mean laboratory potency estimates (as a % of Overall Mean) for the proposed 2 nd IS sample B relative to the 1 st IS. Each box has a laboratory code number and represents the laboratory mean, expressed as a percentage of the overall geometric mean. Fibrinogen Concentrate Total Protein 15 Potencies of Sample B relative to 98/614 as % of overall mean (15.04) B A Potencies of Sample B v 98/614

23 Page 23 Figure 4. Stacking histogram showing Total Protein mean laboratory potency estimates (as a % of Overall Mean) for sample C relative to the 1 st IS. Each box has a laboratory code number and represents the laboratory mean, expressed as a percentage of the overall geometric mean. Fibrinogen Concentrate Total Protein 15 Potencies of Sample C relative to 98/614 as % of overall mean (12.10) B A Potencies of Sample C v 98/614

24 Page 24 APPENDIX I PARTICIPATING LABORATORIES Robert Ariëns, Daniel Whalley, LIGHT Laboratories, Leeds University, Leeds, UK. Rossella Bader, Bianchi Bonomi Hemophilia and Thrombosis Center, University of Milan, Milan, ITALY. Nathalie Barat, François Nicham, Diagnostica Stago R&D, Gennevilliers Cedex, FRANCE. Chris Backhouse, Scott Pearson, Viv Sung, Lisa Evans, Shane Norton, Chemistry Section, Bio Products Laboratory, Elstree, UK. Mariona Bono, R&D Reagents, Haemostasis Department, Diagnostic Grifols, Barcelona, SPAIN. Sossio Costanzo, Alessandro Starace, Chemical Laboratory and Biochemical Laboratory, Kedrion, S. Antimo (NA), ITALY*. Terrie Emson, Chris Charlesworth, Lincoln County Hospital, Lincoln, UK. Denise Foulon, Affinity Biologicals Inc., Ancaster, CANADA. Peter Gärtner, Baxter QC, Vienna, AUSTRIA. Nuria Hosta, Marta Gensana, Instituto Grifols, Barcelona, SPAIN.* Kevin Horner, Steve Kitchen, Royal Hallamshire Hospital, Sheffield, UK. Andreas Hunfeld, Renate Nawrot, Paul-Ehrlich Institut, Langen, Germany*. David Keeling, Kampta Sukhu, Oxford Haemophilia and Thrombosis Centre, Oxford, UK. Sarah Kingsland, Ruth Archer, Hector Bolsa, Bio Products Laboratory, Elstree, UK*. Steffen Kistner, Karin Fuchs, Biotest AG, Dreieich, GERMANY*. Wolfgang Korte, Karin Jung, Institut für Klinische Chemie und Hämatologie, St. Gallen, SWITZERLAND. Andrew Lawrie, Haemostasis Research Unit, University College London, London UK. Andrea Lichte, Dieter Koch, Siemens Healthcare Diagnostics Products GmbH, Marburg, GERMANY. Moniek P.M. de Maat, Erasmus University Medical Centre, Rotterdam, THE NETHERLANDS. Joost C.M. Meijers, Academic Medical Center, University of Amsterdam, THE NETHERLANDS. Catherine Michalski, Aline Laulan, Ingrid Zede, LFB Biomédicaments, Lille-Cedex, FRANCE*. Sukesh C Nair, Christian Medical College Hospital, Vellore, INDIA. Neri Philosof, Anda Granot, OMRIX, Kiryat Uno, ISRAEL*. Marlien Pieters, North West University, Potchefstroom, SOUTH AFRICA. Sanj Raut, Sarah Daniels, NIBSC, Hertfordshire, UK*. Pierre-Harald Schmidt, Spectroscopy & Elementary Analysis, CSL Behring, Marburg, GERMANY*. Christelle Thouvenin, Stago, Laboratoire Etudes et Essais, Franconville, FRANCE. * Laboratories that participated in the Field Study. Laboratories that participated in a separate Stability Study.

25 Page 25 APPENDIX II INTERNATIONAL COLLABORATIVE STUDY TO CALIBRATE THE WHO 2 nd INTERNATIONAL STANDARD FOR FIBRINOGEN CONCENTRATE: VALUE ASSIGNMENT FOR ANALYTE: FIBRINOGEN PROTOCOL & ASSAY INSTRUCTIONS 1. AIMS OF STUDY To calibrate the proposed WHO 2 nd International Standard (IS) for Fibrinogen Concentrate and value assign for Fibrinogen Potency (Clottable and Total Protein). 2. MATERIALS There are 3 test materials provided (10 ampoules of each): CODE 1 st IS WHO 1 st IS Fibrinogen Concentrate (98/614), assigned potencies: - Clottable Protein = 10.4 mg/ampoule - Total Protein = 15.0 mg/ampoule B WHO IS Candidate Fibrinogen Concentrate B (09/242): - Clottable Protein ~ 8-12 mg/ampoule - Total Protein ~ mg/ampoule C WHO IS Candidate Fibrinogen Concentrate C (10/100): - Clottable Protein ~ 8-12 mg/ampoule - Total Protein ~ mg/ampoule 3. RECONSTITUTION AND STORAGE 1. On receipt, store all samples (unopened ampoules) at -20 o C. 2. On the day of assay, bring ampoules to room/ambient temperature. 3. Reconstitute at room/ambient temperature with 1.0 ml of distilled H 2 O or water for injection (WFI) and allow contents to solubilise for minutes at room/ambient temperature. 4. Mix gently and thoroughly to ensure complete solubilisation and transfer entire contents of ampoule to a plastic tube. 5. Keep samples at room temperature during assays.

26 Page NUMBER OF ASSAYS AND AMPOULES A minimum of four independent assays are requested for each method (i.e. 4 for Clottable Protein and 4 for Total Protein) from each laboratory, preferably carried out over separate days rather than all on the same day, in accordance with your routine methodology. Although 10 ampoules of each test material are provided, only 8 ampoules of each are needed for both methods. A fresh separate ampoule of each material should be used for each potency assay (i.e. 4 ampoules for Clottable Protein assays and 4 ampoules for Total Protein assays). Some laboratories may carry out only the Clottable Protein assays or only the Total Protein assays, in which case there will be 6 spare ampoules of each sample. If you are using more than one Clottable Protein method or more than one Total Protein method, material from the same ampoules should be used for each specific method, provided this can be done within 2 hours of reconstitution. If this is impossible, please contact us. Alternatively, the spare ampoules can be used for this purpose. 5. ASSAY METHODS Laboratories are asked to use their routine assay method for the measurement of both Clottable Protein (e.g. Clauss Assay) and Total Protein (e.g. Kjeldahl assay) concentrations, for those laboratories that can perform both. Please indicate potency method used on the Result Sheets. 6. CLOTTABLE PROTEIN: CLAUSS ASSAYS - DESIGN AND ANALYSIS Assay Dilution of Samples All working dilutions of each sample should be carried out using routine buffers. At least three dilutions should be made. After making all assay dilutions, discard reconstituted material and make up a fresh ampoule for the next assay. Assays using Parallel Line Analysis A balanced order of testing should be adopted wherever possible (see Assay Design below). In the following assay design, each letter represents a fresh set of 3 or more different (e.g. 1/5, 1/10, 1/20) dilutions. Where a letter is repeated twice within an assay, a replicate fresh set of dilutions from the same ampoule must be made. Remember to use a fresh set of ampoules for each assay. Please perform assays using the following design only. If this is not possible or you would like to use a different design please contact us.

27 Page 27 Assay Design for Clauss Assay 6 Samples per Assay (4 independent estimates for B and C in total per method) Assay No Order of Testing 1 1 st IS B amp 1 C amp 1 C amp 1 B amp 1 1 st IS 2 C amp 2 1 st IS B amp 2 B amp 2 1 st IS C amp 2 3 B amp 3 C amp 3 1 st IS 1 st IS C amp 3 B amp3 4 1 st IS B amp 4 C amp4 C amp 4 B amp 4 1 st IS Key to Samples: B - WHO IS Candidate (09/242); C - WHO IS Candidate (10/100); 1 st IS - WHO 1 st IS Fibrinogen Concentrate (98/614). Assays involving Interpolation to a Standard Curve Sample 1 st IS should be used to construct a standard curve preferably using at least two independent sets of dilutions (replicates). Samples B and C should be assayed using at least two and preferably three different dilutions tested in replicate. Other Methods Please carry out a minimum of four independent assays using your normal routine methodology for determination of Clottable Protein concentration. For clot dissolution/supernatant + E280 measurement for determination of Clottable Protein concentration, please record Extinction Coefficient(s) used on the result sheet template provided (page 8). 7. TOTAL PROTEIN (e.g. Biuret, Gravimetric, Dye binding/ods, E 280, Kjeldahl or Immunological methods) Please carry out a minimum of four independent assays using your normal routine methodology. For E280 measurement for determination of Total Protein concentration, please record Extinction Coefficient(s) used on the Result Sheet 3 provided (page 7).

28 Page SUBMISSION OF ASSAY RESULTS AND METHODOLOGICAL DETAIL Please record your results and all raw data (i.e. OD values or clotting times etc) including all details of dilutions performed, on the appropriate accompanying Results Sheet templates (pages 5, 7 & 9). Please retain your raw assay data for possible future reference. CLOTTABLE PROTEIN ASSAYS In addition, for laboratories carrying out CLAUSS assays, you are invited to return your own calculations for potency estimates of Fibrinogen Clottable Protein (using Result Sheet 2, page 6) for samples B & C against the current WHO 1 st IS Fibrinogen Concentrate (98/614) 1 st IS, with the latter having a Clottable Protein assigned potency of 10.4 mg/ampoule. For laboratories carrying out other methods, you are invited to return your own calculations for potency estimates (Fibrinogen Clottable Protein) for samples B, C and 1 st IS (WHO 1 st IS Fibrinogen Concentrate - 98/614, using Results Sheet 4 (page 8). TOTAL PROTEIN ASSAYS For laboratories carrying out Total Protein determination, you are invited to return your own calculations for potency estimates (Total Protein) for samples B, C and 1 st IS (WHO 1 st IS Fibrinogen Concentrate - 98/614, using Results Sheet 6 (page 10). Please also complete the attached questionnaire and provide details of your assays (page 9). PLEASE SEND YOUR RESULTS ELECTRONICALLY BY 30 th OCTOBER 2010 TO: Dr S. Raut Sanj.Raut@nibsc.hpa.org.uk

29 Page 29 EXAMPLE STUDY TO CALIBRATE THE WHO 2 nd INTERNATIONAL STANDARD FOR FIBRINOGEN CONCENTRATE RESULTS SHEET 1: RAW DATA Laboratory: Date: Potency Method: CLOTTABLE Protein - CLAUSS ASSAY Assay No.: Preparation (Code No.) WHO IS Candidate Concentrate B (09/242) B ampoule 1 WHO IS Candidate Fibrinogen Concentrate C (10/100) C ampoule 1 WHO 1 st IS Fibrinogen Concentrate (98/614) 1 st IS Working Dilutions: e.g. 1/5 1/10 1/20 e.g. 1/5 1/10 1/20 e.g. 1/5 1/10 1/20 Replicate Response (s) Clotting Time (Sec) For laboratories performing Interpolation to a Standard Curve, please give following details: Standard Used: Dilution Range of Standard Replicate Response (s)

30 Page 30 EXAMPLE STUDY TO CALIBRATE THE WHO 2 nd INTERNATIONAL STANDARD FOR FIBRINOGEN CONCENTRATE Laboratory: RESULTS SHEET 2: LABORATORY s OWN CALCULATED POTENCIES Potency Method: CLOTTABLE Protein - CLAUSS ASSAY Date: Method of Calculation: Parallel Line/other Test Preparation WHO IS Candidate Fibrinogen Concentrate B (09/242) B WHO IS Candidate Fibrinogen Concentrate C (10/100) C Assay Date Fibrinogen CLOTTABLE Protein (mg/ml) vs 1 st IS Estimate 1 Estimate 2 Estimate 3 Estimate 4 MEAN Estimate Estimate 1 Estimate 2 Estimate 3 Estimate 4 MEAN Estimate

31 Page 31 EXAMPLE STUDY TO CALIBRATE THE WHO 2 nd INTERNATIONAL STANDARD FOR FIBRINOGEN CONCENTRATE RESULTS SHEET 3: RAW DATA for OTHER METHODS Laboratory: Date: Potency Method: CLOTTABLE Protein: Assay No.: (e.g. Clot Dissolution/Supernatant + Gravimetric, Biuret, Dye binding/ods, E 280, Kjeldahl) Preparation (Code No.) WHO IS Candidate Concentrate B (09/242) B ampoule 1 WHO IS Candidate Fibrinogen Concentrate C (10/100) C ampoule 1 WHO 1 st IS Fibrinogen Concentrate (98/614) 1 st IS ampoule 1 Total Dilution Factor e.g. ODs at 280nm Replicate Responses e.g. ODs at 315nm For laboratories performing Interpolation to a Standard Curve, please give following details: Standard Used: Dilution Range of Standard Replicate Response (s)

32 Page 32 EXAMPLE STUDY TO CALIBRATE THE WHO 2 nd INTERNATIONAL STANDARD FOR FIBRINOGEN CONCENTRATE RESULTS SHEET 4: LABORATORY s OWN CALCULATED POTENCIES (OTHER METHODS) Laboratory: Potency Method: CLOTTABLE Protein: Date: (e.g. Clot Dissolution/Supernatant + Gravimetric, Biuret, Dye binding/ods, E 280, Kjeldahl) Method of Calculation: e.g. Extinction Coefficient/other Test Preparation Assay Date Fibrinogen CLOTTABLE Protein (mg/ml)* WHO IS Candidate Fibrinogen Concentrate B (09/242) B WHO IS Candidate Fibrinogen Concentrate C (10/100) C WHO 1 st IS Fibrinogen Concentrate (98/614) 1 st IS Estimate 1 Estimate 2 Estimate 3 Estimate 4 MEAN Estimate Estimate 1 Estimate 2 Estimate 3 Estimate 4 MEAN Estimate Estimate 1 Estimate 2 Estimate 3 Estimate 4 MEAN Estimate *For participants using E280 measurement for determining CLOTTABLE Protein concentration, please give Extinction Coefficient used:

33 Page 33 EXAMPLE STUDY TO CALIBRATE THE WHO 2 nd INTERNATIONAL STANDARD FOR FIBRINOGEN CONCENTRATE RESULTS SHEET 5: RAW DATA for TOTAL PROTEIN Determination Laboratory: Date: Potency Method: TOTAL Protein Assay No.: (e.g. Biuret, Gravimetric, Dye binding/ods, E 280, Kjeldahl or Immunological assays) Preparation (Code No.) WHO IS Candidate Concentrate B (09/242) B ampoule 1 WHO IS Candidate Fibrinogen Concentrate C (10/100) C ampoule 1 WHO 1 st IS Fibrinogen Concentrate (98/614) 1 st IS ampoule 1 Total Dilution Factor e.g. ODs at 280nm Replicate Responses e.g. ODs at 315nm For laboratories performing Interpolation to a Standard Curve, please give following details: Standard Used: Dilution Range of Standard Replicate Response (s)

34 Page 34 EXAMPLE STUDY TO CALIBRATE THE WHO 2 nd INTERNATIONAL STANDARD FOR FIBRINOGEN CONCENTRATE Laboratory: RESULTS SHEET 6: LABORATORY s OWN CALCULATED POTENCIES (TOTAL PROTEIN) Potency Method: TOTAL Protein: Date: (e.g. Biuret, Gravimetric, Dye binding/ods, E 280, Kjeldahl or Immunological assays) Method of Calculation: e.g. Extinction Coefficient/other Test Preparation WHO IS Candidate Fibrinogen Concentrate B (09/242) B WHO IS Candidate Fibrinogen Concentrate C (10/100) C WHO 1 st IS Fibrinogen Concentrate (98/614) 1 st IS Assay Date Fibrinogen CLOTTABLE Protein (mg/ml)* Estimate 1 Estimate 2 Estimate 3 Estimate 4 MEAN Estimate Estimate 1 Estimate 2 Estimate 3 Estimate 4 MEAN Estimate Estimate 1 Estimate 2 Estimate 3 Estimate 4 MEAN Estimate *For participants using E280 measurement for determining TOTAL Protein concentration, please give Extinction Coefficient used:

35 Page 35 STUDY TO CALIBRATE THE WHO 2 nd INTERNATIONAL STANDARD FOR FIBRINOGEN CONCENTRATE QUESTIONNAIRE NAME AND LABORATORY: 1. Method Details: 2. Clottable Protein Assay: - Source of Thrombin reagent - Instrumentation/analyser 3. Total Protein Assay: - Critical reagents - Instrumentation/analyser 4. Dilution Buffer: - Buffer details including source 5. Please record any deviation from your routine method used: 6. Comments on any aspects of the study:

36 Page 36 APPENDIX III Participants methods and reagents used: (A) Clottable Protein Lab Method Method details Source of Thrombin Reagent Dilution: Buffer Instrumentation 1 Clauss Assay - - Factor diluent (IL) ACL 200 (IL) 2 Clauss Assay - STA - Fibrinogen 5/lyophilized human calcium thrombin (~80 NIH units/ml) STA - Owren-Koller buffer ph~7.4 STA Analyzer: STAR-R 3A CLOTr Clot removal + Biuret assay of supernatant Bovine Thrombin 50NIH units (Diagnostic Reagents Ltd) 3B Clauss Assay - Thrombin Reagent - Diagnostica Stago Fibrinogen 5 3C CLOTr Clot removal + OD of Bovine Thrombin (Diagnostica dissolved clot Stago) 5 CLOTr Clot removal + Biuret assay of supernatant IIa component of Beriplast (CSL Behring) Saline Veronal Buffered Saline Clot Solvent (0.15M NaCl, 7M Urea, 0.2M NaOH) Saline Unicam 300 series UV/VIS spectrophotometer (absorbance 540nm) ST4 Coagulometer L10284 (Diagnostica Stago ) Unicam 300 series UV/VIS spectrophotometer (absorbance 280nm) Spectramax microplate reader (Molecular Devices) (absorbance 550nm) STAR-R Evolution (Diagnostica Stago) Siemens BCS XP 6 Clauss Assay - Stago Thrombin Owren-Koller buffer (Diagnostiga Stago) 7A Clauss Assay - Bovine Thrombin (Siemens Saline 0.85% Multifibren U) 7B Clauss Assay - Bovine Thrombin (rfibrinogen, Owren-Veronal buffer Siemens BCS Siemens In-House assay) (Dade) 8A Clauss Assay - Bovine Thrombin HemosIL Saline + BSA 1% ACL TOP Analyser (IL) 9 Clauss Assay - Bovine Thrombin (Siemens) Owren-Veronal buffer Siemens CA7000 Analyser 10A Clauss Assay - Bovine Thrombin (Diagnostica 0.05M Imidazole buffer, ph ACL TOP 500 Analyser (IL) Stago) B CLOTr Clot removal + OD of Bovine Thrombin (Diagnostica Clot Solvent (0.15M NaCl, Cecil spectrophotometer & C dissolved clot Stago) 7M Urea, 0.2M NaOH) (absorbance 280nm) 11 Clauss Assay - Bovine Thrombin - 35 NIH units/ml Imidazole buffer ACL TOP Analyser (IL) (IL) 13 Clauss Assay - Thrombin Reagent - Diagnostica Owren-Koller buffer STA Compact Stago STA Fibrinogen 5 Assay (Diagnostiga Stago) 14A Clauss Assay - Thrombin Reagent - Thromborel S Owren s buffer Sysmex 1500 Analyzer (Siemens) 14B ELISA Ant-human fibrinogen - PBS % Tween20 - Ab (DAKO) 15 CLOTr Clot removal + Protein Thrombin, 300IU/ml (Baxter S&R) Citrate NaCl buffer (5.4g/L Steam Distillation Apparatus (N 2) determination of Citrate;7.5g/L NaCl) Titration Unit dissolved clot (Kjeldahl) Digestion Unit 17 Clauss Assay - Thrombin Reagent - Diagnostica Stago Fibrinogen 5 Veronal Buffered Saline ST4 Coagulometer L10284 (Diagnostica Stago ) 18A Clauss Assay STA-Fib 2 assay Titrated Human Calcium Thrombin STA-Owren-Koller buffer STA-R (40 NIH units/ml), ph B Clauss Assay STA-Fibrinogen assay Titrated Human Calcium Thrombin STA-Owren-Koller buffer STA-R (80 NIH units/ml), ph Clauss Assay - Thrombin (Sigma) Veronal buffer & Imidazole Amelung KC10 buffer 20A Clauss Assay - Thrombin: HemosIL Fibrinogen-C Saline Solution (HemosIL ACL TOP 700 LAS reagent (IL) Factor Diluent) 20B ELISA Ant-human fibrinogen - Phosphate Buffer containing Siemens BNII Analyser Ab (Siemens) NaCl (N Diluent - Siemens) 21 Clauss Assay - Thrombin (Dade Behring) - Sysmex CS2100i 22 Clauss Assay - Thrombin Reagent (Siemens) (Siemens) Sysmex CS-2000i 23 Clauss Assay - IL Thrombin reagent (FibC kit - IL Factor Diluent ACL TOP (IL) IL) 25 Clauss Assay - Human Liquid Thrombin FIB L Owren buffer (Diagnostic Amax, Mechanical Human (Diagnostic Grifols) Grifols) 26A Clauss Assay - Fibri-prest (Stago) Owren-Koller buffer BCS-XP (Behring) 26B CLOTr Clot removal + Drying Helena Thrombin Reagent 0.15M NaCl M CaCl Precision Balance and Weighing Method 27A Clauss Assay - Bovine Thrombin (IL) 0.9% NaCl solution (w/v) ACL Elite Pro (IL) & B 28 Clauss Assay - Thrombin Reagent (Siemens) Siemens Kit (Siemens) Siemens BCS XP 29 CLOTr Clot removal + OD of dissolved clot Thrombin Reagent (In-house) Clot Solvent (0.15M NaCl, 7M Urea, 0.2M NaOH) + PBS Evolution 300 spectrophotometer (absorbance 280nm) 30 CLOTr Clot removal + Protein (N 2) determination of dissolved clot (Kjeldahl) 31 Clauss Assay - Thrombin Reagent (Helena Bioscience) Thrombin Reagent (In-house) - Steam Distillation Apparatus Titration Unit Digestion Unit Owren buffer (Helena BioScience.) STA Compact Coagulometer

37 Page 37 (B) Total Protein Lab Method Dilution: Buffer Instrumentation 1 BCA Protein Assay Factor diluent (IL); BCA Method (Sigma BCA-1) Spectrophotometer (absorbance 565nm) 3 Biuret Assay Saline Unicam 300 series UV/VIS spectrophotometer (absorbance 540nm) 5 Biuret Assay Saline Spectramax microplate reader (Molecular Devices) (absorbance 550nm) 10 Protein Absorbance Clot Solvent (0.15M NaCl, 7M Cecil spectrophotometer (absorbance 280nm) Assay Urea, 0.2M NaOH) 15 Kjeldahl method Sulphuric Acid digestion for N 2 - determination 20 Synchron M-TP Factor Diluent (HemosIL) UniCel DxC 800 (Beckman/Coulter) Microprotein Assay 25 Biuret Assay Physiological Saline Spectramax microplate reader (Molecular Devices) 26A Kjeldahl Method Sulpuric Acid digestion for N 2 Buchi AutoKjeldahl Unit K370 determination 26B Biuret Assay - Spectrophotometer (Varian Cary) + Hamilton Diluter 27 BioRad Protein Assay 0.9% NaCl solution (w/v) Spectrophotometers UV-VIS models: UV-1700 & UV-1603 (Shimadzu)

38 Page 38 APPENDIX IV Laboratories Own Calculations Clottable Protein Potency relative to IS (10.4 mg/ml) Lab No. No of Assays (n) Proposed 2 nd IS, Sample B (09/242) [Data from Main & Field Studies] Geometric Mean (mg/ml) Intra- Lab GCV% No of Assays (n) Sample C (10/100) [Data from Main Study] Geometric Mean (mg/ml) Intra- Lab GCV% Clauss Assay Data from Main Study Data from Field Study A B A A A B A A A B C B CLOTr Method Data from Main Study Data from Field Study Data from Main Study 3A 4 4-5A B B C B C ELISA 14B B 4-4 -

39 Page 39 Laboratories Own Calculations Total Protein Potency relative to IS (15.0 mg/ml) Sample B Sample C Lab No. No of Assays (n) Laboratory s Own Mean (mg/ml) No of Assays (n) Laboratory s Own Mean (mg/ml) A B

40 Page 40 APPENDIX V Field Study Table A. Field Study data on Fibrinogen Products measured using CLOTr Methods. Fibrinogen (Fg) Clottable Protein potencies are calculated relative to the 1 st IS and the proposed WHO 2 nd IS (sample B), with proposed and alternative assigned potencies. Potency Ratios of assays relative to the 1 st IS and the proposed WHO 2 nd IS (sample B), with proposed and alternative assigned potencies, are also shown. CLOTr Potency relative to the proposed WHO 2nd IS, sample B (with various assigned potency options) Lab No. Proposed Assigned value based on CLOTr Method 10.9 mg/ml Alternative Assigned value based on Combined Method Mean 12.1 mg/ml Alternative Assigned value based on Clauss Assay 12.9 mg/ml Manufacturer Lab 3 OMCL Lab 5 OMCL Lab 10 Manufacturer Lab 29 % Shift (range) in potency relative to B Fibrinogen Products (n= No of assays) Mean Fg Potency vs 1st IS (98/614) (assigned Potency 10.4 mg/ml) Mean Fg Potency vs B (09/242) (10.9 mg/ml) Potency Ratio: Assays vs 1st IS/B (B=10.9 mg/ml) Mean Fg Potency vs B (09/242) (12.1 mg/ml) Potency Ratio: Assays vs 1st IS/B (B=12.1 mg/ml) Mean Fg Potency vs B (09/242) (12.9mg/ml) Potency Ratio: Assays vs 1st IS/B (B=12.9 mg/ml) R 1 (n=4) S 1 (n=4) T 1 (n=4) R 2 (n=4) S 2 (n=4) T 2 (n=4) R 3 (n=4) S 3 (n=4) T 3 (n=4) R 4 (n=4) S 4 (n=4) T 4 (n=4) R, S & T represents different batches of products from each manufacturer. 2 4% 8 14% 17 19%

41 Page 41 Table B. Field Study data on Fibrinogen Products measured using Clauss Assays. Fibrinogen (Fg) Clottable Protein potencies are calculated relative to the 1 st IS and the proposed WHO 2 nd IS (sample B) with proposed and alternative assigned potencies. Potency Ratios of assays relative to the 1 st IS and the proposed WHO 2 nd IS (sample B), with proposed and alternative assigned potencies, are also shown. Clauss Potency relative to the proposed WHO 2nd IS, sample B (with various assigned potency options) Lab No. Proposed Assigned value based on CLOTr Method 10.9 mg/ml Alternative Assigned value based on Combined Method Mean 12.1 mg/ml Alternative Assigned value based on Clauss Assay 12.9 mg/ml Manufacturer Lab 3 Manufacturer Lab 26 Manufacturer Lab 27 Manufacturer Lab 31 % Shift (range) in potency relative to B Fibrinogen Products (n= No of assays) Mean Fg Potency vs 1st IS (98/614) (assigned Potency 10.4 mg/ml) Mean Fg Potency vs B (09/242) (10.9 mg/ml) Potency Ratio: Assays vs 1st IS/B (B=10.9 mg/ml) Mean Fg Potency vs B (09/242) (12.1 mg/ml) Potency Ratio: Assays vs 1st IS/B (B=12.1 mg/ml) Mean Fg Potency vs B (09/242) (12.9mg/ml) Potency Ratio: Assays vs 1st IS/B (B=12.9 mg/ml) R 5 (n=4) S 5 (n=4) T 5 (n=4) R 6 (n=4) S 6 (n=4) T 6 (n=4) R 7 (n=4) S 7 (n=4) T 7 (n=4) R 8 (n=4) S 8 (n=4) T 8 (n=4) R,S & T represents different batches of products from each Laboratory % 2 9% 2 4%

42 Page 42 APPENDIX VI Draft Instruction for Use form for the proposed WHO 2 nd International Standard for Fibrinogen concentrate (09/242).

43 Page 43