EXPERT COMMITTEE ON BIOLOGICAL STANDARDIZATION Geneva, 17 to 21 October 2016
Calibration of the Proposed 5
thInternational Standard for Thromboplastin, human, recombinant, plain (rTF/16; study code 14/001) and the Proposed 5
thInternational Standard for Thromboplastin, rabbit, plain (RBT/16; study
code 15/001)
Antonius M.H.P. van den Besselaar1, Veena Chantarangkul2 and Armando Tripodi2
1Department of Thrombosis and Hemostasis, Leiden University Medical Center, Leiden, The Netherlands
2Angelo Bianchi Bonomi Hemophilia and Thrombosis Centre, Department of Clinical Sciences and Community Health, Università degli Studi di Milano and IRCCS Cà Granda Maggiore
Hospital Foundation, Milano, Italy NOTE:
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 16 September 2016 and should be addressed to the World Health Organization, 1211 Geneva 27, Switzerland, attention:
Technologies, Standards and Norms (TSN). Comments may also be submitted electronically to the Responsible Officer: Dr C M Nübling at email: nueblingc@who.int.
© World Health Organization 2016
All rights reserved. Publications of the World Health Organization are available on the WHO web site (www.who.int) or can be purchased from WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264; fax: +41 22 791 4857; e-mail: 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:
(http://www.who.int/about/licensing/copyright_form/en/index.html).
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
2
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.
Summary
Background and aim:
Stocks of the current 4th International Standard for Thromboplastin, human, recombinant, plain (coded rTF/09) and the current 4th International Standard for Thromboplastin, rabbit, plain (coded RBT/05) are running low. Candidate replacement materials have been prepared
(provisionally coded 14/001 and 15/001, respectively). This report describes the calibration of the proposed 5th International Standard for Thromboplastin, human, recombinant, plain, and the proposed 5th International Standard for Thromboplastin, rabbit, plain.
Methods
An international collaborative study was carried out for the assignment of International Sensitivity Index (ISI) to the candidate materials. Results were obtained from 20 laboratories.
The validity of each and every ISI assessment was checked using criteria published by the WHO Expert Committee on Biological Standardization. Only valid ISI assessments were used to calculate the mean ISI for each candidate.
Results
The mean ISI for candidate 14/001 was 1.11 (between-laboratory CV: 5.7%) and the mean ISI for candidate 15/001 was 1.21 (between-laboratory CV: 4.6%).
Accelerated degradation tests were performed to assess the stability of the freeze-dried candidate materials. There was no significant change of the prothrombin time after storage of the candidate materials at 5°C for 56 days. After storage at elevated temperatures (31°C - 42°C) there was a slight but significant change of the prothrombin time. The stability after reconstitution was assessed at room temperature, over a time interval of 0.5 - 4 hours. Between 1 and 4 hours after reconstitution, there was no change of the prothrombin time with candidate 14/001, and a slight change (-1.5%) with candidate 15/001.
Conclusion and proposal
It is proposed that candidate 14/001 is accepted as the WHO 5th International Standard for Thromboplastin, human, recombinant, plain, and candidate 15/001 as the WHO 5th International Standard for Thromboplastin, rabbit, plain. Upon establishment, it is proposed to replace the provisional coding of the candidates by rTF/16 and RBT/16, respectively.
Introduction
Treatment with vitamin K-antagonists (VKA) must be monitored to adjust the dose if necessary.
The primary laboratory method for monitoring of VKA is the prothrombin time test (PT). The PT test is performed with a thromboplastin reagent and a method or instrument to determine the coagulation time in seconds. Various types of thromboplastin are prepared commercially and, in order to be able to interpret the results of the PT test, it is essential that each reagent is correctly calibrated (1). This will ensure that the results of tests with different products and batches can be compared. A procedure for the calibration of thromboplastins using a logarithmic plot of PTs has been developed. With this procedure, the definition of a calibration parameter called
International Sensitivity Index (ISI) became feasible. It is possible to express PT results on a common scale, i.e. the International Normalized Ratio (INR), if the ISI of the thromboplastin used is known.
Two International Standards for thromboplastin are presently available for calibration of secondary standards or commercial thromboplastin preparations: the Fourth International
Standard for thromboplastin, rabbit, plain (coded RBT/05) and the Fourth International Standard for thromboplastin, human, recombinant (coded rTF/09)(2, 3). In general, the calibration of a given thromboplastin reagent is more precise if performed against an International Standard of similar composition and from the same species. Thus, plain rabbit thromboplastins should be calibrated against RBT/05 and plain human thromboplastins against the human recombinant material rTF/09.
Stocks of both rTF/09 and RBT/05 are now running low and replacements are required.
Candidate replacement materials have been prepared for thromboplastin, human, recombinant, plain (provisionally coded 14/001) and for thromboplastin, rabbit, plain (provisionally coded 15/001). The calibration of the candidate replacement materials should be carried out as part of international multicentre collaborative studies using fresh coumarin and fresh normal plasmas and using manual techniques for PT determination. The ISI assigned to each replacement material should be the mean of the ISIs obtained by calibration with both existing International Standards, i.e. rTF/09 and RBT/05. The present report deals with the results of an international collaborative study organized under the auspices of the Scientific and Standardization
Committee (SSC) of the International Society on Thrombosis and Haemostasis (ISTH) for the calibration of the two candidate replacement materials.
4 Details of Candidate Preparations
Detail Proposed 5th IS Human,
Recombinant, Plain (14/001)
Proposed 5th IS Rabbit Plain (15/001)
Presentation
Two components sealed in glass Hodes-Lange type ampoules under two purified nitrogen flushes:
Dried Thromboplastin (14/001) and Reconstitution Fluid (14/002)
Two components sealed in glass Hodes-Lange type ampoules under two purified nitrogen flushes:
Dried Thromboplastin (15/001) and Reconstitution Fluid (15/002) Excipients/additives
Dried Thromboplastin:
Recombinant human tissue factor, phospholipid mixture with Tris buffer, sodium chloride, trehalose, bovine gamma globulin, and
antimicrobials pH 7.4. Manufactured 27 Jan-01 April 2014.
Reconstitution Fluid: calcium chloride, polybrene and
antimicrobials. Manufactured 07-08 April 2014.
Dried Thromboplastin:
Rabbit brain thromboplastin powder with Hepes-sodium buffer,
polybrene, glycine, pH 7.3.
Manufactured 05-06 March, 2015.
Reconstitution Fluid: calcium chloride with Hepes-sodium buffer, glycine, polybrene, sodium chloride, PEG 4000 and antimicrobials, pH 7.1. Manufactured 17-18 March, 2015.
Liquid filling weight (g)
Thromboplastin Fill Mean:
0.261g (weight range 0.243-0.273g), target = 0.265g (target range±10% of target =0.238-0.292)
Reconstitution Fluid Fill:
Mean >1.6 mL (exceeds 1.0 mL required)
Thromboplastin Fill Mean:
0.479g (weight range 0.468-0.493g), target = 0.481g (target range±7.5%
of target =0.445-0.517)
Reconstitution Fluid Fill:
Mean >1.6 mL (exceeds 1.0 mL required)
Coefficient of variation of the fill weight (%)
Dried Thrombopalstin CV = 1.9%
based on 117 weight checks per 17 trays/>22000 ampoules
Reconstitution Fluid CV = not applicable, filled in excess
Dried Thromboplastin CV = 1.2%
based on 52 weight checks per 12 trays/>16700 ampoules
Reconstitution Fluid CV = not applicable, filled in excess
Coeff. Of Variation of Thromboplastin PT across fill (% of Normal Control)
Mean 9.95 sec (n=18, 17 trays) CV 1.2% (Spec<4%)
Mean 13.57 sec (n=24, 12 trays) CV 1.1% (Spec<2%)
Residual moisture after lyophilisation (%)
Moisture in Dried Thromboplastin by Laser Absorption (17 samples across 17 trays, 5 measurements per sample):
Mean p(H2O) = 3.3 torr (range 2.9- 4.6 torr, Target <5.6 torr)
Mean %(w/w) Moisture (approximate, estimated by KF correlation) = 3.5%
Moisture in Dried Thromboplastin by Laser Absorption (12 samples across 12 trays, 3 measurements per sample):
Mean p(H2O) = 1.8 torr (range 1.2- 2.3 torr, Target <2.51 torr)
Mean %(w/w) Moisture (approximate, estimated by KF correlation) = 0.72%
Headspace oxygen (%)
Mean 0.2 %atm (range 0.1-0.4 % atm)
Mean 0.4 %atm (Range 0.1-1.3%
atm)
Reconstitution Procedure
Deliver 1.0 ml of rTF/16 Reconstitution Fluid per one (1) Dried rTF/16 Thromboplastin ampoule with swirling, allow 20 minutes at Room Temperature (~15- 25oC) to make 1.0 mL Working rTF/16 Thromboplastin
Deliver 1.0 ml of RBT/16 Reconstitution Fluid per one (1) Dried RBT/16 Thromboplastin ampoule with swirling, allow 30 minutes at Room Temperature (~15- 25oC) to make 1.0 mL Working RBT/16 Thromboplastin
Number of ampoules in stock
10,000 10,000
Manufacturing site Instrumentation Laboratory, Orangeburg NY, USA
Instrumentation Laboratory, Orangeburg NY, USA Custodian NIBSC, Potters Bar, UK NIBSC, Potters Bar, UK
Storage temperature -20 °C (Thromboplastin) 2-8 °C (Reconstitution Fluid)
-20 °C (Thromboplastin) 2-8 °C (Reconstitution Fluid)
Design of the collaborative study for ISI value assignment
Twenty laboratories from Europe, North America, South America, Asia and Australia were invited to participate in the collaborative study. Fourteen of these had participated in the previous collaborative study for the calibration of rTF/09 (3). The candidate materials (14/001 and
15/001) and the current International Standards (rTF/09 and RBT/05) were tested in each laboratory by the same operator with the manual (tilt tube) technique, according to a study protocol with detailed instructions. Test plasmas were freshly prepared from healthy (“normal”) subjects and patients on long-term therapy with VKA. Participants were instructed to select patient plasmas with PT corresponding to an INR interval from 1.5 to 4.5. In addition to the fresh plasmas, four lyophilized control plasmas were included. To account for the effect of inter-day variation, PT measurements were performed in each laboratory on 10 different days. Participants were instructed to include on each day plasmas from two healthy individuals and six patients treated with VKA. Healthy individuals and patients had to be different on each working day. To minimize the effect of preanalytical plasma instability on the relationship between
thromboplastins, the order of testing was changed each day. The order of testing was specified in the data-collection form. The collaborative study was carried out from October 2015 until
February 2016.
Statistical methods
Calculation of ISI for candidate materials
The statistical methods have been described in the Guidelines for thromboplastins and plasma used to control oral anticoagulant therapy with vitamin K antagonists (1). The sequence of steps in the statistical evaluation for each participant’s calibration is as follows.
6 1. Calculate preliminary orthogonal regression line (20 normal samples + 60 patient
samples).
2. Detect outliers defined as points with a perpendicular distance greater than 3 residual standard deviations from the preliminary line.
3. Remove outliers in one step and recalculate the orthogonal regression line (normal samples + patient samples) and ISI as the product of the slope of the regression line and the ISI of the International Standard.
4. Calculate each patient’s INR using the PT determined with the International Standard.
5. Calculate each patient’s INR using the PT determined with the candidate material and ISI from step 3.
6. Calculate each patient’s mean INR from steps 4 and 5.
7. Remove patients with mean INR < 1.5 or mean INR > 4.5.
8. Recalculate the orthogonal regression line for the remaining normal samples + patient samples and ISI as the product of the slope of the regression line and the ISI of the International Standard.
Mean ISI and between-laboratory coefficient of variation
The arithmetic mean and between-laboratory standard deviation (SD) of ISI assessments was calculated. The between-laboratory coefficient of variation was calculated as the SD divided by the mean ISI and multiplied by 100.
Validity of ISI
Each and every ISI assessment was checked for validity. The within-laboratory coefficient of variation (CV) of the slope of the orthogonal regression line for normal samples + patient samples should be 3% or less (1). ISI assessments with a within-laboratory CV of the slope larger than 3% were rejected. Secondly, the adequacy of the ISI model was checked. While there is good evidence that the calibration relationship defined in a double-logarithmic plot of PTs is usually linear, and that the same line represents data points for both patients and healthy subjects, the possibility of departure from these assumptions cannot be ruled out. To assess the magnitude of INR deviations, the orthogonal regression line was calculated for patients’ samples only: Y = a’ + b’·X, in which Y is the natural logarithm of patient’s PT determined with the International Standard and X is the natural logarithm of patient’s PT determined with the candidate material.
INRs calculated with the patients’ only line were compared to INRs calculated with the ISI. In case of marked INR deviation, the assignment of an ISI would not be meaningful. For practical purposes, the assignment of an ISI is acceptable if INRs calculated with the ISI do not differ by more than 10%, in the INR range 2 – 4.5, from INRs calculated with the equation describing the above-mentioned regression line for patients’ samples only (1). Conversely, the assignment of an ISI is not acceptable if INRs differ by more than 10%. ISI assessments resulting in INR
deviations larger than 10% were rejected.
Results of ISI value assignment
Geometric mean normal PT (MNPT) and mean patients’ PT are shown in Table 1. The Pearson correlation between the four different thromboplastins was significant (P<0.001), suggesting that the mean PT differences between the laboratories were significant.
ISI values calculated including all data are shown in Table 2. The between-laboratory variation (CV) of like-to-like calibrations (i.e. 2.2% and 4.3% for 14/001 and 15/001 respectively) was smaller than the CVs for the cross-species calibrations (i.e. 6.4% and 5.6% for 14/001 and 15/001 respectively). The number of outliers (> 3 SD) and samples with INR<1.5 or INR>4.5 is shown in Table 3. ISI values calculated after exclusion of outliers (>3 SD) are shown in Table 4.
ISI values calculated after exclusion of outliers (> 3 SD) and samples with INR<1.5 or INR>4.5 are shown in Tables 5 and 6. In the cross-species calibration of candidate 15/001 against rTF/09, there was one laboratory with a CV of the slope (CVB) larger than 3.0% (Table 5). In the
calibration of candidates 14/001 and 15/001 with RBT/05, there were two results with a CV of the slope (CVB) larger than 3.0% (Table 6).
Intercept (a’) and slope (b’) of the orthogonal regression lines for patients’ samples only are shown in Table 7. The deviations of the INR calculated with the ISI for the “true INR”=2 and the
“true INR” = 4.5 are shown in Table 8. For the like-to-like calibration of candidate 14/001 against rTF/09, all laboratories had deviations <10%. For the cross-species calibration of
candidate 14/001 against RBT/05, 6 laboratories had deviations larger than 10%. For the like-to- like calibration of candidate 15/001 against RBT/05, only one laboratory had a deviation larger than 10%. For the cross-species calibration of candidate 15/001 against rTF/09, 8 laboratories had deviations larger than 10%.
Table 9 shows the mean ISI and inter-laboratory CV for the candidates before and after
exclusion of non-valid ISI assessments. For candidate 14/001, exclusion of non-valid assessment had no important effect on the mean ISI. The final mean ISI for candidate 14/001 is 1.11 with a CV of 5.7%. For candidate 15/001, exclusion of non-valid assessments resulted in a 1% higher mean ISI and a reduction of the between-laboratory variation of the ISI. The final mean ISI for candidate 15/001 is 1.21 with a CV of 4.6%.
For each of the lyophilized control plasmas and for each participant, INR was calculated with the local MNPT and the mean ISI for each Thromboplastin reagent. Table 10 shows the mean INR and the inter-laboratory SD. The differences between the mean values with the 4 thromboplastin reagents were relatively small. Further analysis of the INR values for the lyophilized control plasmas will be performed later.
Stability of the proposed International Standards.
Accelerated degradation study
Tissue factor is a glycoprotein that needs the association with phospholipids for full expression of its procoagulant activity. Lyophilized tissue factor preparations (thromboplastin) for PT determination usually contain many other components such as residual water which may influence the stability of the reagent. In general, the long-term stability of lyophilized tissue
8 factor from human or animal brain stored at low temperature is excellent. The stability of
lyophilized biological materials may be predicted from accelerated degradation tests. The purpose of an accelerated degradation test is to measure the relative rates of potency loss at several temperatures and to extrapolate the rate to the desired temperature of storage. For the deterioration of lyophilized tissue factor preparations complex kinetics are expected because tissue factor is a lipoprotein. It may be difficult to predict the stability of tissue factor at low temperature from the results of an accelerated degradation test. Nevertheless, the results of an accelerated degradation test may be useful to assess the relative stability under transportation conditions at various temperatures. The accelerated degradation test is a standard procedure to check the stability of lyophilized thromboplastin preparations (1).
Following shipment of both candidates (coded 14/001 and 15/001) to Leiden, they were stored at minus 20°C. The respective reconstitution fluids (coded 14/002 and 15/002) were stored at +5°C.
For the accelerated degradation test, a number of ampoules of 14/001 and 15/001 were stored at +5°C, +31°C, +36°C, and +42°C, for different time intervals. The time intervals ranged from 1 to 59 days. In fact, the ampoules were placed in the incubators on different dates, so that they all could be analysed in the same session at the end of the different incubation time intervals. After storage of the candidates at these temperatures, they were reconstituted and tested with two deep- frozen pooled plasmas (one normal, coded NP141202, and one coumarin plasma obtained from patients treated with VKA , coded AP141204). The tests were performed with a semi-automatic, electro-mechanical coagulometer according to Schnitger and Gross manufactured by Amelung GmbH (Lemgo, Germany). All ampoules stored at a given temperature were tested in the same session. In addition, ampoules stored at -20°C were tested in each session. The results obtained with the ampoules stored at -20°C were considered as the initial value at zero storage time, assuming that deterioration at -20°C was negligible. The time between reconstitution and of each ampoule and testing was 2 hours. For each storage temperature and time, 3 ampoules were used.
Each ampoule was tested in single PT determination. The mean values of the 3 PT’s were used for statistical analysis. For each storage temperature and time, a clotting time ratio (PT ratio) was calculated as the mean PT of the coumarin plasma divided by the mean PT of the normal plasma (AP/NP). Correlation coefficients according to Spearman were used to test the change of PT or PT-ratio with incubation time. A significance level of 5% was used.
PT’s and PT-ratios as a function of storage time at various temperatures are shown in Tables 11 and 12. For candidate 14/001, a slight decrease of the PT of normal and coumarin plasmas at +36°C and +42°C was observed (Table 11). There was no significant change of PT after storage at +5°C. A decrease of PT-ratio was observed for candidate 14/001 after storage at +31°C, +36°C, and +42°C. No significant change of PT-ratio was observed after storage at +5°C (Table 11). For candidate 15/001, a slight increase was observed of the PT of normal and coumarin plasmas stored at +31°C, +36°C, and +42°C (Table 12). There was no significant change of PT and PT-ratio after storage at +5°C. The change of PT-ratio at +36°C was significant (Table 12).
In conclusion, the PT-ratio and hence the ISI of both candidates may change after storage at +31°C, +36°C and +42°C. In contrast, no significant change was observed after storage at +5°C for 56 days, demonstrating that both candidates are stable during shipment at +5°C. For shipment at ambient temperature, however, certain precautions may be advised. The shipment duration
should be limited to a few days. Cooling packs may be included in the parcel to avoid deleterious effects of elevated temperature.
Stability after reconstitution
To assess the stability of the candidates after reconstitution, they were reconstituted with their respective reconstitution fluids and kept at room temperature (+22 ± 1°C) for various time intervals and tested in a PT test. The time intervals were 4, 3, 2, 1, and 0.5 hours, respectively.
For each time interval, three ampoules of each candidate were used. Two deep frozen pooled plasmas (one normal, coded NP141202, and one coumarin plasma obtained from patients treated with VKA, coded AP141204) were used for the PT tests. The tests were performed in duplicate with a semi-automatic, electro-mechanical coagulometer according to Schnitger and Gross manufactured by Amelung GmbH (Lemgo, Germany). Correlation coefficients according to Spearman were used to test the change of PT with incubation time. A significance level of 5%
was used.
Mean PT’s as a function of time interval between reconstitution and testing are shown in Tables 13 and 14. There was no change of the mean PT and PT ratio for candidate 14/001 (Table 13).
There was a slight decrease of the mean PT and PT ratio for candidate 15/001 in the first 3 hours after reconstitution (Table 14). The decrease of PT between 0.5 and 1 hour was greater than that in the following time intervals (Table 14). It is recommended to use the reagent between 0.5 and 4 hours after reconstitution if it is stored at room temperature.
Conclusion
Upon establishment of the proposed candidate preparations, it is proposed to replace the
provisional codes 14/001 and 15/001 by rTF/16 and RBT/16 respectively. This will facilitate the recognition and continuity of terminology for international standards for thromboplastins.
References
1. WHO Expert Committee on Biological Standardization. Guidelines for thromboplastins and plasma used to control oral anticoagulant therapy with vitamin K antagonists. WHO Technical Report Series 2013; No. 979; pp 271-316.
2. Chantarangkul V, van den Besselaar AMHP, Witteveen E, Tripodi A. International collaborative study for the calibration of a proposed international standard for thromboplastin, rabbit, plain. J Thromb Haemost 2006;4:1339-45.
3. Tripodi A, Chantarangkul V, van den Besselaar AMHP, Witteveen E. International collaborative study for the calibration of a proposed international standard for thromboplastin, human, plain. J Thromb Haemost 2010;8:2066-8.
10 Acknowledgements
We wish to thank all participants (see Appendix) of the international collaborative study for their contribution and the ISTH/SSC Subcommittee on Control of Anticoagulation for their support.
We further acknowledge the following contributions:
Provision of candidates: Instrumentation Laboratory (Orangeburg, NY);
Provision of International Standards: National Institute for Biological Standards and Control (Potters Bar, United Kingdom);
Provision of control plasmas: Technoclone Herstellung von Diagnostika und Arzneimitteln GmbH (Vienna, Austria);
Additional support: Stichting Kwaliteitsbewaking Medische Laboratoriumdiagnostiek (Nijmegen, The Netherlands).
Appendix: List of participants of collaborative calibration study 1. F. Angeloni (Hemostasis Reference Laboratory, Hamilton, Canada) 2. N.B. Binder (Technoclone GmbH, Vienna, Austria)
3. M. Byrne (St James’s Hospital, Dublin, Ireland)
4. V. Chantarangkul and A. Tripodi (University of Milano, Milano, Italy) 5. R. Dauer (Alfred Health, Melbourne, Australia)
6. B.R. Gudmundsdóttir (Landspitali University Hospital, Reykjavik, Iceland) 7. K. Gustafsson and T. Lindahl (Universitetssjukhuset, Linköping, Sweden)
8. J. Jespersen and K. Overgaard (Hospital of South West Denmark, Esbjerg, Denmark) 9. S. Kitchen (Royal Hallamshire Hospital, Sheffield, United Kingdom)
10. C. Legnani and M. Cini (University Hospital S. Orsola-Malpighi, Bologna, Italy) 11. R.A. Manning (Hammersmith Hospital, East Acton, London, United Kingdom) 12. M. Martinuzzo (Laboratorio Central del Hospital Italiano, Buenos Aires, Argentina) 13. O. Panes and D. Mezzano (Escuela de Medicina, Univ. Católica de Chile, Santiago,
Chile)
14. V. Pengo and E. Bison (Laboratorio Trombosi – Clinica Cardiologica, Padova, Italy) 15. A. Riddell and A. Drebes (Royal Free Hospital, London, United Kingdom)
16. S. Subramanian (St John’s Medical College Hospital, Bangalore, India) 17. A. Szederjesi (St Laszlo Hospital – Campus, Budapest, Hungary) 18. C. Tantanate (Siriraj Hospital, Mahidol University, Bangkok, Thailand) 19. A.M.H.P. Van den Besselaar (Leiden University Medical Center, Leiden, The
Netherlands)
20. R. Zerback and P. Herbel (Roche Diagnostics GmbH, Mannheim, Germany)
TABLE 1
Mean normal prothrombin time (MNPT) and mean VKA patients’ prothrombin time (s).
Mean values were obtained after excluding outliers (>±3SD) and INR<1.5 and INR>4.5.
MNPT Mean VKA patients’ PT Lab rTF/09 RBT/05 Candidate
14/001
Candidate 15/001
rTF/09 RBT/05 Candidate 14/001
Candidate 15/001 1 14.266 18.064 14.127 19.790 35.428 43.310 34.940 44.095 2 14.205 16.851 14.102 19.193 36.650 45.001 34.072 51.604 3 14.539 16.699 13.940 18.474 34.573 40.893 34.447 44.007 4 11.253 14.367 11.105 14.709 29.292 35.156 28.528 33.800 5 13.199 15.503 12.719 17.090 28.753 35.125 28.014 37.411 6 13.384 15.845 13.290 16.976 31.391 37.415 30.547 40.890 7 12.295 16.328 12.262 17.634 25.935 33.220 25.176 36.500 8 12.101 15.083 11.684 16.535 27.189 35.143 25.426 37.322 9 13.317 16.380 12.703 17.864 30.526 37.060 28.791 38.054 10 13.768 17.738 13.654 18.299 30.789 40.854 29.424 41.340 11 13.159 15.970 12.681 16.976 35.081 38.837 33.285 39.703 12 14.065 16.997 13.547 18.492 32.494 37.584 31.565 39.427 13 13.766 17.722 13.069 18.187 33.357 42.687 31.690 40.794 14 15.490 17.447 14.947 20.294 35.392 40.437 33.802 44.104 15 14.417 17.669 14.188 20.803 38.084 47.223 37.849 48.498 16 12.931 15.938 12.339 16.301 29.776 38.671 28.445 37.271 17 14.236 16.127 14.001 17.913 36.506 39.905 35.303 43.229 18 14.355 16.781 13.803 19.421 36.050 41.255 34.877 46.278 19 15.808 18.734 14.779 20.734 34.470 43.951 32.149 46.402 20 15.262 21.197 15.009 20.877 34.353 43.523 32.975 44.190 Mean 13.79 16.87 13.40 18.33 32.80 39.86 31.57 41.75 Min 11.25 14.37 11.11 14.71 25.93 33.22 25.18 33.80 Max 15.81 21.20 15.01 20.88 38.08 47.22 37.85 51.60
12 TABLE 2
ISI for Candidates 14/001 and 15/001 including all data. N is the number of normal and patients’ samples. CV(b) is the coefficient of variation of the orthogonal regression line slope.
Lab Reference: rTF/09 Reference: RBT/05
Candidate 14/001 Candidate 15/001 Candidate 14/001 Candidate 15/001 N ISI CV(b) N ISI CV(b) N ISI CV(b) N ISI CV(b) 1 80 1.108 2.4 80 1.266 2.7 80 1.078 2.6 80 1.231 1.7 2 80 1.132 1.3 80 1.079 2.1 80 1.289 2.7 80 1.226 2.1 3 80 1.052 2.2 80 1.149 2.9 80 1.094 2.3 80 1.194 1.9 4 80 1.091 1.7 80 1.262 2.6 80 1.046 2.6 80 1.210 2.1 5 80 1.068 1.7 80 1.179 3.0 80 1.096 2.8 80 1.210 2.2 6 80 1.093 2.0 80 1.082 2.4 80 1.152 2.5 80 1.140 1.9 7 80 1.123 1.8 80 1.154 2.5 80 1.124 3.9 80 1.155 3.2 8 80 1.122 0.7 80 1.140 2.2 80 1.206 2.2 80 1.222 1.0 9 80 1.087 1.0 80 1.203 2.0 80 1.117 1.8 80 1.235 1.2 10 79 1.131 2.0 79 1.129 2.9 79 1.188 2.8 79 1.184 1.5 11 80 1.096 1.5 80 1.236 2.9 80 1.053 2.4 80 1.184 1.6 12 80 1.070 1.0 80 1.229 1.7 80 1.043 1.6 80 1.198 1.2 13 80 1.077 1.6 80 1.304 3.2 80 1.039 2.7 80 1.253 1.6 14 80 1.093 1.4 80 1.201 2.2 80 1.139 2.1 80 1.249 1.4 15 80 1.069 0.8 80 1.259 1.2 80 1.126 1.2 80 1.326 0.9 16 80 1.094 2.1 80 1.174 2.7 80 1.155 3.0 80 1.237 1.6 17 80 1.105 1.9 80 1.194 3.0 80 1.080 2.1 80 1.165 2.2 18 80 1.073 1.3 80 1.165 2.1 80 1.115 2.0 80 1.210 1.9 19 80 1.063 2.2 80 1.064 2.8 80 1.211 2.5 80 1.213 2.2 20 88 1.120 1.6 86 1.224 4.0 89 0.988 2.8 87 1.065 3.3
Mean 1.093 1.185 1.117 1.205
CV(interlab) 2.2 5.6 6.4 4.3
TABLE 3
Number of outliers > ±3SD and INR<1.5 and INR>4.5
Lab Reference: rTF/09 Reference: RBT/05
Candidate 14/001 Candidate 15/001 Candidate 14/001 Candidate 15/001
>±3SD <1.5 >4.5 >±3SD <1.5 >4.5 >±3SD <1.5 >4.5 >±3SD <1.5 >4.5
1 1 1 1 1 1 1 0 1 2 2 0 1
2 1 0 1 0 0 9 1 0 1 0 0 6
3 0 1 0 0 1 0 0 1 0 0 0 0
4 0 0 1 1 0 0 0 0 2 1 0 0
5 0 1 1 1 0 0 0 0 0 1 0 0
6 0 0 0 0 0 0 0 0 0 2 0 0
7 0 3 2 0 2 2 0 1 2 0 0 2
8 0 0 0 1 0 1 1 0 0 1 0 0
9 0 3 0 0 1 0 0 1 0 1 0 0
10 1 0 0 0 0 0 0 0 0 0 0 0
11 1 1 0 0 1 0 1 0 1 0 0 1
12 0 1 1 1 0 0 1 0 0 0 0 0
13 1 1 3 0 0 1 1 0 1 0 0 0
14 0 0 0 0 0 0 1 0 0 0 0 0
15 1 0 0 2 0 1 1 0 0 1 0 1
16 1 0 0 1 0 0 0 0 0 1 0 0
17 0 0 0 0 0 0 1 0 0 2 0 0
18 0 0 0 0 0 0 1 0 0 1 0 0
19 1 3 1 0 2 2 2 3 0 1 2 1
20 1 8 1 0 6 0 0 3 1 2 3 0
14 TABLE 4
ISI after excluding outliers >±3SD
Lab ISI values
Candidate 14/001 Candidate 15/001 Reference:
rTF/09
Reference:
RBT/05
Reference:
rTF/09
Reference:
RBT/05
1 1.079 1.070 1.266 1.254
2 1.135 1.289 1.068 1.226
3 1.052 1.094 1.149 1.194
4 1.091 1.037 1.262 1.218
5 1.068 1.086 1.179 1.198
6 1.093 1.152 1.082 1.141
7 1.123 1.124 1.154 1.155
8 1.122 1.218 1.128 1.225
9 1.087 1.117 1.203 1.242
10 1.118 1.188 1.129 1.184
11 1.093 1.053 1.231 1.184
12 1.070 1.046 1.215 1.198
13 1.082 1.039 1.280 1.253
14 1.093 1.139 1.193 1.249
15 1.073 1.130 1.257 1.334
16 1.088 1.171 1.174 1.230
17 1.105 1.080 1.178 1.172
18 1.073 1.115 1.184 1.232
19 1.078 1.211 1.073 1.237
20 1.106 0.988 1.224 1.067
Mean 1.091 1.117 1.181 1.210
CV (interlab) 2.0 6.6 5.4 4.5
TABLE 5
ISI (final) after excluding outliers >±3SD and INR<1.5 and INR>4.5 (Reference: rTF/09).
N is the number of normal and patients’ samples. CV(b) is the coefficient of variation of the orthogonal regression line slope. CV is the between-laboratory coefficient of variation.
Lab Candidate 14/001 Candidate 15/001
N ISI CV(b) N ISI CV(b)
1 77 1.080 1.2 77 1.255 2.8
2 78 1.139 1.2 78 1.066 2.0
3 79 1.054 2.2 79 1.149 2.9
4 79 1.090 1.7 78 1.264 2.7
5 78 1.063 1.7 80 1.179 3.0
6 80 1.093 2.0 80 1.082 2.4
7 75 1.136 1.9 76 1.138 2.7
8 80 1.122 0.7 79 1.128 2.0
9 77 1.088 1.0 79 1.208 2.0
10 78 1.118 1.9 79 1.129 2.9
11 78 1.095 1.4 78 1.242 2.7
12 78 1.069 1.0 79 1.215 1.6
13 75 1.074 1.2 78 1.274 3.0
14 80 1.093 1.4 79 1.193 2.1
15 79 1.073 0.7 79 1.257 1.2
16 79 1.088 1.9 80 1.174 2.7
17 80 1.105 1.9 79 1.178 2.8
18 80 1.073 1.3 79 1.184 2.0
19 75 1.082 2.2 75 1.073 2.4
20 78 1.110 1.4 82 1.206 4.2
Mean 1.092 1.18
CV (%) 2.1 5.4
16 TABLE 6
ISI (final) after excluding outliers >±3SD and INR<1.5 and INR>4.5 (Reference: RBT/05).
N is the number of normal and patients’ samples. CV(b) is the coefficient of variation of the orthogonal regression line slope.
Lab Candidate 14/001 Candidate 15/001
N ISI CV(b) N ISI CV(b)
1 77 1.075 2.5 77 1.254 1.4
2 71 1.357 2.6 74 1.228 2.3
3 79 1.096 2.3 80 1.194 1.9
4 79 1.037 2.3 79 1.218 1.9
5 79 1.086 2.5 79 1.198 1.7
6 80 1.152 2.5 78 1.141 1.5
7 77 1.146 4.2 78 1.153 3.4
8 78 1.233 2.0 79 1.225 1.0
9 79 1.119 1.8 79 1.242 1.1
10 79 1.188 2.8 79 1.184 1.5
11 79 1.056 2.4 79 1.189 1.6
12 79 1.046 1.5 80 1.198 1.2
13 79 1.050 2.7 80 1.253 1.6
14 80 1.139 2.1 80 1.249 1.4
15 77 1.130 1.1 78 1.334 0.9
16 79 1.171 2.8 79 1.230 1.5
17 80 1.080 2.1 78 1.172 1.8
18 80 1.115 2.0 79 1.232 1.7
19 76 1.240 2.6 76 1.240 2.1
20 83 0.994 2.9 82 1.062 2.9
Mean 1.125 1.210
CV (%) 7.5 4.5
TABLE 7
Parameters of orthogonal regression line for patients’ samples only
Lab Reference: rTF/09 Reference: RBT/05
Candidate 14/001 Candidate 15/001 Candidate 14/001 Candidate 15/001
a´ b´ a´ b´ a´ b´ a´ b´
1 0.138 0.965 -1.266 1.277 0.958 0.791 -0.242 1.060
2 0.134 0.983 -1.237 1.223 0.233 1.034 -1.197 1.281
3 -0.164 1.047 -1.928 1.446 0.899 0.795 -0.351 1.073 4 -0.003 1.009 -1.189 1.295 0.667 0.862 -0.280 1.088
5 0.107 0.976 -1.731 1.406 1.060 0.748 -0.281 1.059
6 0.264 0.931 -1.309 1.282 0.828 0.817 -0.667 1.155
7 -0.367 1.123 -1.332 1.274 0.500 0.936 -0.493 1.113 8 -0.007 1.023 -1.727 1.388 0.739 0.876 -0.621 1.158
9 0.140 0.976 -1.289 1.289 0.844 0.828 -0.344 1.087
10 0.094 0.986 -1.791 1.402 1.081 0.777 -0.353 1.092 11 0.107 0.984 -0.953 1.220 0.654 0.859 -0.168 1.036 12 0.075 0.987 -1.489 1.350 0.846 0.803 -0.441 1.104 13 0.140 0.974 -2.656 1.661 1.398 0.679 -0.429 1.127 14 0.004 1.012 -1.525 1.344 0.824 0.817 -0.431 1.091 15 0.002 1.001 -1.103 1.224 0.442 0.936 -0.480 1.117 16 -0.056 1.030 -2.101 1.520 1.114 0.761 -0.180 1.059 17 -0.072 1.030 -1.263 1.288 1.043 0.742 -0.375 1.075 18 0.081 0.987 -1.131 1.230 0.320 0.957 -0.846 1.191 19 0.059 1.003 -1.384 1.283 0.697 0.891 -0.755 1.186 20 0.023 1.005 -2.033 1.460 1.416 0.675 -0.115 1.016
18 TABLE 8
Deviation (%) of INR values using orthogonal regression line parameters a´ and b´
Lab Candidate 14/001 Candidate 15/001
Reference: rTF/09 Reference: RBT/05 Reference: rTF/09 Reference: RBT/05 INR = 2 INR=4.5 INR = 2 INR=4.5 INR = 2 INR=4.5 INR = 2 INR=4.5
1 -1.6 1.2 -8.5 6.0 4.9 -2.6 -1.4 0.9
2 -4.4 1.3 -8.7 2.3 11.3 -4.9 11.8 -2.3
3 3.4 -2.4 -10.0 5.8 13.3 -8.7 1.7 -0.9
4 -0.6 -0.7 -3.0 0.7 5.9 -2.2 1.0 -1.1
5 -0.4 0.1 -13.0 7.6 11.9 -6.8 0.7 -0.6
6 -3.2 3.7 -9.2 9.1 8.2 -9.5 6.1 -5.4
7 1.7 -3.5 -1.8 3.6 4.7 -9.1 2.6 -5.3
8 -0.6 0.6 -9.3 8.7 9.8 -10.3 3.8 -2.7
9 -1.3 1.2 -8.0 6.0 6.0 -4.9 0.2 -0.4
10 -2.1 1.9 -13.0 13.6 9.4 -11.1 2.1 -2.5
11 -1.5 0.8 -3.2 2.3 3.5 -1.3 0.4 0.3
12 -0.2 -0.1 -4.5 6.3 5.9 -7.7 2.0 -2.6
13 -1.1 0.4 -16.3 10.7 16.3 -8.2 1.7 -1.0
14 -0.2 -0.3 -9.1 7.9 7.4 -7.1 0.3 -0.0
15 0.5 -0.2 -3.0 0.9 2.9 -1.3 -2.5 0.6
16 1.1 -0.9 -14.9 11.9 13.3 -10.1 -0.3 0.4
17 0.6 -0.1 -12.7 8.3 8.0 -4.7 2.2 -2.1
18 -0.3 0.1 -0.6 0.5 6.6 -2.6 6.7 -1.6
19 -0.2 -0.5 -8.4 8.5 5.6 -12.2 3.1 -4.2
20 -0.8 0.9 -8.9 14.4 10.0 -9.1 2.3 -5.0
Mean -0.6 0.2 -8.3 6.8 8.2 -6.7 2.2 -1.8
TABLE 9
Mean ISI before and after exclusion of non-valid assessments. SD and CV is the between- laboratory standard deviation and coefficient of variation, respectively. SE is the standard error of the mean (= SD/√N).
Lab Before exclusion After exclusion
Candidate 14/001 Candidate 15/001 Candidate 14/001 Candidate 15/001 Ref:
rTF/09
Ref:
RBT/05
Ref:
rTF/09
Ref:
RBT/05
Ref:
rTF/09
Ref:
RBT/05
Ref:
rTF/09
Ref:
RBT/05
1 1.080 1.075 1.255 1.254 1.080 1.075 1.255 1.254
2 1.139 1.357 1.066 1.228 1.139 1.357
3 1.054 1.096 1.149 1.194 1.054 1.096 1.194
4 1.090 1.037 1.264 1.218 1.090 1.037 1.264 1.218
5 1.063 1.086 1.179 1.198 1.063 1.198
6 1.093 1.152 1.082 1.141 1.093 1.152 1.082 1.141
7 1.136 1.146 1.138 1.153 1.136 1.138
8 1.122 1.233 1.128 1.225 1.122 1.233 1.225
9 1.088 1.119 1.208 1.242 1.088 1.119 1.208 1.242
10 1.118 1.188 1.129 1.184 1.118 1.184
11 1.095 1.056 1.242 1.189 1.095 1.056 1.242 1.189
12 1.069 1.046 1.215 1.198 1.069 1.046 1.215 1.198
13 1.074 1.050 1.274 1.253 1.074 1.253
14 1.093 1.139 1.193 1.249 1.093 1.139 1.193 1.249
15 1.073 1.130 1.257 1.334 1.073 1.130 1.257 1.334
16 1.088 1.171 1.174 1.230 1.088 1.230
17 1.105 1.080 1.178 1.172 1.105 1.178 1.172
18 1.073 1.115 1.184 1.232 1.073 1.115 1.184 1.232
19 1.082 1.240 1.073 1.240 1.082 1.240 1.240
20 1.110 0.994 1.206 1.062 1.110 1.062
Mean 1.092 1.125 1.180 1.210 1.092 1.138 1.201 1.212
CV (%) 2.1 7.5 5.4 4.5 2.1 8.0 4.6 4.6
Mean all 1.109 1.195 1.110 1.208
SD 0.0632 0.0605 0.0629 0.0552
CV (%) 5.7 5.1 5.7 4.6
SE 0.0100 0.0096 0.0109 0.0103
20 TABLE 10
Mean INR and between-laboratory standard deviation for control plasmas A, B, C, and D
Plasma A Plasma B Plasma C Plasma D
rTF/09 0.95 ± 0.05 1.99 ± 0.12 2.92 ± 0.19 4.35 ± 0.35 Candidate 14/001 0.94 ± 0.03 2.00 ± 0.11 3.00 ± 0.19 4.51 ± 0.35 RBT/05 0.98 ± 0.04 2.18 ± 0.14 2.96 ± 0.20 4.18 ± 0.28 Candidate 15/001 0.98 ± 0.04 2.22 ± 0.12 3.03 ± 0.19 4.26 ± 0.34
TABLE 11
Mean prothrombin time and prothrombin time ratio with candidate 14/001 after storage at various temperatures. The PT ratio was calculated as the mean PT of plasma AP141204 divided by the mean PT of plasma NP141202.
Storage temperature: +5°C Storage temperature: +31°C Storage
time (d)
Mean PT (s) PT ratio Storage time (d)
Mean PT (s) PT ratio
NP141202 AP141204 NP141202 AP141204
0 12.33 32.27 2.617 0 12.33 32.90 2.668
1 12.30 31.80 2.585 1 12.30 31.70 2.577
2 12.30 32.03 2.604 2 12.23 31.37 2.565
3 12.30 31.97 2.599 3 12.17 31.33 2.574
4 12.07 32.07 2.657 4 12.30 30.23 2.458
7 12.77 32.57 2.551 7 12.40 30.33 2.446
14 12.27 31.60 2.575 14 12.27 30.70 2.502
21 12.30 32.93 2.677 21 11.93 29.93 2.509
28 12.33 32.73 2.655 28 12.47 29.17 2.339
42 12.57 32.07 2.551 42 12.23 29.40 2.404
56 12.30 32.33 2.628 56 12.27 29.60 2.412
p-value‡ 0.695 0.174 0.770 p-value 0.687 <0.001 <0.001 Storage temperature: +36°C Storage temperature: +42°C Storage
time (d)
Mean PT (s) PT ratio Storage time (d)
Mean PT (s) PT ratio
NP141202 AP141204 NP141202 AP141204
0 12.53 33.23 2.652 0 12.37 32.20 2.603
1 12.13 30.33 2.500 1 11.9 29.50 2.479
2 12.00 29.67 2.473 2 11.63 29.57 2.543
3 12.07 29.93 2.480 3 11.83 28.93 2.445
4 11.77 29.57 2.512 4 11.83 28.47 2.407
7 11.60 28.67 2.472 7 11.20 27.73 2.476
14 11.37 28.83 2.536 14 11.47 27.20 2.371
21 11.70 28.30 2.419 21 11.20 27.10 2.420
28 11.67 27.83 2.385 28 11.00 26.20 2.382
42 11.33 27.20 2.401 42 10.80 25.43 2.355
56 11.23 26.93 2.398 56 11.07 26.67 2.409
p-value‡ <0.001 <0.001 0.005 p-value <0.001 <0.001 0.004
‡ Spearman’s bivariate correlation between storage time and mean PT or PT ratio
22 TABLE 12
Mean prothrombin time and prothrombin time ratio with candidate 15/001 after storage at various temperatures. The PT ratio was calculated as the mean PT of plasma AP141204 divided by the mean PT of plasma NP141202.
Storage temperature: +5°C Storage temperature: +31°C Storage
time (d)
Mean PT (s) PT ratio Storage time (d)
Mean PT (s) PT ratio
NP141202 AP141204 NP141202 AP141204
0 18.30 41.77 2.282 0 18.2 41.37 2.273
4 18.23 40.7 2.232 3 18.57 41.57 2.238
5 18.43 41.5 2.251 4 18.97 42.97 2.265
6 18.40 41.5 2.255 5 18.7 43.13 2.307
7 18.20 40.63 2.233 6 18.73 43.4 2.317
9 18.6 41.23 2.217 8 18.73 42.97 2.294
17 18.63 41.4 2.222 15 19.37 44.47 2.296
24 18.2 41.2 2.264 22 19.27 44.2 2.294
31 18.5 41.1 2.222 28 19.97 45.67 2.287
44 18.4 41.67 2.264 43 19.97 45.87 2.297
59 18.47 41.73 2.260 57 20.27 47.63 2.350
p-value‡ 0.303 0.926 0.915 p-value <0.001 <0.001 0.056 Storage temperature: +36°C Storage temperature: +42°C Storage
time (d)
Mean PT (s) PT ratio Storage time (d)
Mean PT (s) PT ratio
NP141202 AP141204 NP141202 AP141204
0 18.1 40.87 2.258 0 18.1 41.43 2.289
2 18.87 42.53 2.254 1 19.3 43.53 2.256
3 19.1 43.83 2.295 2 19.23 43.53 2.263
4 19.9 43.8 2.201 3 19.67 44.23 2.249
5 19.1 43.77 2.291 4 19.57 44.4 2.269
7 19.03 43.63 2.293 6 19.4 44.23 2.280
15 19.83 45.8 2.310 13 20.1 45.83 2.280
22 20.07 46.57 2.321 20 21.1 47.53 2.253
29 20.07 46.77 2.331 26 21.33 48.7 2.283
42 20.5 47.57 2.320 41 21.43 50.73 2.367
57 21.37 49.93 2.337 55 22.17 51.97 2.344
p-value‡ <0.001 <0.001 0.001 p-value <0.001 <0.001 0.133
‡ Spearman’s bivariate correlation between storage time and mean PT or PT ratio
TABLE 13
Mean prothrombin time and prothrombin time ratio at various times after reconstitution of candidate 14/001. The PT ratio was calculated as the mean PT of plasma AP141204 divided by the mean PT of plasma NP141202.
Time (h) Mean PT ± standard deviation (s) PT ratio
NP141202 AP141204 AP/NP
0.5 12.33 ± 0.31 33.42 ± 0.82 2.709
1 12.50 ± 0.18 33.87 ± 0.54 2.709
2 12.60 ± 0.11 33.65 ± 0.24 2.671
3 12.43 ± 0.27 34.08 ± 0.41 2.741
4 12.40 ± 0.30 33.87 ± 0.39 2.731
p-value‡ 0.873 0.219 0.391
‡ Spearman’s bivariate correlation between storage time and mean PT or PT ratio
24 TABLE 14
Mean prothrombin time and prothrombin time ratio at various times after reconstitution of candidate 15/001. The PT ratio was calculated as the mean PT of plasma AP141204 divided by the mean PT of plasma NP141202.
Time (h) Mean PT ± standard deviation (s) PT ratio
NP141202 AP141204 AP/NP
0.5 17.83 ± 0.28 42.15 ± 0.23 2.364
1 17.60 ± 0.32 41.17 ± 0.62 2.339
2 17.55 ± 0.15 40.97 ± 0.39 2.334
3 17.40 ± 0.14 40.38 ± 0.53 2.321
4 17.47 ± 0.21 40.55 ± 0.40 2.322
p-value‡ 0.037 0.037 0.037
‡ Spearman’s bivariate correlation between storage time and mean PT or PT ratio
26
28
30
