EVALUATION OF A RAPID ANTE MORTEM BSE TEST

EVALUATION OF A RAPID ANTE MORTEM BSE TEST Scientific Report of the Scientific Expert Working Group of the European Food Safety Authority on Transmissible Spongiform Encephalopathy (TSE) 1 Adopted on 27 November 2006 Question N EFSA-Q-2003-084 Acknowledgement The Experts of the working group are acknowledged for their work for this mandate. The Members are: Wolfgang Phillip, Jim Hope, Jean-Yves Madec, Martin Groschup, Thierry Baron, Peter Lind, Marta Ulvund, Vittorio Silano, Danny Matthews, Johannes Löwer, Emmanuel Vanopdenbosch (Chairman and Rapporteur). The collaborators of the Institute for Reference Materials and Measurements are thanked for the organization, execution and analysis of the results of this evaluation. The supply of the samples by the Friedrich Loeffler Institut, Insel Riems, Germany and the Veterinary Laboratory Agency, Weybridge, United Kingdom, is acknowledged and highly appreciated. 1 For citation purposes: Scientific Report of the European Food Safety Authority on Transmissible Spongiform Encephalopathy (TSE) on a request from the European Commission on the evaluation of a rapid ante mortem BSE test, The EFSA Journal (2006) 95, 1-14 www.efsa.europa.eu Page 1 of 14

Table of contents Acknowledgement... 1 Summary... 3 Background... 4 Terms of Reference... 5 Assessment... 5 1 Evaluation of new rapid ante mortem tests for the diagnosis of BSE... 5 2 Assessment of the application dossier... 5 2.1 Evaluation Protocol... 6 a) Diagnostic Sensitivity... 6 b) Diagnostic Specificity... 6 c) Detection of BSE at early stages of infection... 6 2.2 Selection of serum samples of BSE infected and non-infected animals... 7 a) Serum samples of Infected Reference Animals... 7 b) Serum samples from Uninfected Reference Animals... 7 2.3 Sample Preparation and Evaluation Procedure... 8 2.4 Summary of the results... 9 2.4.1 Diagnostic sensitivity and specificity... 9 2.4.2 Sedation with xylazine... 11 2.4.3 Samples of experimental BSE infected cattle... 11 2.5 Discussion... 13 Conclusions on the Results of the Trial... 13 Recommendation... 14 Documents provided to EFSA... 14 www.efsa.europa.eu Page 2 of 14

Summary The European Food Safety Authority (EFSA) and its Scientific Expert Working Group on Transmissible Spongiform Encephalopathy (TSE) Testing were asked by the European Commission (EC) to take over the mandate of the former Scientific Steering Committee (SSC) for the scientific evaluation of rapid TSE/BSE (Bovine Spongiform Encephalopathy) tests. At present 12 rapid BSE test kits are approved by the EC for the post mortem testing of slaughtered cattle in accordance with the TSE Regulation (EC) No 999/2001. Following an EC call for expression of interest in the Official Journal of the European Union (No C15) on 22 January 2003, several parties indicated their interest in participating in the third European evaluation exercise for newly developed rapid post mortem and live animal TSE/BSE tests. Expressions of interest were invited from those who had tests in advanced stages of development or available for use for the diagnosis of BSE in live cattle. Applications were received from six companies presenting six different tests. In order to ensure that useful tests would be widely available, applicants were also requested to give assurances that they were prepared to make their tests available on a non discriminatory basis following the evaluation. A panel of external scientists assessed all applications based on the pre-defined criteria, covering e.g. the scientific basis of the test, available experimental evidence, practicality of the sampling and testing procedures and stage of development of the test. Following this assessment 1 test was selected for the evaluation exercise. This joint application was presented by two companies: Scil Diagnostics GmbH, Martinsried, Germany and DiaSpec GmbH, Freiburg, Germany for the AquaSpec BSE rapid ante mortem test. Evaluation was based on the EFSA Scientific Report on the Design of a Field Trial Protocol for the evaluation of BSE Tests for Live Cattle adopted on 1 July 2004. On 10 October 2006, EFSA has received the report of the European Commission s Institute of Reference Materials and Measurements (IRMM) on the evaluation of DiaSpec s rapid ante mortem BSE test (IRMM, 2006). Based on the overall assessment covering the application dossier and a phase I laboratory evaluation, the experts of EFSA s Working Group on TSE Testing express their opinion on the evaluated ante mortem BSE test. As the submitted AquaSpec BSE rapid ante mortem BSE Test could not succeed the laboratory trial, the overall assessment is negative. Therefore, the experts of the EFSA WG on TSE Testing concluded that the AquaSpec BSE ante mortem rapid BSE Test did not meet the predefined criteria and therefore do not recommend the AquaSpec BSE rapid ante mortem test (Scil Diagnostics GmbH, Martinsried, Germany and DiaSpec GmbH, Freiburg, Germany) for approval by the European Commission. Key Words: BSE, Bovine Spongiform Encephalopathy, TSE, Transmissible Spongiform Encephalopathy, ante mortem, rapid BSE test, live animal test, evaluation, laboratory trial, Regulation (EC) No 999/2001. www.efsa.europa.eu Page 3 of 14

Background According to EU legislation all slaughtered cattle over the age of 30 months have to be tested using one of the EC approved rapid BSE tests (EC, 2001). In addition, a defined number of fallen stock over 24 months of age as well as all emergency slaughtered cattle over 24 months of age have to be tested for BSE with one of the approved rapid test. At present, test kits of 12 manufacturers are approved by the EC and listed in the Annex to Regulation (EC) No 999/2001. The European Food Safety Authority (EFSA) and its Scientific Expert Working Group on Transmissible Spongiform Encephalopathy (TSE) Testing were asked by the European Commission (EC) to take over the mandate of the former Scientific Steering Committee (SSC) for the scientific evaluation of rapid TSE/BSE (Bovine Spongiform Encephalopathy) tests. Following an EC call for expression of interest in the Official Journal of the European Union (No C15) on 22 January 2003, several parties indicated their interest in participating in the third European evaluation exercise for newly developed rapid post mortem and live animal TSE/BSE tests. On 1 July 2004, the EFSA Working Group on TSE Testing adopted its Scientific Report on the Design of a Field Trial Protocol for the evaluation of BSE Tests for Live Cattle. The expressions of interest were invited from those who had tests in advanced stages of development or available for use for the diagnosis of BSE in live cattle. Applications were received from six organisations presenting six different tests. In order to ensure that useful tests would be widely available, applicants were also requested to give assurances that they were prepared to make their tests available on a non discriminatory basis following the evaluation. A panel comprising 15 external scientists assessed all applications based on the pre-defined criteria, covering e.g.; the scientific basis of the test, available experimental evidence, practicality of the sampling and testing procedures and stage of development of the test. Following this assessment 1 test was selected for the evaluation exercise. This joint application was presented by two companies: Scil Diagnostics GmbH, Martinsried, Germany and DiaSpec GmbH, Freiburg, Germany for the AquaSpec BSE rapid ante mortem test. On 10 October 2006, EFSA has received the report of the European Commission s Institute of Reference Materials and Measurements (IRMM) on the evaluation of DiaSpec s rapid ante mortem BSE test (IRMM, 2006). For the purpose of the report and easiness of reference the test will be referred to as DiaSpec s rapid ante mortem BSE test. The current evaluation exercise describes the second part (the first part started in 2004 with the evaluation of post mortem tests) of an evaluation concentrating on live animal BSE tests. The final goal was to consider possible developments in prion research leading to improved diagnostics. This test evaluation has been designed and managed jointly by the European Commission s Directorate General Joint Research Centre, Institute for Reference Materials and Measurements (IRMM) and the Directorate General Health and Consumer Protection (DG SANCO) in collaboration with a working group on TSE testing of the European Food Safety Authority (EFSA). The results of the laboratory evaluation were reported by IRMM in their document on The evaluation of DiaSpec s rapid ante mortem BSE test GE/RM Unit/12/2006/August/03. The design of the field trial was based on the Scientific Report on the Design of a Field Trial www.efsa.europa.eu Page 4 of 14

Protocol for the evaluation of BSE Tests for Live Cattle, the EFSA Working Group on TSE Testing adopted on 1 July 2004 (Question N EFSA-Q-2003-084). Terms of Reference EFSA was requested by the EC to take on the responsibility for the scientific aspects of the evaluation of new rapid TSE/BSE tests. In a letter of 4 October 2006, EFSA was invited to assess the IRMM document on The evaluation of DiaSpec s rapid ante mortem BSE test, (IRMM, 2006). Besides the evaluation of the technical dossiers of the newly developed test and the results of a phase I laboratory evaluation, the EFSA Working Group on TSE Testing was also asked to provide a conclusion on its contents, in order that the Commission can conclude the rapid test evaluation. Assessment 1 Evaluation of new rapid ante mortem tests for the diagnosis of BSE A panel comprising 15 external scientists assessed all applications, following a call for the expression of interest published in the Official Journal of the European Communities (No C15 of 22 January 2003), based on the pre-defined criteria, covering e.g. the scientific basis of the test, available experimental evidence, practicality of the sampling and testing procedures and stage of development of the test. Following this assessment 1 test out of 6 candidate tests, was selected for the evaluation exercise. This joint application was presented by two companies: Scil Diagnostics GmbH, Martinsried, Germany and DiaSpec GmbH, Freiburg, Germany for the AquaSpec BSE rapid ante mortem test. On 10 October 2006, EFSA received the report of the European Commission s Institute of Reference Materials and Measurements (IRMM) on the evaluation of DiaSpec s rapid ante mortem BSE test. The test is based on the detection of disease associated features in infrared spectroscopic patterns of bovine serum; therefore, a Fourier-transform infrared spectroscopy (FT-IR) spectrum in the middle infrared range (MIR) of a liquid bovine serum sample is measured in the AquaSpec flow cell. The principal component of the AquaSpec Analyser is a microflow cuvette that allows reproducible measurements of liquid samples in their native state. The resulting FT-IR spectrum is analysed with software developed by DiaSpec, using chemo-metrical method. The result is then compared to the cut-off in order to classify the sample. This EFSA Scientific Report provides an overview on the evaluation of the AquaSpec rapid ante mortem BSE test and conclusions and recommendations of the EFSA TSE Testing WG. The evaluation of this new BSE test covers the diagnostic sensitivity, the diagnostic specificity and the detection of BSE at early stages of infection. 2 Assessment of the application dossier The trial was performed according to the EFSA Scientific Report on the Design of a Field Trial Protocol for the evaluation of BSE Tests for Live Cattle (EFSA, 2004). www.efsa.europa.eu Page 5 of 14

This protocol defines that the performance of any new rapid BSE test should statistically not be inferior to that of currently approved post mortem tests. In that protocol it is mentioned that a rapid BSE test for live cattle could be approved for the purpose of consumer protection, for epidemiological screening or for both. However, because of too much uncertainty, the protocol did not consider the evaluation of live animal tests for the purpose of epidemiological studies, for which a lower degree of specificity and sensitivity could be acceptable. Therefore, it was required that if a live animal test is approved for consumer protection purposes, it should have at least the same performance as approved post mortem tests on clinically affected animals. The text hereunder, is extracted and reproduced entirely from the IRMM report. The full report is available at the given link. http://prod.irmmext.wip.irmm.jrc.be/html/activities/tse_testing/index.htm 2.1 Evaluation Protocol In a first step the evaluation exercise was designed to evaluate the diagnostic sensitivity and diagnostic specificity of the selected test. All samples were coded and tested blindly. The evaluation protocol was designed by IRMM and discussed and agreed by EFSA's working group on 'TSE testing'. a) Diagnostic Sensitivity The diagnostic sensitivity of a test is calculated by the proportion of reference samples of infected animals that are tested positive by the assay. 54 samples of 53 confirmed BSE positive animals were used for the evaluation. b) Diagnostic Specificity The specificity of a test is the proportion of reference samples of uninfected animals that test negative with the assay. A total of 558 samples of 488 individual animals were included in the exercise for this purpose. These samples belonged to three categories: CAT 1 NEG: BSE negatives from New Zealand, healthy animals at slaughter; 450 samples of 386 animals; CAT 2 SUSP: BSE suspects: clinical symptoms, but BSE negative; 70 samples of 67 animals; CAT 3 DD: differential diagnosis: samples of animals with other diseases or infections than BSE; 38 samples of 35 animals; c) Detection of BSE at early stages of infection Samples of 3 experimentally orally BSE infected cattle and of 1 unchallenged control animal were included in this part of the evaluation. The samples were collected 24, 30, 34 and 40 months post infection, respectively. In the meantime, it has been confirmed that all three infected animals IT 02, IT 03 and IT 12 were IHC positive in the obex when they were culled www.efsa.europa.eu Page 6 of 14

at 44 months post-infection (mpi), 46 mpi and 46 mpi respectively (Martin Groschup, personal communication, dd 8 11 December 2006). 2.2 Selection of serum samples of BSE infected and non-infected animals The biggest challenge for the evaluation of an ante mortem BSE test is the selection of suitable test material, which shall be collected according to the manufacturer's specifications. DiaSpec's specifications included limitations such as age of samples (24 months), no exposure to CO 2 (dry ice for transport), collection with specific devices (Sarstedt Monovette), a protocol on serum preparation and related time limits, the degree of hemolysis etc. The number of samples as foreseen in the original evaluation protocol could not be reached for all categories, but the age of already collected samples forced us to start the evaluation with the present sample set. The reduced number of samples does, however, not influence the final assessment of the assay on to whether the test can separate serum samples populations of BSE positive and negative animals. a) Serum samples of Infected Reference Animals Blood samples were selected from bovines showing clinical signs of BSE and confirmed to be affected by BSE. Such samples were supplied by the Veterinary Laboratory Agency, Weybridge, United Kingdom. 33 samples were collected in the frame of the BSE surveillance program and stored in VLA's archive; another 20 samples were collected specifically for the evaluation exercise. Samples of experimentally infected cattle were collected at defined time points post infection according to the test developer's specifications. Samples for this evaluation exercise were collected respecting the specifications set up by the manufacturer. b) Serum samples from Uninfected Reference Animals CAT 1 NEG: Serum samples of 400 healthy cattle of mixed breeds aged from 2 to 7 years were collected in New Zealand. Additional samples of another 20 healthy animals were collected before (-) and after (+) a challenge with Xylazine. CAT 2 SUSP: Samples collected from animals showing clinical symptoms of BSE at inspection. 44 out of 70 serum samples were collected and prepared according to the manufacturer's specifications. The remaining serum samples of BSE suspect animals were kindly provided by VLA's tissue archive. CAT 3 DD: All samples were collected according to the technical specifications. Most animals were confirmed to be BSE free and a diagnosis was provided with the samples. The samples derived from animals with the following infections / pathologies / symptoms: www.efsa.europa.eu Page 7 of 14

Table 1. Serum samples for CAT 3 DD derived from animals with the listed symptoms, diseases and infection. Symptoms / Diseases / Infections Depression caused by purulent sinusitis CNS caused by severe metabolic acidosis Dysphagia-Bulbar paralysis Parese hind quarters + Enteritis Central Nervous Symptoms Impossible to stand up Excitation-fever-slight, dysphagia Dysphagia-fever-Listeriosis -brain abscess Spasms due to spinal cord lesion Paresis/depression/hypokalemia/alkalosis Sinusitis-meningitis Meningitis Listeriosis Botulism Enteritis Tetanos Dystocia, paralysis Ataxia Hypocalcemia Malignant catarrhal fever Wherever samples were collected for the purpose of the evaluation, the organisations were requested to follow strictly the collection protocol set-up by the company. CAT 1 NEG animals were healthy and showed no clinical signs suggestive of nervous disease. As a final confirmation, the obex of each animal was removed and subjected to histological examination. For all CAT 2 SUSP (BSE suspect but negative) and most CAT 3 DD (differential diagnosis) the BSE status was confirmed as negative. 2.3 Sample Preparation and Evaluation Procedure All serum samples were kept frozen at -70 ºC from the time of preparation to the day of evaluation. An evaluation is always divided into two main parts: the preparation and the execution. The preparation started in August 2003, when the DiaSpec test was retained by the expert committee for evaluation, the first sample was collected in January 2004, and the last samples were collected in May 2006. One important question to be resolved was whether sedation with Xylazine of animals has an influence on the quality of test samples. Following a Principal Component Analysis (PCA) of the measured spectra it was concluded that results for samples of individual cows differed more than samples of the same cow independent of sedation. As a result it was assumed for the purpose of the evaluation that results were not affected by the use of xylazine for sedation. The evaluation was carried out at the participants premises and was supervised by Commission officials. All additional features are outlined in the test protocol (EFSA, 2004). The evaluation was done in the period of 30 th August 2005 to 6 th September 2005 and between 26 th June 2006 and 4 th July 2006. www.efsa.europa.eu Page 8 of 14

2.4 Summary of the results The sensitivity and specificity estimates are based on pre-determined cut off points defined by the test developers. In this evaluation the diagnostic sensitivity and specificity were calculated on the basis of 54 positive and 558 negative specimens, respectively. The cut-off is divided in two zones for the initial analysis (1 st measurement), samples with values > 0.9 are immediately declared negative and all other samples are re-tested twice. Is the mean value of these measurements < 0.5 the sample is classified positive, samples with values > 0.5 are classified negative. The test developer requested the use of modified software for the classification during the evaluation. This second software version contains a so called 'ph corrector', which is used to normalise spectra and to exclude age and sample quality related features from the classification. The test developer assured that this 'ph corrector' had been used over the last months for the final development of the database but it was removed from the final software only days before the evaluation started. A full report with detailed explanations was provided on 01 July 2006 during the evaluation. 2.4.1 Diagnostic sensitivity and specificity The diagnostic specificity and diagnostic sensitivity of the AquaSpec rapid ante mortem BSE test are summarized in Table 2. Table 2: Summary of results obtained with AquaSpec rapid ante mortem BSE test. NO ph = software version without 'ph corrector', + ph = version with 'ph corrector' NO 'ph corrector' + 'ph corrector' Diagnostic sensitivity 88.7 % 88.2 % 95 % confidence interval 75.4-95.8 % 74.4-95.7 % Diagnostic specificity 63.5 % 78.4 % 95 % confidence interval 58.1-68.9 % 74.3-82.3 % To allow a better demonstration of the capacity and limitations of this assay, results obtained with samples of BSE negative animals are presented separately for three categories: CAT 1 NEG = samples of healthy, non-symptomatic cattle CAT 2 SUSP = clinically suspect, confirmed BSE negative CAT 3 DD = animals with additional infections / symptoms / pathologies than BSE. All spectra were analyzed with two software versions, with and without 'ph corrector'. The version ' NO ph corrector' was used exclusively in the first 3 days of the evaluation, from the 4 th day on all spectra were analyzed with the version including the 'ph corrector'. Results categorized 'invalid' were not considered for the calculations, but numbers are presented here for completeness. www.efsa.europa.eu Page 9 of 14

Table 3: Summary of results obtained with samples of BSE negative animals. NO ph = software version without 'ph corrector', + ph = version with 'ph corrector' NO 'ph corrector' NEGATIVES Classification Category negative positive invalid TOTAL Diagnostic Specificity % 95 % confidence interval CAT 1 NEG 288 144 18 1 450 66.7 62.2-72.7 % CAT 2 SUSP 25 45 0 70 35.7 14.0-53.1 % CAT 3 DD 28 7 3 2 38 80.0 58.8-93.4 % TOTAL 341 196 21 558 63.5 58.1-68.9 % 1 7 samples were declared invalid because of a too small sample volume for a third analysis and 11 samples were declared invalid because of instrument or handling errors 2 1 sample was declared invalid because of a too small sample volume for a third analysis + 'ph corrector' NEGATIVES Classification Category negative positive invalid TOTAL Diagnostic Specificity % 95 % confidence interval CAT 1 NEG 365 68 17 3 450 84.3 80.1-87.8 % CAT 2 SUSP 24 44 2 70 35.3 13.1-53.0 % CAT 3 DD 31 4 3 4 38 88.6 70.7-96.9 % TOTAL 420 116 22 558 78.4 74.3-82.3 % 3 11 samples were declared invalid because of a too small sample volume for a third analysis and 6 samples were declared invalid because of instrument or handling errors 4 1 sample was declared invalid because of a too small sample volume for a third analysis Surprisingly, samples of CAT 2 SUSP were mainly classified positive. 41 out of 44/45 samples were classified positive with both software versions, only 19 out 24 samples classified negative without (-) ph corrector were also classified negative with (+) ph corrector, 2 were now declared invalid, 3 samples were positive. 4 samples classified negative with (+) ph corrector were classified positive without (-) ph corrector. The samples in question were mainly samples collected according to the specifications, only one sample from the archive was classified differently with both versions, whereas (+) ph corrector generated a false positive. www.efsa.europa.eu Page 10 of 14

Table 4: Summary of results obtained with samples of BSE positive animals. NO ph = software version without 'ph corrector', + ph = version with 'ph corrector' NO 'ph corrector' POSITIVES Classification Category negative positive invalid TOTAL Diagnostic Specificity % 95 % confidence interval POSITIVES 6 47 1 1 54 88.7 75.4-95.8 % 1 1 sample was declared invalid because of a too small volume for a third analysis + 'ph corrector' POSITIVES Classification Category negative positive invalid TOTAL Diagnostic Specificity % 95 % confidence interval POSITIVES 6 45 3 2 54 88.2 74.4-95.7 % 2 2 samples were declared invalid because of a too small sample volume for a third analysis and 1 sample was declared invalid because of instrument or handling errors 2.4.2 Sedation with xylazine Serum samples collected before and after the sedation of cattle with Xylazine were included in the population of negatives. Here, all spectra analysed with the 'ph corrector' were classified negative, when analyzed without 'ph corrector' 3 samples were classified false positive. These results underline again that sedation with xylazine does not influence the final classification of a serum sample with the 'AquaSpec BSE test'. This confirmed the first results of March 2005, where DiaSpec could demonstrate on another set of samples that sedation with Xylazine would not influence the classification of samples. 2.4.3 Samples of experimental BSE infected cattle. In order to determine in a first step the ability of the assay to diagnose a BSE infection as early as possible post infection, samples of three orally BSE infected cattle and of a nonchallenged control animal were analyzed. Even if that kind of infection experiment does not guarantee that an infection is successfully established in all animals of the cohort, it is still the most promising approach to generate a BSE infection in a controlled manner. However, animals # 02, # 03 and # 12 are still alive, and all animals show first clinical signs of a BSE infection 42 months p.i. Samples of four time points post infection (p.i.), 24, 30, 34 and 40 months, were measured in parallel and all spectra were analyzed using the program with and without ''ph corrector''. www.efsa.europa.eu Page 11 of 14

Table 5. Results obtained on samples of experimentally infected cows (animals # 12, # 02, # 03). Positive results (< 0.5) are in RED boxes, all other samples were classified negative (values > 0.5). NO ph = software version without 'ph corrector', + ph = version with 'ph corrector'. 1 st, 2 nd = first and second duplicate of the same serum sample. Animal # 12 - challenged Animal # 02 - challenged Months p.i. 24 30 34 40 Months p.i. 24 30 34 40 NO ph 1st 0.42 0.28 0.45 0.31 NO ph 1st 0.89 0.91 0.96 0.83 2nd 0.42 0.28 0.57 0.31 2nd 0.64 0.97 0.99 0.93 + ph 1st 0.51 0.41 0.74 0.35 + ph 1st 0.95 0.96 0.99 0.89 2nd 0.58 0.33 0.65 0.24 2nd 0.78 0.98 1 0.91 NEG CONTROL # 31 Animal # 03 challenged Months p.i. 24 30 34 40 Months p.i. 24 30 34 40 NO ph 1st 0.96 0.95 0.91 0.89 NO ph 1st 0.71 0.77 0.79 0.85 2nd 0.97 0.99 0.94 0.83 2nd 0.94 0.66 0.87 0.79 + ph 1st 0.99 0.95 0.96 0.89 + ph 1st 0.86 0.92 0.93 0.97 2nd 0.94 0.99 0.96 0.87 2nd 0.97 0.93 0.92 0.91 All samples of the unchallenged control animal and of two challenged animals (# 02 and # 03) were classified negative, irrespective of the software version used. However, samples of animal # 12 were tested positive as early as 24 months p.i. when analyzed without 'ph corrector', whereas only samples of 30 and 40 months p.i. were classified positive when the analysis was done with the 'ph corrector'. The results of animal #12 do not show any gradient towards a 'more positive' value over the time course of infection. These results are not coherent and do not demonstrate that the test recognizes BSE infections earlier than commonly used rapid post mortem BSE tests. www.efsa.europa.eu Page 12 of 14

2.5 Discussion Although the diagnostic sensitivity of respectively 88,7% (ph corrector) and 88,2% (non ph corrector) gives some indication about the capability of AquaSpec BSE test for detecting BSE-or PrPsc "features" in a number of cases, the sensitivity is clearly below the requirements as described in the EFSA report (2004) i.e. the test should not be below 98.5% compared with the approved post mortem test and the requirement of EC 999/2001 ( 100% sensitivity for samples from clinical confirmed BSE cases). The wide 95% confidence intervals of 75% till 95% are also illustrative for the rather low reproducibility and repeatability of the test. The specificity, even after the application of the ph corrector, although improving from 63.5% till 78.4%, scores far from the requirements ( EFSA 2004: between 99.95% and 99.99%), and especially the Category 2 Suspects have an unacceptable low 35% specificity, not influenced by ph corrector. Finally, the incoherent results on preclinical samples of experimentally BSE infected cattle, indicates that the test is inappropriate for one of its main purposes, i.e. detecting incubating BSE cases, although a lower detection rate could be accepted in these cases. In addition, it should be noticed that these performances are obtained on "ideally" taken and conserved samples as specified under 4.2.2 of the IRMM report, which implies that the performances can be expected to be lower if real field samples (with hemolysis, bacterial contamination, etc...) would be analyzed on a large scale with the present test. Moreover, the use of ph corrector does not give consistent improvement of the results, i.e. no improvement of specificity in Cat 2 Suspects and on BSE positive animals, making a straightforward recommendation for the use of ph corrector impossible. One of the main problems could have been due to the effects of age and ph of the samples, but this reflects a weak robustness of the test. For all the above mentioned reasons, the AquaSpec rapid ante mortem BSE test as presented in this dossier, is not acceptable as Live animal BSE rapid test in the frame of regulation 999/2001, mainly because not meeting the requirements of not scoring inferior than the currently approved post mortem BSE tests, but also because not capable to detect in a coherent way pre-clinical cases. It must be admitted however that these samples were not strictly collected following the protocol set-up of the company, but this reflects the rather weak robustness of the test, if applied under (less strict) field conditions. Conclusions on the Results of the Trial 1. EFSA s Working Group on TSE Testing agreed that the AquaSpec rapid ante mortem BSE test (Scil Diagnostics GmbH, Martinsried, Germany and DiaSpec GmbH, Freiburg, Germany) did not perform satisfactorily in the trial. It was concluded that its performance was inferior to already approved tests. More in particular, the test performed inferiorly compared to an approved rapid BSE post mortem test based on 6 false negative test results (diagnostic sensitivity of 88,7% no ph corrector, 88,2% ph corrector), on true positive samples (from animals tested positive with an approved reference test) in the diagnostic sensitivity testing. It must be emphasized that the diagnostic sensitivity is the crucial parameter of a test in terms of consumer protection. Also the poor specificity of 63.5% till 78.4%, scores far from the requirements and especially the Category 2 Suspects have an unacceptable low 35% specificity, not influenced by ph corrector. www.efsa.europa.eu Page 13 of 14

2. With respect to the detection of pre-clinical BSE cases, the test gave incoherent results on pre-clinical samples of experimentally BSE infected cattle. Recommendation Despite these rather disappointing results, compared to approved post mortem tests, this methodology could still have a future as a research tool in the frame of TSE or other diseases. Documents provided to EFSA Letter with the ref. D(2006)SD/khk/520963 from the European Commission - Health & Consumer Protection Directorate-General (DG Sanco) requesting to assess the evaluation of the Live animal BSE test evaluation report. IRMM (2006). European Commission s Directorate General Joint Research Centre, Institute for Reference Materials and Measurements (IRMM). The evaluation of DiaSpec's rapid ante mortem BSE test. Wolfgang J. Philipp, Nadine Kollmorgen, Heinz Schimmel, Pierre van Iwaarden. July 2006. (http://prod.irmmext.wip.irmm.jrc.be/html/activities/tse_testing/index.htm) EFSA, 2004. EFSA Scientific Report on the Design of a Field Trial Protocol for the evaluation of BSE Tests for Live Cattle. EFSA Working Group on TSE Testing, adopted on 1 July 2004. www.efsa.europa.eu Page 14 of 14