EXPERT COMMITTEE ON BIOLOGICAL STANDARDIZATION Geneva, 12 to 16 October 2015

Transcription

1 EXPERT COMMITTEE ON BIOLOGICAL STANDARDIZATION Geneva, 12 to 16 October 2015 WHO/BS/ ENGLISH ONLY Collaborative Study to Evaluate the Proposed 5 th WHO International Standard for Hepatitis C Virus (HCV) RNA for Nucleic Acid Amplification Technique (NAT)-Based Assays Clare Morris 1,3, Graham Prescott 1, Jason Hockley 2 and the Collaborative Study Group* 1 Division of Virology and 2 Biostatistics, National Institute for Biological Standards and Control, South Mimms, Potters Bar, Herts, EN6 3QG, UK 3 Study Coordinator; Tel , Fax , Clare.Morris@nibsc.org * See Appendix 8 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 14 September 2015 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 M Nübling at nueblingc@who.int World Health Organization 2015 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 The previous 4 th HCV IS for NAT (NIBSC code 06/102) was established in 2011, however it was recognised that the material must be shipped on dry ice in order to prevent loss of titre during an ambient shipping process. A small pilot study was conducted at NIBSC to determine the possible effect that the presence or absence of HCV antibodies may have on the stability of a lyophilised preparation. Following the assessment of six lyophilised formulations containing lyoprotectants and stabilizing buffers in the presence and absence of antibodies to HCV, by six laboratories using different commercial assays, no differences were observed. It was subsequently agreed at the 3 rd joint meeting of the Blood Virology and Clinical Diagnostics Standardisation of Genomic Amplification Techniques meeting held on 30/31st May 2014 in Graz, Austria that a candidate standard for the 5 th HCV IS NAT would comprise an antibody negative preparation of genotype 1a. The candidate 5 th HCV IS for NAT (NIBSC code 14/150) was filled and freeze dried at NIBSC in November 2014, 1980 vials were lyophilised with a reconstitution volume of 1.1mL to allow for single vial use on analysers requiring a 1mL input volume. The subsequent international collaborative study was performed from December 2014 to March laboratories from 11 countries participated in the evaluation of 6 samples including the lyophilised candidate standard and the corresponding frozen liquid bulk. HCV antibody positive plasma and a low titre material intended for use as a secondary reference material were also assessed. Assays comprised mainly of commercial real time PCR methods reporting in both quantitative and qualitative values. The current 4 th HCV IS and the previous 2 nd IS were both included giving the option of deriving potency comparative to the 2 nd standard. The range in individual laboratory mean estimates for the candidate material (S5) in quantitative assays was log 10 IU/mL and log 10 IU/mL in qualitative assays. Standard deviation values for intra laboratory comparisons were low. Accelerated thermal degradation studies indicate good stability at 3 months however at 6 months a loss of 0.14Log is observed at +20 C relative to the -20 C baseline sample. Further studies on real time stability are on-going. Overall, the results of this study indicate the suitability of the candidate to be established as the replacement 5th WHO International Standard for HCV (NIBSC code 14/150) with an assigned potency of 5.0 log 10 IU/vial when reconstituted in 1.1mL of nuclease-free water. Introduction The availability of an international standard for HCV RNA for the past 20 years (1-7) has been fundamental in driving the close agreement now seen across HCV nucleic acid tests, giving assurance of results regardless of assays used. HCV RNA assays are used in both a screening and diagnostic capacity, with regulations in developed countries mandating the screening of donated blood and plasma for the presence of HCV RNA, whilst in the field of biotherapeutic medicines (8) European Pharmacopeia guidelines dictate that constituent plasma pools must be

3 Page 3 screened for the presence of HCV RNA and that all assays used for such a purpose must have a sensitivity of at least 100 IU/mL (9). Furthermore, with no therapeutic or prophylactic vaccine available there is a global need to use HCV NAT assays for the diagnosis of patients and the monitoring of those receiving antiviral therapy (8). In both cases accurate and comparable quantification is required across all HCV NAT systems. The 4 th WHO HCV NAT IS (NIBSC code 06/102) was established in 2011 (5); however the product was established with the knowledge that the material was not able to maintain titre when shipped to customers at ambient temperature, a practice commonly employed for lyophilised standards to reduce the shipping costs for the customer. The material did however demonstrate good stability when shipped on dry ice. The 4 th HCV IS was established with the understanding that further stability investigation would be carried out and a 5 th standard developed in advance of the usual time frame of depleted stocks of the previous standard. A difference in stability has been previously reported (5) between the 1 st (99/790) and 2 nd (99/798) established preparations of the HCV NAT IS and the 3 rd (06/100) and 4 th (06/102) preparations. The most notable difference between these is the presence of HCV antibody in preparations 1 and 2 but its absence in preparations 3 and 4. A pilot study was conducted to determine whether the presence or absence of HCV antibody affected the stability of the product. This report describes both the pilot study conducted to investigate the stability issues observed in the previous standard and following the outcome of this study the development and evaluation of the replacement batch of standard to be established as the 5 th WHO International Standard for HCV RNA for NAT. Data from the pilot study were presented at the XXIV Scientific Working Group on the Standardisation of Genome Amplification Techniques for the Safety Testing of Blood and Clinical Diagnostics meeting in Graz, Austria in May The collaborative study to establish the 5 th HCV NAT IS was presented to the 20 th Anniversary meeting of the Scientific Working Group on the Standardisation of Genome Amplification Techniques for the Safety Testing of Blood and clinical diagnostics meeting held on 25/26 th June in London, UK. In line with previous standards, an international unit derived as the result of a comprehensive collaborative study incorporating a range of different assays will be assigned to this material. As with the majority of biological preparations, the material cannot be fully characterized by a physico-chemical reference method. Therefore, biological assays will be used, these methods are heterogeneous and the lack of a reference method does not permit the results to be expressed in absolute values according to the SI system. Pilot study to investigate stability of freeze-dried HCV formulations Background

4 Page 4 WHO HCV NAT IS stability discrepancies have previously been reported (5). Briefly the 1 st (99/790) and 2 nd (99/798) established preparations of the HCV NAT IS displayed good stability throughout their use, whereas a loss of titre over time was observed in the 3 rd (06/100) and 4 th (06/102) IS preparations. The 1 st and 2 nd standards were prepared from identical stock material and lyophilised on two separate occasions. The stock material used to formulate the 3 rd and 4 th standard was different from that used for the 1 st and 2 nd standard, however similarly both 3 rd and 4 th standards were produced from the same stock material and lyophilised on two separate occasions. The main notable difference between the preparations was the presence of HCV antibody in preparations 1 and 2 but its absence in preparations 3 and 4. High moisture and oxygen was observed in all preparations. A pilot study was conducted to determine whether excessive moisture in conjunction with the presence or absence of HCV antibody affected the stability of lyophilised HCV RNA. The pilot study was performed between January 2014 and May Six different HCV RNA candidate formulations were evaluated for stability through accelerated thermal degradation studies and subsequently assessed by laboratories using different commercial assays. Source candidate material The study compared six different HCV RNA preparations comprising of two main groups. 1) HCV window period positive donation, diluted in negative human plasma. 2) Pooled antibody positive plasma donations. Six individual HCV RNA positive plasma packs were purchased from the UK NHS Blood and Transplant Authority. Each pack had previously been screened using the UK blood donation screening algorithm and rejected due to a positive HCV NAT signal. Titre determination was carried out at NIBSC using the COBAS AmpliPrep/COBAS TaqMan HCV v2.0 assay. Three plasma packs, containing a titre that when pooled, would result in a final concentration of approximately 6 log 10 IU/mL of HCV RNA, were selected. The final volume available for the study, following pooling, was 320mL. This material was used to formulate the HCV antibody positive preparations. 800mL of HCV RNA positive, HCV antibody negative (window period) plasma was donated by Public Health England, UK. This was confirmed to be antibody negative at NIBSC using the Murex EIA and was confirmed to have a titre of 6 log 10 IU/mL using the COBAS AmpliPrep/COBAS TaqMan HCV v2.0 assay. Table 1 outlines each candidate formulation. Formulation of candidate pilot study material Antibody negative material

5 Page 5 A 1/10 dilution into HCV RNA negative, HCV antibody negative plasma was performed on the window period donation to give a titre in the region of 5 log 10 IU/mL. 3x 80mL of material was aliquoted into 120mL sterile bottles and labelled 2, 4 and 6. Antibody positive material In order to achieve a titre of 1x10 5 IU/mL as the final titre of the antibody positive material, six HCV positive plasma packs were tested using the Roche COBAS AmpliPrep/COBAS TaqMan HCV v2.0 assay Two HCV antibody positive plasma packs were selected or intial pooling to give a concentration of 1.44x10 3 IU/mL, the flask was labelled pool 6. Pool 6 was then combined with a single HCV antibody positive donation, with atitre of 8.93x10 5 IU/mL. 100mLs of Anti HCV positive plasma at a concentration of 1x10 6 IU/mL was produced and comprised the HCV antibody positive preparation for this pilot study. 3x 80mL of the pooled HCV antibody positive plasma was aliquoted into 120mL sterile bottles and labelled 1, 3 and 5. To assess any stabilising effect during lyophilisation, preparations were made with the presence or absence of Hepes buffer and trehalose. Hepes buffer, at a final concentration of 40mmol/L, was added to bottles 3 and 4. Trehalose, at a final concentration of 10mg/L, was added to bottles 5 and 6. The remaining bottles, 1 and 2, served as unstabilsed control material for each AB+ and AB- preparation. All preparations were mixed for 10 minutes thoroughly to ensure homogeneity. 38 x 0.55mL of each preparation was aliquoted in to 2mL sarstedt vials and frozen at C to provide aliquid frozen baseline sample. Lyophilisation of pilot study candidates Lyophilisation was conducted by Public Health England, Colindale, UK. All preparations were lyophilised in the same run using a three day plasma cycle. Preparations were then returned to NIBSC on dry ice. Accelerated degradation of samples Prior to placing samples at different temperatures, half of each batch of preparation was exposed to atmospheric conditions to deliberately allow the introduction of moisture and oxygen. The vials were placed in a microbiological safety cabinet; stoppers were lifted but not fully removed and the vials left for 60 minutes. After which time all stoppers were closed fully and the screw cap lids replaced. In total 25 samples from each of the six lyophilised preparations were placed at C and C. At the one month time point samples were removed, 3 vials of each temperature were assayed by participants in a small collaborative study. Vials of the current 4 th HCV NAT IS were

6 Page 6 evaluated in parallel. 56 vials of 06/102 were placed at C for one month. The labelling of tubes is summarised in table 1. Statistical analysis Pilot collaborative study Six laboratories, each using a different, commercially available HCV RNA assay were invited to participate in a small collaborative study to assess the lyophilised preparations from each temperature. Frozen liquid bulk material corresponding to each formulation was also included. Study participants and assays used can be found in appendices 1 and 2. Participants were sent up to a total of 28 lyophilised preparations and 6 frozen liquid preparations (Vial numbers varied depending on the volume requirement of the participants chosen assay). Each participant was requested to perform two independent assays, using a fresh vial of material each time. For quantitative assays participants were requested to test all vials neat. For qualitative assays, it was recommended that laboratories tested each sample neat and at up to four 10-fold dilutions in the first assay to establish the end point and at half log dilutions around the end point in subsequent assays. A copy of the study protocol is provided in appendices 3-6. Results Pilot study The mean of all data received from all participants for the analysis of the pilot study samples A- G are presented in table 2 Observed differences in titre between liquid bulk samples and lyophilised preparations are also shown. The mean loss in titre due to the lyophilisation of material of the antibody positive preparations is 0.53 log 10 IU/mL whilst for antibody negative preparations it is 0.22 log 10 IU/mL. The mean differences between the presence or absence of moisture or antibodies ranges from to log 10 IU/mL, this difference is considered within the limits of precision of the assays used. Statistical analysis Statistical analysis using the Arrhenius equation, frequently used to predict loss of titre following long term storage at -20 through the short term loss at higher temperatures, was not possible due to the negligible difference in titre between each sample types. Discussion Pilot Study

7 Page 7 All panel members used in this study were formulated to contain approximately 1x 10 5, however some loss was experienced during lyophilisation giving a range of titres from 4.89 log10 IU/mL to 5.13 log10 IU/mL. All panel members were compared relative to their C counterpart. Data from the pilot study indicated no loss of titre of HCV RNA across all temperatures and formulations used. Given the time period available, a timeframe of one month was chosen to store pilot formulations at C and C. In previous HCV IS formulations a reduction in titre was observed at temperatures up to 20 0 C over a few days during shipping. By storing samples for one month it was felt that this would go over and beyond previous scenarios where by a loss in titre was observed during the shipping process. In previous HCV IS stability studies it has been demonstrated that there has been a loss in titre during routine transport at ambient temperature. In this study a temperature of C represents an average ambient shipping temperature, to hold a sample at this temperature for a period of one month exceeds a usual shipping timeframe and therefore it was expected that if a loss is titre was to be observed, such a time frame would be suitable to witness this. However as no loss was seen in any sample, it can be concluded that the presence or absence of HCV antibodies did not influence the stability of HCV RNA in lyophilised preparations. Indeed the addition of stabilising compounds also displayed no effect. Due to the availability of sufficient window period donation at NIBSC it was proposed that the 5 th HCV RNA IS would be formulated from this donation in order to avoid delays in sourcing sufficient qualities of antibody positive plasma. Delegates of the SoGAT workshop held in Graz, Austria in May 2014 agreed an antibody negative HCV RNA plasma preparation would be fit for purpose and development of such a standard should continue in a timely manner.

8 Page 8 Establishment of a 5th International Standard for HCV RNA for use in NAT - Main collaborative study Introduction Following a pilot study (described above) to investigate the stability of lyophilised HCV RNA, it was agreed at the 3 rd combined Blood Virology and Clinical Diagnostic SoGAT workshop, held in Graz, Austria (30/31st May 2014) that the 5 th HCV RNA IS for NAT should be formulated from a window period donation and comprise HCV antibody negative material spiked into human plasma. It was also agreed that in a change to previous batches of HCV IS, the 5th batch would be formulated in 1.1mL aliquots to allow for undiluted use on current automated analysers that require a 1mL input volume. Material and Methods Candidate standard The proposed candidate standard comprises (NIBSC code 14/150) a single window period donation, confirmed by Sanger sequencing to be HCV genotype 1a (sample and sequence information provided by Public Health England, Colindale UK) spiked into human plasma tested and found negative for HCV, and HIV-1/2 antibodies, HBsAg and HAV. Negative human plasma sourced from the National Blood Service, UK (NBS) in the form of individual plasma packs comprised the diluent for this standard. 12 packs were pooled and refrozen until required. Each plasma pack was previously screened by the NBS and found negative for HIV antibodies, HCV RNA, HBsAg and Syphilis. Once pooled the diluent was further tested at NIBSC and confirmed negative for HIV-1/2 antibodies, HBsAg and HCV RNA. Preparation of bulk material 300mL of window period material was stored at C until required. Previous titre determination using the COBAS AmpliPrep/COBAS TaqMan HCV v2.0 assay, confirmed the bulk stock material to be 6 log 10 IU /ml. Bulk stock material was thawed in a water bath at 37 0 C, with occasional mixing to encourage even heat distribution and increase the thawing process. Following thawing the stock was mixed using a magnetic stirrer and flea for 30 minutes to disperse any aggregates and ensure homogeneity. Stirring also aided the breakdown of residual lipid clots. The bulk stock was diluted in human plasma sourced from the UK Blood Transfusion and Transplant service, all plasma donations were screened and found negative for anti-hiv-1, HBsAg, and HCV RNA. 280mL of bulk stock was diluted into 2220mL of negative plasma. 2500mL of final preparation was formulated to contain approximately 1x10 5 HCV log 10 IU/mL. The final formulation was stirred continuously for 30mins using a magnetic stirrer and flea.

9 Page 9 200mL of liquid bulk stock was dispensed in 1.1mL aliquots into 2mL Sarstedt vials and frozen at C. The remaining bulk preparation was transferred to the Centre for Biological Reference Materials, located at NIBSC for filling and lyophilisation. The final product was coded, NIBSC code 14/150. Filling and lyophilisation of the candidate standard Filling and lyophilisation was performed onsite at NIBSC by the Centre for Biological Reference Materials (CBRM), Potters Bar, UK in September The filling was performed in a Metal and Plastic GmbH (Radolfzell, Germany) negative pressure isolator that contains the entire filling line and is interfaced with the freeze dryer (CS150 12m2, Serail, Arguenteil, France) through a pizza door arrangement to maintain integrity of the operation. The bulk material was dispensed in 1.1mL volumes into 5mL screw cap glass vials using a Bausch & Strobel (Ilfshofen, Germany) filling machine FVF5060. The bulk material stirred constantly using a magnetic stirrer. The homogeneity of the fill was determined by on-line check-weighing of the wet weight, a tolerance of to 1.108g with a target of 1.100g was applied; a sample of 3% was check weighed to determine the homogeneity of the fill. Those outside of this tolerance were discarded. Filled vials were partially stoppered with halobutyl 14mm diameter cruciform closures and lyophilized in a CS150 freeze dryer. Vials were loaded onto the shelves at -50 C and held at this temperature for 4 hrs. A vacuum was applied to 270 ub over 1 hr, followed by ramping to 30 ub over 1 hr. The temperature was then raised to -40 C, and the vacuum maintained at this temperature for 42.5 hrs. The shelves were ramped to 25 C over 15 hrs before releasing the vacuum and back-filling the vials with nitrogen. The vials were then stoppered in the dryer, removed and capped in the isolator, and the isolator decontaminated with formaldehyde before removal of the product. A total of 1980 vials were filled; these vials were then stored at C. Continuous temperature monitoring will proceed for the lifespan of the product. Material characteristics can be found in table 2. Post-fill testing Assessments of residual moisture and oxygen content, as an indicator of vial integrity after sealing, were determined for twelve vials of the freeze-dried product. Residual moisture was determined by non-invasive near-infrared (NIR) spectroscopy (MCT 600P, Process Sensors, Corby, UK). NIR results were then correlated to Karl Fischer (using calibration samples of the same excipient, measured using both NIR and Karl Fischer methods) to give % w/w moisture readings. Oxygen content was measured using a Lighthouse Infra-Red Analyzer (FMS-750, Lighthouse Instruments, Charlottesville, USA). The CV of fill mass and mean residual moisture and Oxygen were within WHO acceptable limits; details can be found in table 3. Stability of the freeze-dried candidate Accelerated degradation studies are currently on going at NIBSC, where the stability of this material can be predicted when stored at C. Vials of the freeze-dried product were stored at -

10 Page C, C, +4 0 C, C, C and C in November 2014 and have been tested at 3 and 6 months. At the three month time point two vials were removed from each temperature, reconstituted and pooled prior to extraction and amplification. At the 6 month time point, three vials were removed from each temperature, reconstituted and tested individually. At both time points the HCV RNA concentration was quantified by NAT using the COBAS AmpliPrep/COBAS TaqMan HCV v2.0. Study samples The lyophilised candidate material 14/150 was evaluated in parallel to the 2 nd and 4 th WHO International Standard for HCV for Nucleic Acid Techniques (NIBSC codes 96/798 and 06/102 respectively). Alongside these materials, 3 liquid preparations were provided in this study; a sample of the candidate liquid bulk material, a HCV antibody positive commutability sample comprising a single donation HCV antibody positive sample purchased from the UK National Blood Authority and a candidate working reagent for HCV (14/ ), consisting of a clinical samples spiked into HCV negative human plasma. Before shipping, liquid preparations were stored at C, whereas freeze-dried materials were stored at C. Collaborative study (NIBSC code CS534) samples were shipped to all participants frozen (on dry ice). Samples were blinded and coded 1-6 as shown below: Sample 1 Candidate HCV International standard liquid bulk material in a 2mL Sarstedt tube Sample 2 HCV antibody positive commutability liquid material in a 2mL Sarstedt tube Sample 3 Lyophilized preparation 96/798 in a 3mL crimp cap vial Sample 4 Lyophilized preparation 06/102 in a 3 ml crimp cap vial Sample 5 Lyophilized preparation 14/150 in a 5mL screw cap vial Sample 6 HCV candidate working reagent 14/ in a 2mL Sarstedt tube Study Design The aim of this collaborative study was to establish a replacement material for the 4th WHO International Standard for HCV RNA (NIBSC code 06/102) and assign a unitage based on comparisons to both 2nd and 4th WHO HCV International Standards. This study also includes a Working Reagent candidate material, which will be calibrated against these preparations in a range of NAT-based assays. Up to 10 vials of samples 1-6 were sent to participants on dry ice. Vial numbers varied depending on the volume requirement of the participants chosen assay. Participants Study samples were sent to 17 participants across 11 countries (Appendix 7). Participants were selected for their experience in previous collaborative (HCV NAT) studies and their representative distribution geographically; providing a range of assays across the testing field (assays used are shown in appendix 8). These included manufacturers of in vitro diagnostic devices (IVD s), control and reference laboratories and blood transfusion centres. Laboratories were designated a number code at random and where a specific laboratory performed more than one assay their data was analysed separately. These labs were given a decimal number code, for

11 Page 11 example Laboratory 17.1, 17.2 etc. Appendix 7 does not represent the order of designated laboratory numbers. Study protocol Participants were requested to test samples 1-6 in their routine HCV NAT-based assays; in three independent tests. Each participant was given sufficient material for their assay, which had previously been established through the participant invitation. Where assays required a higher volume than that of a specific sample, participants were instructed to combine vials and start their dilutions (outlined below) from the combined material (further details can be found in the study protocol, appendices 9-10). Samples 3 and 4 were to be reconstituted in 0.5 ml of deionized, nuclease-free molecular grade water and sample 5 in 1.1 ml. All lyophilized preparations were left for at least 20 minutes with occasional agitation before use. Participants were requested to dilute samples in human plasma and extract each dilution before amplification. For quantitative assays: Participants were requested to test samples 1, 2 and 6 undiluted only. Sample concentrations of each sample were reported in IU/mL. For samples 3 5, it was suggested that participants perform a minimum of two serial ten-fold dilutions within the linear range of their assay (e.g and 10-2). For qualitative assays Participants were requested to test the dilution at the predicted assay end-point (limit of detection), with a minimum of two half-log serial dilutions either side of the end-point (i.e. at least five dilutions in total). Within subsequent assays participants were asked to work around this end point more precisely. Results were reported as either negative or positive. In both case all results were returned to NIBSC for analysis with details of methodology used. Statistical methods Qualitative and quantitative assay results were evaluated separately. In the case of qualitative assays, for each laboratory and assay method, data from all assays were pooled to give a number positive out of number tested at each dilution step. A single end-point for each dilution series was calculated, to yield an estimate of NAT detectable units/ml. It should be noted that these estimates are not necessarily directly equivalent to a genuine genome equivalent number/ml (9). This model assumes that the probability of a positive result at a given dilution follows a Poisson distribution (with mean given by the expected number of copies in the sample tested), and that a single copy will lead to a positive result. Calculations were performed using the statistical package GLIM (10). When asingle value is obtained for each laboratory and sample no assessment of within laboratory, i.e between assay variability can be made. In the case of quantitative assays, analysis was based on the results supplied by the participants, reported as IU/mL. For each assay run, a single estimate of log 10 IU/mL was obtained for each sample, by taking the mean of the log 10 estimates of IU/mL across replicates, after correcting for

12 Page 12 any dilution factor. A single estimate for the laboratory and assay method was then calculated as the mean of the log 10 estimates of IU/mL across assay runs. Overall analysis was based on the log 10 estimates of IU/mL or NAT detectable units/ml. Overall mean estimates were calculated as the means of all individual laboratories. Variation between laboratories (inter-laboratory) was expressed as standard deviations (SD) of the log 10 estimates and % geometric coefficient of variation (%GCV) of the actual estimates. Variation within laboratories and between assays (intra-laboratory) was expressed as standard deviations of the log 10 estimates and %GCVs of the individual assay mean estimates. Potencies of sample 5, the candidate International Standard relative to the two previous International Standards were calculated as the difference in estimated log10 IU/mL (candidate standard-previous standard) plus the value in International Units/mL (IU/mL) of the previous standard. Results Of the 18 laboratories that were sent samples, 17 laboratories returned data, with two laboratories returning data from two different assays giving a total of 19 data sets for analysis. The data returned comprised 12 (63%) quantitative and 7 (37%) qualitative sets. All qualitative data was produced by assays from a commercial supplier, with the most frequently used assay being the Roche MPX, two different platforms were represented; MPX Ampliprep and MPX 8800, such data constituted 86% of qualitative data. The remaining 14% was performed using the Procleix Ultrio system. All main commercial assays were represented in the quantitative data sets, 4 laboratories used the Roche CAP/CTM v2.0, 2 laboratories reported data from the Roche 6800 platform, single laboratories reported data using the versant kpcr HCV RNA assay, Abbott RealTime and QIAGEN QiaSymphony. One laboratory reported quantitative data using two different in house methods. Tables 5 and 6 show the mean laboratory estimates for quantitative and qualitative assays respectively. Blank spaces in the tables indicate no data was returned for that sample. The loss of titre arising from lyophilisation was assessed by participant, mean values for S1 (liquid bulk) with S5 (candidate 14/150). When combining mean values of both quantitative and qualitative assays for these samples a loss of 0.11 log 10 IU/mL is observed. This is similar to that seen across all assays in the pilot study with the same material. However, taking quantitative and qualitative data sets individually a loss of log 10 IU/mL vs NAT detectable units are observed, as demonstrated in table 7. Inter-laboratory variation for both quantitative and qualitative data is presented in table 7. Variation is greater amongst qualitative data than quantitative in 4/6 candidates evaluated demonstrated by higher GCV values. Overall the combined standard deviation values across all samples were <0.50 log 10 IU/mL. Intra-laboratory standard deviations for each quantitative laboratory data set are presented in table 8, values are very low for each sample indicating excellent precision.

13 Page 13 Histogram representation of laboratory mean estimates for each sample are shown in figures 1-6.shaded squares represent qualitative assays and clear squares quantitative assays. Value agreement is seen across all samples tested with data from qualitative assays flanking the main consensus group comprising of quantitative assays. Figure 7 shows the relative potencies for the candidate standard (S5) relative to the previous 2nd HCV IS (S4), which has previously been assigned a value of 5.00 Log 10 IU/ml. In general for both quantitative and qualitative assays excellent agreement was observed at approximately 5 log 10 IU/mL. One outlying data set, (laboratory 15) representing data from the qualitative MPX assay, gave a value of 3.41 log 10 IU/mL. Figure 8 represents quantitative data only; a variation in potency of only 0.5 log is seen from the mean. Figure 8 shows the estimated relative potency for the candidate sample (S5) relative to the current 4 th HCV IS (S4). Harmonisation is seen in all quantitative assays. However outliers are seen in three separate qualitative assays: the MPX on the 8800 platform, Procleix Ultrio assay and the MPX on the S201 platform. A reduction in relative potency of -2.05, and log 10 from the mean was observed respectively. Table 10 shows relative potency estimates for the candidate sample (S5) relative to both the 2 nd and 4 th HCV IS. Mean potencies for qualitative, quantitative and combined data, plus 95% confidence intervals based on the mean are displayed. Standard deviations for both the qualitative data sets are higher than quantitative. Results for the candidate standard are marginally higher across both data sets relative to the 4 th IS than the 2 nd with a difference of 0.59 and 0.44 respectively. A -0.2 log difference is apparent between the 4 th IS and the 2 nd IS qualitative data sets and potency is also reduced, although to a lesser degree, in the quantitative data (0.11 log), although these differences are not statistically significant. Accelerated degradation studies Table 4 shows titres of the candidate standard at following 3 and 6 months storage at elevated temperatures. Differences between titres compared with the C sample at each time point are shown in parentheses. Changes in titre of up to 0.07 log 10 IU/mL were observed in temperatures up to C at the three month time point and titres at the 6 month time point are indicating a 0.14 log 10 IU/mL loss at C, but all are within the variability of the assays, as presented in table 4. Discussion In this study, a range of NAT-based assays for HCV have been used to evaluate the stability and suitability of the candidate standard as the replacement 5th WHO International Standard for HCV for NAT-based assays. A WHO IS for HCV NAT has been available since During this time, 4 replacement batches have been established; two comprising antibody positive material and two comprising

14 Page 14 antibody negative, the rationale for the different formulations has circulated around the primary end user. Initially, due to the availability of material, the 1 st and 2 nd standards were developed with antibody positive sourced plasma from the UK Blood Transfusion Service. The sourcing of sufficient material for the formulation of a standard can be problematic, with either a small volume of high titre material needed for spiking or a large volume, <2000mL, that can be aliquoted and lyophilised directly with no need for a dilution to be made. Following the development of the first two standards it was identified that the main use of the standard was to calibrate assays used by the blood donation field, in this case the detection of a positive HCV NAT sample would be likely to be a window period donation, thus it was subsequently decide to produce the 3 rd replacement HCV IS using antibody negative material. The 3 rd and 4 th HCV IS preparations were formulated from at the same time from the same bulk stock material, they were evaluated in the same collaborative study, suitable performance was demonstrated by both candidates and thus the 4 th followed on when the 3 rd became depleted. As assay sensitivity improved a loss in titre in the region of 0.2log 10 IU/mL in the 4 th IS became detectable. A look back study across all HCV IS s also identified stability issues with the 3 rd preparation, however this was not mirrored in the older 1 st and 2 nd standard preparations. The main difference between these preparations was the presence of HCV antibody in standards 1 and 2 and its absence in 3 and 4, ie standards 3 and 4 were formulated from a window period donation. It was therefore hypothesised that the presence of HCV antibody created a stabilising effect on the virus during the lyophilisation process, perhaps by the formation of complexes that protects the fragile RNA. Prior to the formulation and study to establish the 5 th standard, a pilot study was performed, however as demonstrated in this report, with the formulations and protocols used for the study no loss in titre was observed. The question over antibody positive or negative formulations was revisited; the rationale for an antibody negative material suiting blood donation centres detecting window period donations still applies. However it is also recognised that assays would also be used for patient diagnosis and treatment management, in which case the clinical sample would most likely be antibody positive. Having established that the presence or absence of stability did not affect the lyophilised stability of the end product it was agreed to formulate the replacement 5th HCV NAT IS with window period donation. Sufficient volume was available at NIBSC and its use would enable a timely production of material to ensure continuity of supply. The candidate was therefore prepared from a new bulk material that had been evaluated for stability in a pilot study previously described in this report. In line with previous WHO HCV NAT standards, the selected candidate material was genotype 1a, a circulating strain that still predominates in the developed world. The selection of genotype was approved at the 3 rd joint workshop of Blood Virology and Clinical Diagnostics held in Graz, Austria on 30/31st May It was accepted that whilst a variety of genotypes are in circulation material, the availability of a WHO International Reference panel for HCV NAT is available for the optimisation of assays against different genotypes and continuity between WHO HCV NAT international standards would be prudent.

15 Page 15 Data generated during the pilot study indicated a 0.2 log drop was to be expected during the lyophilisation of the HCV positive antibody negative material. Appropriate calculations were made for the production of the final material to adjust for this loss to give a final candidate at 5 log 10 IU/mL, as can be seen in table 1, the combined difference for difference between frozen liquid bulk (S1) and the lyophilised standard (S5) is 0.17 log, however if the data for quantitative and qualitative data sets is analysed independently there is a larger decrease in titre observed with qualitative assays (0.43 log 10 ) than quantitative (0.09 log 10 ). The reason for this is unknown, qualitative assays used in this study utilise similar extraction and amplification mechanisms compared with those used in quantitative systems so this phenomenon is unlikely to be attributable to the mechanisms and perhaps more due to analysis methods associated with end point dilutions. In this study both the 2 nd and 4 th (current) HCV IS were included for possible relative potency assignment. Historically it is common practice to assign potency to a replacement standard based on the immediate previous standard. Theoretically for true continuity of the IU from one standard to another a direct comparison should always be made back to the 1 st IS. However, in practice, this is not always possible due to limited availability of material. It has discussed previously in this report that stability issues had been identified with the current 4 th IS. Whilst these appear to be limited to degradation only at temperatures above C it was decided to include a previous standard to give an alternative reference for relative potency assignment should the 4 th IS underperform for any reason in this study. Unfortunately there was insufficient material available to include samples from the 1 st HCV IS, however the 1 st and 2 nd standards were produced from the same bulk material at the same dilution concentration, they were both assigned a unit of 5 log 10 IU/mL in the initial study; the only difference between them being that the lyophilisation was performed on two different days. Both 1 st and 2 nd standard have remained stable at C since their production 19 years ago. Thus it was appropriate to include the stable 2 nd standard in addition to the current 4 th. Table 4 demonstrates the accelerated degradation stability testing performed to date. Such data has subsequently been discussed that the 20 th Anniversary SoGAT workshop, previously referenced, and a WHO collaborating centres meeting held at NIBSC on 2-3 rd July At both meetings it was agreed that the data is inconclusive and additional stability testing should be performed prior to establishment. The six month time point will be repeated and addition 9 month time point performed in advance of the WHO ECBS meeting. In both assays the thermal degradation samples will be tested alongside the current IS for direct potency calibration instead of using the assays calibration. Indications at 3months storage at C suggest that the material would remain stable during normal transportation timeframes, this will be confirmed by simulated transport analysis. Figures 7-10 and table 10 focus on the relative potency comparisons of both standards to the proposed 5 th. It can be seen in figure 7 (combined quantitative and qualitative data) that when comparing the relative potency of S5 relative to S3 (2 nd IS) there is a single outlying data point from a laboratory using the Roche MPX assay performed on the S201 platform (lab 15). A second laboratory also provided data for this study using the same platform (lab 9) and it can been seen that these results are within the main consensus grouping, suggesting that the reduced

16 Page 16 relative potency is a factor of the way the assay has been performed on this occasion and not the initial assay calibration. Similarly, when observing data in figure 9, relative potency of S5 compared to S4 (current 4 th IS), qualitative outliers are observed. In addition to the example above, the Procleix Ultrio assay (lab 10) and Roche MPX performed on the 8800 platform also reports lower potencies than would be expected. A comparison can be made for the Procleix Ultrio assays with a second laboratory also using this assay (lab 7), it can be seen that the data set falls within the main consensus group, therefore as above this is not considered an artefact of the assay system. It is however harder to draw a conclusion for the final outlier (lab 8) as no other laboratories have performed this assay in this study. Given the above comparisons and the potency data presented in table 10, it is apparent that a unit could be assigned to S5 based on either the 2 nd or 4 th standard. However with the consideration of stability issues previously identified with the 4 th standard and the ideal scenario, that relative potencies should be compared directly with the 1 st IS, it would seem appropriate to assign the relative potency of S5 compared with S3 (2 nd IS). S3 has demonstrated good stability and is identical in formulation and unit assignment to the 1 st IS. Furthermore the HCV RNA IS will be used by users of both qualitative and quantitative assays; it is therefore recommended that at potency should be derived from the combined data of both data sets. In a change to previous HCV NAT IS s, this 5 th replacement has been formulated to be reconstituted in a final volume of 1.1mL. Previous standards have been reconstituted in a 0.5mL volume, from the advent of IS development in the blood field 0.5mL provided sufficient volume for a 1 vial to be required. Current assays now require up to 1mL sample volume, thus with previous 0.5mL volumes customers were ordering at least two vials to run one assay and stocks were becoming depleted too quickly. This change will be clearly stated in the information for use sheet that accompanies each shipment of material. In addition a notification will also be sent to all customers who have received the 4 th standard within the last year, notifying them of the volume change and of the availability of the new standard. The batch of candidate 5 th HCV IS comprises 1980 vial. With an anticipated dispatch rate of 250 vials/annum; this stock represents a supply that should have is a little over 7.5 years supply. Whilst it may seem prudent to make a much larger batch, as highlighted above with a biological preparation such as this, it is always a challenge sourcing a sufficient volume and titre of stock material. In these situations one way of ensuring the provision of the same batch of material for an extended period of time is to provide end users with well calibrated secondary reference materials. Within this study, sample 6 was evaluated for this purpose. In conclusion a relative potency assigned compared to the 2 nd HCV IS would equal S5 being assigned a unit of 5.0 log 10 IU/mL when reconstituted in 1.1ml. If it was decided that a potency should be assigned based on the comparison to the current 4 th IS, then S5 would be assigned a unit of 4.90 log 10 IU/ ml when reconstituted in 1.1ml

17 Page 17 Conclusion Based on the results of this collaborative study, S5 observed in this report (NIBSC code 14/150) demonstrates continued harmonisation from the availability of a standard for HCV RNA. The HCV antibody negative formulation used for this candidate performs equally in all assays used. Further investigation of stability needs to be carried out and data will be available for discussion at the 2015 WHO ECBS meeting with a recommendation as to how this material should be sent to participants, supplementary analysis will be submitted to the committee no later than one month before the meeting. This standard will continue to be used by IVD manufacturers, blood transfusion centres, control authorities, and clinical laboratories using both quantitative and qualitative assays, to calibrate secondary reference materials and in the validation of HCV NAT assays. It is therefore appropriate to derive a relative potency value from a combined data set comprising quantitative and qualitative data. A potency should be assigned against the 2 nd HCV RNA IS (S3), this would determine that when properly reconstituted, the preparation contains of 5.0 log 10 IU/ml Proposal It is proposed that a value of 5.0 log 10 IU/mL is assigned to the 5th International Standard for HCV RNA for use in NAT (NIBSC code 14/150). Comments from Participants This report was circulated to all participants for comment. Six participants returned a variety of comments. One laboratory resubmitted data having discovered a calculation error in their data previously submitted, this required some reanalysis by our statistical department. Three laboratories retuned minor typographical corrections and clarification of correct table references in the text. One laboratory asked for the intended unit assignment per ml to be clarified in the final proposal as the per/vial and per/ml unit was confusing. Finally one laboratory suggested the qualitative data should not be included in the final relative potency assignment due to the variability that was observed in these data sets. Whilst this standard is intended for use in both quantitative and qualitative assays it is agreed that calibration of a unit assignment in IU/ml will impact largely on the quantitative assays, therefore, in conjunction with the statistical group at NIBSC the decision has been made to make a unit proposal based on quantitative data only. It was also suggested that further statistical methods should be included for some of the data sets, this comment was address by the statistical group at NIBSC, in this instance further references have been included.

18 Page 18 Acknowledgements We gratefully acknowledge the important contributions of the collaborative study participants and the organisations donating material for use in this study References 1. Saldanha J, Lelie PN and Heath AB. Report on the evaluation of a candidate standard for GAT assays for HCV RNA. WHO ECBS Report 1997;WHO/BS/ Saldanha J, Lelie N and Heath AB. Establishment of the First International Standard for Nucleic Acid Amplification Technology (NAT) assays for HCV RNA Vox Sang. 76: Saldanha J, Heath A, Aberham C, Albrecht J, Gentili G, Gessner M, and Pisani G. WHO collaborative study to establish a replacement WHO international standard for HCV RNA NAT. WHO ECBS Report 2003;WHO/BS/ Saldanha J, Heath A, Aberham C, Albrecht J, Gentili G, Gessner M, Pisani G. World Health Organization collaborative study to establish a replacement WHO international standard for hepatitis C virus RNA nucleic acid amplification technology assays. Vox Sang. 2005;88: Fryer J, Heath AB, Wilkinson DE, Minor PD, Collaborative Study Group. Collaborative Study to Evaluate the Proposed 4th WHO International Standard for Hepatitis C Virus (HCV) for Nucleic Acid Amplification Technology (NAT)-Based Assays. WHO ECBS Report 2011;WHO/BS/ Baylis SA, Heath AB and the Collaborative Study Group. WHO collaborative study to establish a replacement WHO International Standard for Hepatitis C virus RNA nucleic acid amplification technology (NAT)-based assays. WHO ECBS Report 2007;WHO/BS/ Baylis SA, Heath AB; Collaborative Study Group. World Health Organization collaborative study to calibrate the 3rd International Standard for Hepatitis C virus RNA nucleic acid amplification technology (NAT)-based assays. Vox Sang. 2011;100: Guidance for Industry Nucleic Acid Testing (NAT) for Human Immunodeficiency Virus Type 1 (HIV-1) and Hepatitis C Virus (HCV): Testing, Product Disposition, and Donor Deferral and Reentry. U.S. Department of Health and Human Services Food and Drug Administration Center for Biologics Evaluation and Research May Monograph on human plasma for fractionation. European Pharmacopoeia. 07/2008: Collet, D., Modelling Binary Data, Chapman Hall, London. 11. Francis, B., Green, M., Payne, C. (Eds), The GLIM System Release 4 Manual. Oxford Science Publications, Clarendon Press, Oxford.