WHO/BS/ ENGLISH ONLY

Transcription

1 ENGLISH ONLY EXPERT COMMITTEE ON BIOLOGICAL STANDARDIZATION Report on the WHO collaborative study to establish the 1 st International Standard for anti-ev71 serum (Human) Gillian Cooper 1,4, Qunying Mao 2, Laura Crawt 1, Yiping Wang 2, Thomas Dougall 3 Peter Rigsby 3, Zhenglun Liang 2, Miao Xu 2, Philip Minor 1, Junzhi Wang 2 and Javier Martin 1,4 and study participants (see Appendix 1) Division of Virology 1 and Biostatistics 3 National Institute for Biological Standards and Control (NIBSC), South Mimms, Potters Bar, Herts, EN6 3QG, UK National Institute for Food and Drug Control (NIFDC), Beijing , China 2 4 Study Coordinators; Tel: , Fax: Gill.Cooper@nibsc.org Javier.Martin@nibsc.org 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 Kai Gao, at gaok@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 of 39 Summary A collaborative study was conducted with an aim to establish the 1 st International Standard (IS) for anti-human enterovirus A71 (EV71) serum (Human). Two candidate samples were produced from plasma samples donated by healthy individuals in China and collected and characterized by NIFDC. A third serum preparation containing low anti-ev71 antibody titre was also produced with a view to be assessed as a WHO Reference Reagent. The serum samples are intended to help laboratories standardize virus neutralization methods to measure antibody levels against EV71 in human sera. These assays might be suitable for the evaluation of the immunogenicity of EV71 vaccines and for the assessment of the seroprevalence of human populations against EV71 virus. The serum samples were processed, filled and freeze-dried at NIBSC. The study consisted of seventeen laboratories from UK, USA, China, Taiwan, Japan and Malaysia,which national control laboratories, manufacturers of EV71 vaccines and public health laboratories were included. All participants used their own in-house virus neutralization method and a challenge EV71 strain provided by NIBSC. The study samples comprised the two serum candidates, namely 14/138 and 14/140, the low titre sample, 13/238, the Chinese National Standard (Liang et al, 2011), four clinical sera from naturally infected individuals and a sample prepared with a mixture of sera from volunteers who had received EV71 vaccine as part of a clinical trial. Both candidates were included as coded duplicates and also as a single vial in liquid format. Preliminary assays at NIBSC showed that all three serum samples exhibited good levels of neutralizing antibodies against a wide range of EV71 strains of various genotypes. The subsequent collaborative study showed that between laboratory variations in neutralization titres were significantly reduced when values were expressed relative to those of either of the two candidate sera. Stability studies were undertaken to further characterize the candidate materials. Real time stability of samples maintained for 6 months at different temperatures showed no significant loss of activity (relative to that at -20 o C storage temperature) at temperatures of +20 o C and below. From these data, 14/140 is recommended as the 1 st IS for anti-ev71 serum (Human) and 14/138 as a potential replacement for 14/140 in the future, with an assigned unitage of 1,000 and 1,090 International Units (IU) of anti-ev71 neutralizing antibodies per vial, respectively. 13/238 was proposed as a WHO Reference Reagent, serving as an assay control for EV71 neutralization assays with an assigned value of 300 IU per ampoule. Introduction Vaccines for the effective prevention of Hand-Foot and Mouth disease (HFMD) caused by EV71 are under development across South East Asia (Li et al., 2014; Chong et al., 2015). Quantification of anti-ev71 neutralizing antibodies in human sera is an important marker and diagnostic tool in assessing the immune status of individuals. Such data are invaluable for the evaluation of the immunogenicity of EV71 vaccines. With the surge in vaccine production it has been recognized that there is a need to establish suitable reference standards to ensure that methods used to measure the serum neutralizing activity or antibody levels against EV71 are accurate, sensitive and reproducible. This will contribute to the standardized assessment of the quality and efficacy of vaccines used in immunization programs globally. Standardized neutralization assays can also be used to estimate the seroprevalence of human

3 Page 3 of 39 populations against EV71. The proposal for establishing the 1 st IS of anti-ev71 serum (Human) was endorsed by ECBS in 2012 (WHO TRS 980). Candidate materials were identified among sixty two plasma samples donated by healthy individuals in China. As it has been shown in a number of clinical and seroprevalence studies, the cross-reactivity between EV71 strains of different genotypes is not easy to predict as antigenic variations between viruses can be detected but does not have a clear pattern (Huang et al., 2013). For this reason, EV71 strains representative of a wide range of B and C genotypes were used for the initial characterization of the candidate serum samples. EV71 strains of C4, B2 and B4 genotypes are used for EV71 candidate inactivated vaccines that are at different stages of clinical development and have been used in clinical trials in humans (Li et al., 2014; Chong et al., 2015). Aim of the study The aim of the study was to characterize two candidate anti-ev71 sera in virus neutralization assays to assess their suitability to be used as the 1st IS for anti-ev71 serum (Human). A third serum preparation containing low anti-ev71 antibody titres was also evaluated as a WHO Reference Reagent. Both reference materials are intended to help laboratories standardizing virus neutralization methods. Materials and Methods Bulk materials and processing Product summary A donation of sixty-two plasma samples from healthy individuals with glutamic-pyruvic transaminase (ALT) <25U, confirmed to be negative for HBsAg, HIV, HCV and syphilis antibodies were assayed for EV71 neutralizing antibody (NTAb) titres. Twentyfive plasmas which tested negative for Coxsackivirus A16 (CA16)-NTAb and positive for EV71- NTAb were chosen to be pooled to produce the two high candidate IS and a low anti- EV71 antibody reference. Ten plasmas with similar high levels of anti-ev71 neutralising antibodies were selected to become the two high titre candidate standards and five plasmas with similar levels low levels of antibody were selected for the low titre reference. The initial screening of plasma suitability was performed at NIFDC. The final selection of plasma for the candidate international references was performed in conjunction with NIBSC. The serum samples were then processed, filled and freeze-dried at NIBSC. Individual plasmas were shipped to NIBSC frozen on dry-ice and stored at -80 o C until processing. Aliquots (1ml) of each plasma were also sent for formulation purposes. The samples were screened for safety markers by the NIBSC Blood Virology section and all found to be negative for HBsAg, Anti HIV 1+2, HCV RNA NAT test and Anti-HCV, and therefore suitable to serve as candidate International references. A pilot trial fill was undertaken to validate if the defibrination method used was suitable for the freeze drying of the filled product. There is little previous experience for the freeze drying of EV71 serum and therefore it was important to establish a cycle that would give a reliable and, stable product limiting the loss of EV71 NTAb activity. This was completed on

4 Page 4 of 39 27/01/14. The three candidates gave yellowish, robust cake that remained intact on agitation. The fill Coefficient of Variation (CV%) was 1.80, 1.21 and 1.49 for the two high and low respectively. These figures are within the acceptable range for a product of this nature. Moisture and oxygen levels were also within range. Ampoules from the pilot fills were also tested by the NIBSC Biotherapeutics department for presence of thrombin. All samples tested negative, validating the method used. Defibrination and Pooling Plasma samples were defrosted at +4 o C prior to defibrination and pooling. Once thawed the samples for each reference were pooled and aliquoted in 300ml volumes into 500ml sterile blood bottles. They were treated with 10% CaCl 2 (0.125 M) and incubated for 30 minutes at 37 o C. The bottles were then placed at +4 o C overnight to form a clot. The liquid was removed and the clot was squeezed to release any retained liquid. The clot was then returned to +4 o C overnight and the process of squeezing was repeated over several days. The serum was spun at 4000rpm for 30 mins and the supernatant removed. The resulting sera were stored at +4 o C prior to filling and freeze drying. Filling, Freeze-drying and Sealing Filling was completed for all three candidates from a homogenous stirred bulks maintained at +4 0 C throughout the filling using a Bausch and Strobel AFV5090 machine. 3ml ampoules were filled with 0.5ml of material and in-line samples. For every 90 ampoules filled 3 ampoules (4-5% of total filled ampoules) were taken for measurements of the fill volume. Freeze-drying was carried out directly after filling using a 2 day cycle. After completion of the freeze drying the candidate materials were put at long term storage of -20 o C. Ampoules were sealed under boil-off gas from high purity liquid nitrogen (99.99%) and measurement of the mean oxygen head space after sealing served as a measure of ampoule integrity. The mean oxygen head space was measured non-invasively by frequency modulated spectroscopy (FMS 760, Lighthouse Instruments, Charlottesville, USA). Residual moisture content was measured using the colorimetric Karl Fischer method in a dry box environment (Mitsubishi CA100, A1 Envirosciences, Cramlington, UK) with total moisture expressed as a percentage of the mean dry weight of the ampoule contents. Product summary details for each of the filled samples are shown in Table 1. Fill dates for the candidate materials are detailed below: Anti-EV71 serum Low (NIBSC 13/238) 14/02/2014 Anti-EV71 serum High (NIBSC 14/138) 29/05/2014 Anti-EV71 serum High (NIBSC 14/140) 29/05/2014 Post filling testing of freeze-dried samples One ampoule for each of the candidate ISs 14/138 and 14/140 from the beginning, middle and end of the filling run were tested in three repeat assays to check the EV71 NTAb activity after the freeze drying process. This testing was performed against the NIBSC C4 virus strain. The GMT s of the beginning, middle and end samples of 14/138 were; 724, 574 and 456, and for 14/140 were; 512, 512 and 645. Overall there were no apparent losses in activity for each of the candidates for any of the time points during the filling process. The filled material was therefore fit to be used in the collaborative study as candidate ISs.

5 Page 5 of 39 Study Samples A total of thirteen samples were provided to participants. The samples were shipped in dryice and storage at -70 o C was recommended. Participants were provided with three ampoules of sample A, B, D, E and F, two ampoules of samples H and I, one vial of sample T, W, X, Y and Z. A Sample Receipt Form was included with the study protocol to give feedback on the date of the shipment and condition the samples were received. Instructions for Use and Material Data Sheets were also provided. There were no issues with the shipment and the receipt of samples, although one laboratory reported later that their samples were received without dry-ice. The study samples are listed below: Low candidate for anti EV71 antibodies, sample A A collection of 5 plasmas from Chinese donations that had similar low EV71 antibody serum titres and tested negative for CA16 antibodies, were pooled to produce this standard (NIBSC -13/238). WHO 1 st candidate for anti EV71 antibodies High, samples B and D A collection of 5 plasmas from Chinese donations that had similar High EV71 antibody serum titres and tested negative for CA16 antibodies were pooled to produce this standard (NIBSC -14/138) WHO 1 st candidate for anti EV71 antibodies High, samples E and F A collection of 5 plasmas from Chinese donations that had similar High EV71 antibody serum titres and tested negative for CA16 antibodies were pooled to produce this standard (NIBSC -14/140) NIFDC National anti EV71 antibody standard, sample G Produced and validated in a collaborative study in 2010: code 2010/N o Produced from a single plasma from a naturally infected donor with appropriate safety levels. Assigned 1,000 EV71 U/ml (NTAb units). Freeze dried preparation and is reconstituted in 200ul sterile glass distilled water. Custodian is NIFDC and the long term storage is -20 o C. WHO 1 st candidates for anti EV71 antibodies High liquid presentation, samples H and I Samples of 14/138 and 14/140 were reconstituted by NIBSC in 0.5ml Sterile distilled water and frozen at -20 o C prior to the study. Human serum from EV71 vaccinees, sample T 10 sera from Chinese donors with more than 0.5ml volume and NTAb higher than 1:32 were selected and pooled to produce vaccine-immunized sera for the collaborative study. Clinical sera from naturally infected individuals, samples W, X and Z Sera from healthy Chinese adults naturally infected with EV71.

6 Page 6 of 39 Low titre EV71 serum, sample Y Serum with low EV71 NTAb from an individual from a different geographical region provided by NIBSC. Challenge virus EV71 strains from most available B and C genotypes, representative of viruses that have been widely circulating in the world in recent years (Chong et al., 2015), were selected for the initial analysis of the sera. The virus panel covered a wide range of isolation dates and geographical locations (Table 2). No genotype A or D viruses, that have circulated to a much lesser extent in recent years, were available for this study. Viruses were kindly provided by Dr Qunying Mao (NIFDC, China), Dr Hiro Shimizu (NIID, Japan) and Dr Harrie van der Avoort (RIVM, The Netherlands). The C4 523 strain was chosen from the panel of viruses. This particular strain is routinely used by laboratories in China for the NTAb test of EV71 and was shown to be a suitable challenge virus for the test based on the results from the collaborative study for the establishment of the current Chinese National Standard (Liang et al., 2011). This choice was further justified by the fact that most EV71 vaccines in advance stages of clinical development (mostly in China) are manufactured using EV71 C4 strains which have been shown to induce good levels of cross-reactive neutralizing antibodies against a wide range of EV71 genotype B and C strains (Zhang et al., 2014). Participants were provided with one vial of EV71 C4 523 strain for use as the main challenge virus for the study. The decision to provide a common challenge virus for all participants was made with a view to reduce the variability in neutralization results between laboratories. This C4 523 strain could serve as a future reference virus for neutralization assays if there is demand. Laboratories 14 and 17 also used EV71 B4 C7/Osaka/1997 strain as a second challenge virus. An EV71 B4 strain is used for vaccine production by at least one manufacturer in Taiwan (Li et al., 2014; Chong et al., 2015). Design of Collaborative Study Participants Twenty one laboratories were invited to participate in the study. Eighteen laboratories from seven countries accepted to participate. They included two National Control laboratories, ten manufacturers and five Public Health Institutes. Seventeen retuned data. They are referred to by a code number, allocated at random, and not reflecting the order of listing in Appendix 1. Study time-frame The study samples were sent to participants mid-march 2015 and the coordinator requested data to be returned by the 27 th April Due to issues with some participants requiring permits, some study samples were sent late so the deadline for these participants was extended. The majority of participants were able to return data by the deadline. Study Plan Participants were requested to:

7 Page 7 of 39 - Follow recommended protocols for storage, reconstitution and preparation of dilutions of the study samples. - Determine the neutralization titre against EV71 of each of the 13 coded human sera in the panel by performing three independent assays using their in-house method and RD cells if possible. - Use a fresh ampoule of samples A, B, D, E, and F for each assay. - Use the EV71 C4 strain included in the panel as the challenge strain in the neutralization assay (if participants routinely use a genotype B strain in the laboratory, they were asked to also evaluate the panel of sera against an EV71 B virus strain as well). Laboratory methods Participants used the same basic methodology where the study samples were assayed for EV71 neutralizing antibodies using a constant virus varying serum method using RD cells. The main challenge virus strain was provided by the study coordinator. The challenge dose recommended for the assay was between log 10 TCID 50. The assay method requires the serum and virus to be incubated in a 96 well. After an initial incubation step the cell substrate is added and a further incubation step is required to allow for the development of cytopathic effect (CPE). The assay is then read to determine the end point dilution based on the last positive well showing CPE. Participants provided calculated titres based on this method. Details of participant s in-house methods can be found in Table 3. Overall the methods were relatively consistent across parameters. Laboratories 11, 12 and 13 used only 4 days post infection for the development of CPE versus 7 days for all other laboratories. The major variation is the use of the Plaque Reduction Neutralization Test (PRNT) using Vero cells by Laboratory 15.RD cells were used by all participants except laboratory 12 who used Vero cells to perform the study. Participants included their in-house assay positive and negative controls and reported results. Documentation of Study results Participants were requested to report their results electronically using standard forms provided by the coordinator. Results including potency calculations and statistical validity criteria were to be sent using the Results Forms. The Raw Data Form provided a template so that participants could document the scoring of the samples. Information on the method and assay reagents used was also requested using the Method Form. Statistical methods For each sample, end point titres were based on the last positive well in a dilution series. Titres were converted directly into relative potencies by dividing the titre value by that obtained for the appropriate standard. Potencies relative to the two candidate high titre

8 Page 8 of 39 standards, 14/138 and 14/140, were calculated relative to the geometric mean (GM) of the results obtained for the coded duplicates (samples B and D for 14/138, samples E and F for 14/140). Mean estimates for each sample in each laboratory were taken as the geometric mean (GM) of the three assays performed. Variability in results (end point titres and relative potencies) between assays within laboratories and between laboratories was assessed using geometric coefficients of variation (GCV = {10 s -1} 100% where s is the standard deviation of the log 10 transformed results), within assay variation was assessed using the relative potencies of the coded duplicates for the two candidate standards 14/138 and 14/140. Pooled GCVs across all laboratories have been calculated for the clinical samples T, W, X and Y for the end points titres and the potencies relative to the candidate standards. Stability Studies Samples of the candidate standards were stored at -150 C, -70 C, -20 C, +4 C, and +20 C and tested at 1 and 6 months. Samples will be tested yearly to assess long term stability. Samples were also placed at + 37 C and +45 C for up to one month and these were also tested for EV71 NTAb. The stability of candidate standards reconstituted in liquid form was also assessed. Ampoules of the three candidate materials were reconstituted in 0.5ml of sterile glass distilled water and maintained at +4 C or 20 C for a period up to four weeks. Results Preliminary characterization of serum samples Virus stocks for this study were prepared by growth in cell cultures and titres were established using a validated infectivity assay (TCID 50 ) using RD cells. Nearly full genome sequences of stock EV71 strains were obtained by deep sequencing analysis using the MiSeq sequencing platform. The donated serum samples were initially tested against four different EV71 genotype C4 strains and one CA16 virus at NIFDC. Data from this analysis were used to select the serum samples that were used to produce the three candidate sera. The data shown in Table 4 represent the average for the pooled plasmas for each group expressed as GMT s against the different EV71 viruses. All serum samples were negative against the CA16 virus (data not shown). All serum samples showed lower neutralization titres against virus EV71-1 different to those against EV71-2 to 4 strains. There is not a clear explanation for this result as all four viruses were of the same genotype (C4), all from China and even EV71-1 to 3 strains were obtained from the same outbreak in Anhui, China in However, mutations found at capsid amino acids VP1-98, 145, 167 and 183 in EV71-1 strain might have an effect on the antigenicity of this virus as they all locate at or close to antigenic site 1, near the 5-fold axis of symmetry. This can explain, at least in part, the observed results. Following plasma processing, filling and freeze-drying, the candidate serum samples were tested against a panel of EV-71 viruses (Table 2) to check cross reactivity. The results

9 Page 9 of 39 showed that all three serum preparations had neutralization activity against all virus strains tested (Figure 1). Generally, all three serum preparations showed higher neutralization titres against genotype C EV71 strains than they did against genotype B EV71 viruses although differences for serum samples 13/238 and 14/140 were small. Differences were greater for serum sample 14/138. All three sera showed the lowest neutralization titres against EV71 B strain isolated in The Netherlands in A large stock of EV71 C4 523 virus was prepared at NIBSC by growth in RD cells for distribution to collaborative laboratories. The virus titre was given to participants as a guide, and they were requested to verify that they could obtain a titre within +/-0.3log 10 of the estimate using their routine RD cells assay. Laboratories were asked to also test EV71 genotype B strains if available. Results of collaborative study Study data returned Data were received from 17 participants who calculated end point titres for all samples using their in-house EV71 NTAb method in 3 assays (Table 5A and Figure 2). Lab 7 returned two sets of results from two different operators, but one set was only partially complete and was excluded from further analysis. Labs 14 and 17 returned two sets of results using virus strains B4 and C4 in each of them. Re-analysis of the assay raw data at NIBSC gave close agreement to the results reported by participants and therefore the participant s results were used for subsequent analysis. Intra and inter-assay variability Relative potencies of the coded duplicate samples of 14/138 (B and D) and 14/140 (E and F) were used to assess intra-assay variability and are shown in Figures 3 and 4, respectively. With the exception of 14/140 in labs 12, 13 and 17, relative potencies were in the range [0.5, 2.0] in 97% of cases, showing a good level of intra-assay precision among participating laboratories. Inter-assay variability, as illustrated by the between-assay GCVs in Table 5A, ranged from 0% (no difference in results obtained for sample in all three assays) to 396% (16-fold difference in results obtained for sample W by lab 2). Despite some instances of high interassay variability, no further exclusions were made for the analysis presented in this report. End point titres Table 5B and Figure 5 summarise the calculated laboratory mean end point titre estimates. With the exception of sample T, all between-laboratory GCVs were above 60%. The pooled between-laboratory GCV for the clinical serum samples T, W, X and Y is 64%. Potencies relative to 14/138 (samples coded B and D) Table 6 and Figure 6 summarise potency estimates relative to 14/138 (samples coded B and D). These show better agreement between laboratories than end point titres for all samples, with the exception of sample Z which had the lowest titre of the samples tested. With the exception of samples W and Z, between-laboratory GCVs were between 33% and 49%. The pooled between-laboratory GCV for the clinical samples T, W, X and Y was reduced to 41%.

10 Page 10 of 39 Potencies relative to 14/140 (samples coded E and F) Table 7 and Figure 7 summarise potency estimates relative to 14/140 (samples coded B and D). These also show better agreement between laboratories than end point titres for all samples, with the exception of sample T which showed an equivalent level of agreement (GCV of 45% compared to 43%). With the exception of sample Z, the low titre sample, between-laboratory GCVs were between 29% and 45%. The pooled between-laboratory GCV for the clinical samples T, W, X and Y was reduced to 35%. Comparison of methods using Vero and RD cells Analysis of variance with Tukey s multiple comparison test was used to compare the differences in methodology used between laboratories, using log transformed potencies. The analysis showed that there is no significant difference in using Vero cells compared to RD cells for the EV71 NTAb test. The comparison shows that using a PRNT method with Vero cells, give results at the low end of the range, but these are not significantly different from the majority of other laboratories. Several laboratories use only 4 days post infection to score for CPE which could impact on the end point titres. The analysis showed that these laboratories were within the middle of the range so not significantly different compared to other methods. Comment on data returned from Laboratory 7 for second operator Comparison of the data between the two operators for laboratory 7 for samples A F, H and I were consistent between operators and across samples. The results for operator 2 were midrange for the spread of end point titres reported by participants for all samples. Results for study samples using EV71 B4 strain as a challenge virus A comparison was performed by calculating the ratio between neutralization titres against B4 and C4 viruses for each study sample using data from each laboratory. The analysis (Table 8 and Figure 8) showed that there was good correlation between the results obtained by laboratories 14 and 17 using C4 and B4 strains. However, agreement between the two laboratories was not as good for samples A, W, X and Z. This is likely because those samples contain low antibody neutralization titres against both B4 and C4 strains so assay variability would have a greater impact on the overall result. As shown during preliminary work at NIBSC, all study samples showed good levels of neutralizing activity against both B4 and C4 EV71 strains. Neutralization titres against the B4 strain were generally lower than those against the C4 strain, particularly for serum samples B, D and H (representing candidate standard 14/138). The results likely reflect the serum composition as different individual donors would have been exposed to different EV71 strains. Stability study results At the time of this report only data up to the 6 month time point are available for evaluation. Three independent assays were performed at each temperature and time point, and potencies relative to the -20 C baseline for all three candidates were obtained. Geometric means with confidence limits are shown in Tables 9A, B and C. The data show that there is no significant loss of activity at +4 C and +20 C (temperatures used during laboratory manipulation for assays), relative to the -20 C baseline for any of the three candidates. A program to measure real-time stability is ongoing by regularly measuring the potency of samples stored at -70 C, -20 C and +4 C.

11 Page 11 of 39 The stability of candidate standards reconstituted in liquid form was also assessed. Ampoules of the three candidate materials that had been reconstituted in 0.5ml of sterile glass distilled water and maintained at different temperatures were tested. Three independent assays were performed at each temperature and time point, and potencies relative to that of an ampoule stored at -20 C and freshly reconstituted were obtained for all three candidates. The data show that there is no significant loss of activity for any of the reconstituted materials that had been stored at -20, +4 C or +20 for any of the three candidate standards (Table 10). Discussion Results shown in this study indicate that the candidate serum samples, particularly 14/140, are suitable as reference standards to measure the neutralization activity in human sera against a wide range of EV71 strains from different genotypes. Based on the lower overall GCV for clinical samples relative to 14/140 and the higher level of cross-reactivity of this serum, it was decided to recommend 14/140 as the 1 st IS for anti-ev71 serum (Human) with an assigned potency of 1,000 IU of anti-ev71 neutralizing antibodies/ampoule. Based on this proposal, the candidate 14/138 has a potency of 1,090 IU of anti-ev71 neutralizing antibodies/ampoule. The low titre reference 13/238 was found to be suitable as an assay control for the neutralization test. Based on its mean relative potency it can be assigned a unitage of 300 IU of anti-ev71 neutralizing antibodies/ampoule. Further collaborative studies might be required to measure more accurately the neutralizing activity against any EV71 strain other than the EV71 C4 523 virus analysed here. The collaborative study data showed high GCVs across laboratories. This high variability is expected due to the methodology used to assess neutralizing antibody in a cell based system. However, once the data are expressed as relative potencies against the candidate materials, the between laboratory GCVs were reduced. Stability data for the candidate materials showed that up to six months the standards were stable. However, more data needs to be generated to be able to establish a long term stability profile. Stability data also showed that the candidate standards were also stable after reconstitution and re-freezing, although a slight loss in potency might have occurred after four weeks at +4 C or two weeks at +20 C for some samples. The aims of this collaborative study have been met, and a candidate IS can be proposed with an IU unitage. The study has also shown that the neutralizing antibody method used by participants was a suitable tool to assess the study materials and that expressing data as relative potencies improves between laboratory standardization. This is important as the role of this IS is to ensure that methods used to measure the serum neutralizing activity or antibody levels are accurate, sensitive and reproducible for the assessment of vaccines used in immunization programs. Commutability for the candidate serum samples for clinical samples has not been specifically addressed in this collaborative study. To address this would require the reference material to be evaluated in a range of assay methods, alongside all sample types for which it might be used for the detection of EV71 Ab. However, the data presented in this report for several clinical serum samples, which include samples with low, medium and high antibody content, demonstrate that inter-laboratory agreement is improved when results are expressed as relative potencies and so the 1 st IS for anti-ev71 serum (Human) is fit for its intended use.

12 Page 12 of 39 Recommendations It is proposed that the candidate 14/140 should be established as the 1 st IS for anti-ev71 serum (Human) to be used for the standardization of virus neutralization assays. The assigned potency for this should be: 1,000 IU/ampoule It is proposed that the candidate 14/138 should be retained as a potential secondary/replacement IS for anti-ev71 serum (Human) to be used for the standardization of virus neutralization assays. The assigned potency for this should be: 1,090 IU/ampoule It is proposed that 13/238 should be ratified as a low anti-ev71 WHO Reference Reagent for anti-ev71 serum (Human) to be used for the standardization of virus neutralization assays. The assigned potency for this should be: 300 IU /ampoule A summary of participant s comments is provided in Appendix 2. A sample of the Instructions for Use and the Material Safety Data sheet can be found in Appendix 3. References - Chong P, Liu CC, Chow YH, Chou AH, Klein M. Review of enterovirus 71 vaccines Clin Infect Dis. 60(5): doi: /cid/ciu Huang ML, Chiang PS, Chia MY, Luo ST, Chang LY, et al. Cross-reactive neutralizing antibody responses to enterovirus 71 infections in young children: implications for vaccine development PLoS Negl Trop Dis 7. doi: /journal.pntd Li JX, Mao QY, Liang ZL, Ji H, Zhu FC. Development of enterovirus 71 vaccines: from the lab bench to Phase III clinical trials Expert Rev Vaccines.(5): doi: / Liang Z, Mao Q, Gao Q, Li X, Dong C, Yu X, Yao X, Li F, Yin W, Li Q, Shen X, Wang J. Establishing China's national standards of antigen content and neutralizing antibody responses for evaluation of enterovirus 71 (EV71) vaccines Vaccine. 29(52): doi: / j.vaccine Zhang H, An D, Liu W, Mao Q, Jin J, et al Analysis of Cross-Reactive Neutralizing Antibodies in Human HFMD Serum with an EV71 Pseudovirus-Based Assay. PLOS ONE 9(6): e doi: /journal.pone

13 Page 13 of 39 Table 1. Product Summary. Product Summary of candidate standards used in the study NIBSC code 13/238 14/138 14/140 Presentation 3ml DIN ampoules 3ml DIN ampoules 3ml DIN ampoules No. of containers 3,508 5,004 5,010 Number of containers weighed Validation of ampoule integrity Visual inspection plus Oxygen headspace on 12 random samples Visual inspection plus Oxygen headspace on 12 random samples Visual inspection plus Oxygen headspace on 12 random samples Sealing gas Nitrogen from liquid Nitrogen 99.99% pure Nitrogen from liquid Nitrogen 99.99% pure Nitrogen from liquid Nitrogen 99.99% pure Measured by Near Infra-Red spectroscopy Near Infra-Red spectroscopy Near Infra-Red spectroscopy Residual Moisture measured by Karl Fischer reagent Karl Fischer reagent Karl Fischer reagent Mean fill mass 0.5g 0.5g 0.5g CV fill mass 0.34% 0.44% 0.30% Mean dry weight 0.04g 0.04g 0.04g CV of dry weight 0.23% 0.82% 1.08% Mean residual moisture 0.58% 0.70% 0.70% CV residual moisture 14% 8% 11% Mean oxygen headspace 0.37% 0.28% 0.35% CV of oxygen headspace 34% 33% 21% Date of fill 14/02/ /05/ /05/2014 Storage temperature -20 C -20 C -20 C Microbial contamination None detected None detected None detected

14 Page 14 of 39 Table 2. List of EV71 strains in the panel. Genotype EV71 Strain Year Location B0 B0 ( ) 1966 Netherlands B1 B1 ( ) 1971 Netherlands B2 B2 ( ) 1986 Netherlands B3 B3 (MAL-97-B3) 1996 Malaysia B4 B4 (JPN-97-B4) 1997 Japan C1 C1 (91-480) 1991 Netherlands C2 C2 (36-92) 2007 Netherlands C4 C4 (JPAN-3-C4) 2003 Japan C4 C4 (China 523 NIBSC) 2007 China C4 C4 (75-YAMAGATA) 2003 Japan C5 C5 (E TW) 2006 Taiwan

15 Page 15 of 39 Table 3. Details of methods used by participants. Lab Code Cell line used Cell information In-house Positive Serum Controls In-house Negative Serum Controls virus-serum incubation (hours & ) CPE Assay (Days & ) 1 RD E5 cells/well for 2 hrs 37 for 7 days 2 RD 1.5E5 cells/well for 2 hrs 37 for 7 days 3 RD 1.0E5 cells/well No information No information 4 RD E5 cells/well for 2 hrs 37 for 7 days 5 RD E5 cells/well for 2 hrs 36 for 7 days 6 RD 1.5E5 cells/well for 2 hrs 35 for 7 days 7 RD 1.5E5 cells/well for 2 hrs 35 for 7 days 8 RD 1.0E5 cells/well for 2 hrs 37 for 7 days 9 RD E5 cells/well for 2 hrs 35 for 7 days 10 RD 2.0E5 cells/well for 2 hrs 35 for 7 days 11 RD 0.05E5 cells/well for 1 hrs 37 for 4 days 12 Vero 0.4E5 cells/well for 1 hrs 37 for 4 days 13 RD 0.3E5 cells/well for 1 hrs 37 for 4 days 14 RD 1E5 cells/well for 2 hrs 35 for 7 days 15 Vero 1.5E5 cells/well for 1 hrs 37 for 6 days 16 RD 0.3E5 cells/well for 1 hrs 37 for 7 days 17 RD 1.0E5 cells/well for 3 hrs 37 for 7 days

16 Page 16 of 39 Table 4. Data for Candidate Materials tested against four strains of EV71 Virus. EV71-1 EV71-2 EV71-3 EV71-4 High High Low

17 Page 17 of 39 Table 5A. Neutralization results in all laboratories. Lab Assay Sample Code A B D E F G H I T W X Y Z GM GCV GM GCV GM GCV GM GCV GM GCV GM GCV GM GCV / / / / / / / / / / / / / / / / / / GM / / / / / / GCV / / / / / / GM GCV

18 Page 18 of GM GCV GM GCV GM GCV GM GCV GM GCV C GM B4 GCV GM GCV GM GCV GM GCV GM GCV

19 Page 19 of 39 Table 5B. Laboratory geometric means end point titres. Lab A B D E F G H I T W X Y Z Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab GM GCV 74% 70% 61% 60% 78% 74% 76% 76% 43% 75% 70% 94% 82%

20 Page 20 of 39 Table 6. Laboratory geometric mean potencies relative to 14/138 (samples coded B and D). Lab A E F G H I T W X Y Z GM (E:F) Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab GM GCV 49% 33% 46% 36% 36% 43% 39% 57% 34% 46% 85% 33%

21 Page 21 of 39 Table 7. Laboratory geometric mean potencies relative 14/140 (samples coded E and F). Lab A B D G H I T W X Y Z GM (B:D) Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab Lab GM GCV 41% 44% 30% 30% 41% 29% 45% 35% 36% 35% 66% 33%

22 Page 22 of 39 Table 8. Neutralization titres using B4 and C4EV71 strains data for laboratories 14 and 17. B4 Virus Strain Lab code Assay A B D E F G H I T W X Y Z C4 Virus Strain GM GM GM GM Ratio between strains (%) lab 17 35% 15% 28% 59% 94% 49% 20% 49% 31% 119% 52% 66% 66% lab 14 63% 13% 25% 63% 100% 63% 16% 50% 31% 50% 100% 63% 100% Stability studies

23 Page 23 of 39 Table 9A. Potencies of 13/238 relative to -20 o C sample (%). Time GM Temp (months) Assay 1 Assay 2 Assay 3 (Overall) LCL UCL -20 C C C C C C Table 9B. Potencies of 14/138 relative to -20 o C sample (%). Time GM Temp (months) Assay 1 Assay 2 Assay 3 (Overall) LCL UCL -20 C C C C C C Table 9C. Potencies of 14/140 relative to -20 o C sample (%). Time GM Temp (months) Assay 1 Assay 2 Assay 3 (Overall) LCL UCL -20 C C C C C C

24 Page 24 of 39 Table 10. Stability after reconstitution. Potencies of 13/238, 14/138 and 14/140 relative to - 20 o C sample (%) reconstituted on day of assay. Temp Time (weeks) 13/238 14/138 14/ C C C C C C

25 Page 25 of 39 Figure 1. Data for the candidate samples tested against a panel of EV-71 virus strains.