EUROPEAN COMMISSION HEALTH & CONSUMER PROTECTION DIRECTORATE-GENERAL

EUROPEAN COMMISSION HEALTH & CONSUMER PROTECTION DIRECTORATE-GENERAL Directorate F - Food and Veterinary Office DG(SANCO)/2007-7315 MR Final FINAL REPORT OF A MISSION CARRIED OUT IN BRAZIL FROM 27 FEBRUARY TO 8 MARCH 2007 CONCERNING THE EVALUATION OF THE CONTROL OF RESIDUES AND CONTAMINANTS IN LIVE ANIMALS AND ANIMAL PRODUCTS, INCLUDING CONTROLS ON VETERINARY MEDICINAL PRODUCTS Please note that factual errors in the draft report have been corrected. Clarifications provided by the Brazilian Competent Authority are given as footnotes, in bold, italic, type, to the relevant part of the report. 19/06/07-73876

EXECUTIVE SUMMARY This report describes the outcome of a mission carried out by the Food and Veterinary Office (FVO) in Brazil 27 February to 8 March 2007. The mission was part of a series of FVO missions on residue controls in third countries. The overall objective of the mission was to (a): evaluate the control of residues and contaminants in live animals and animal products including the controls on the distribution and use of veterinary medicinal products and feed additives, the use of which may give rise to residues in such products and (b): to follow-up on the actions taken by the Competent Authorities in response to recommendations made in previous FVO missions covering this topic. The evaluation was based on the standards set out in Council Directive 96/23/EC, and other relevant Community legislation in this field, including legislation on the control and distribution of veterinary medicinal products. The mission assessed the performance of the competent authorities and other officially authorised entities involved in residues and veterinary medicinal product controls and the legal and administrative measures put in place to give effect to the relevant Community requirements with regard to import of food of animal origin into the EU. The report concludes that much progress has been made in the realisation of the national residues control plan in comparison with previous years. The timely implementation of the plan and its supervision has markedly improved. The comprehensive scope of the complementary residues testing programme has addressed all of the previously identified deficiencies in the coverage of the residues testing programme relative to Community requirements. Under this programme, the use of private accredited laboratories to carry testing using validated analytical methods has resulted in the satisfactory completion of residues testing in honey for 2006. For the other species, covered by the complementary programme the implementation is not complete as methods have yet to be validated to the standard set by the competent authority and there are gaps in key areas such as malachite green in aquaculture, and topical chemicals for ectoparasite control in cattle. In tandem with this programme, the reorganisation of laboratory co-ordination and substantial investment in state of the art equipment and training of staff in the governmental laboratory network clearly demonstrates the commitment of the competent authority to raise performance standards in this area. Progress is being made in efforts to accredit the governmental laboratories to ISO 17025. Nevertheless, and in contrast to the private laboratories, accreditation to this international standard is still not in place for any of the governmental laboratories and the use of inappropriately validated methods was seen. With regard to the authorisation, distribution and control of veterinary medicines, the competent authority has been active in reviewing, overhauling and improving the regulatory framework for veterinary medicines and there are plans in place to significantly extend the range of veterinary medicines for which a veterinary prescription is required and in regard to cattle eligible for export to the EU, there is now a requirement for the keeping of medicines records. Overall, it can be concluded that the competent authority has worked hard and committed substantial resources to rectify the deficiencies in residues and veterinary medicines controls identified in previous FVO residues missions, thereby increasing confidence in the chemical safety of food of animal origin exported to the EU. The report makes a number of recommendations to the Brazilian competent authorities, aimed at rectifying the shortcomings identified and enhancing the implementing and control measures in place. i

TABLE OF CONTENTS 1. INTRODUCTION...1 2. OBJECTIVES AND SCOPE OF THE MISSION...1 3. LEGAL BASIS FOR THE MISSION...1 4. BACKGROUND...2 4.1. COUNTRY STATUS IN RELATION TO SUBMISSION OF RESIDUES CONTROL PLANS...2 4.2. SUMMARY OF PREVIOUS MISSION RESULTS...2 4.3. RAPID ALERT SYSTEM FOR FOOD AND FEED (RASFF) NOTIFICATIONS FOR CONSIGNMENTS FROM BRAZIL CONCERNING RESIDUES...3 4.4. PRODUCTION AND TRADE INFORMATION...3 5. MAIN FINDINGS...4 5.1. NATIONAL RESIDUE CONTROL PLAN (NRCP) AND COMPLEMENTARY (RESIDUES TESTING) PLAN...4 5.2. OTHER RESIDUES CONTROL PROGRAMMES...9 5.3. FOLLOW-UP OF NON-COMPLIANT RESULTS...9 5.4. LABORATORIES...10 5.5. VETERINARY MEDICINAL PRODUCTS AND MEDICATED FEEDINGSTUFFS...18 6. CONCLUSIONS...22 6.1. NATIONAL RESIDUE CONTROL PLAN AND THE COMPLEMENTARY PLAN...22 6.2. OTHER RESIDUES CONTROL PROGRAMMES...22 6.3. LABORATORIES...23 6.4. VETERINARY MEDICINAL PRODUCTS AND MEDICATED FEEDINGSTUFFS...23 6.5. OVERALL CONCLUSION...24 7. CLOSING MEETING...24 8. RECOMMENDATIONS...25 9. COMPETENT AUTHORITY RESPONSE TO RECOMMENDATIONS 26 Annex I Applicable Community Standards 27 ii

ABBREVIATIONS & SPECIAL TERMS USED IN THE REPORT ANVISA AOZ, AMOZ, AHD, SEM Agência Nacional de Vigilância Sanitária marker residues for the nitrofurans furazolidone, furaltadone, nitrofurantoin and nitrofurazone respectively HPLC (DAD) (UV) IMA (C)CA (Central) Competent Authority INMETRO CCα/ß Decision limit/ Detection Capability ISO CCRC CGAL DFIP DG(SANCO) DIPOA DSA Coordenação de Controle de Resíduos e Contaminantes Co-ordination of the control of residues and contaminants Coordenação Geral de Apoio Laboratorial General Coordination of Laboratories Departamento de Fiscalização de Insumos Pecuários Department of Livestock Input Inspection Health and Consumer Protection Directorate General Departamento de Inspecao de Produtos de Origem Animal - Department of Inspection of Products of Animal Origin Departamento de Saude Animal - Department of Animal Health JECFA LANAGRO LoD LoQ MAPA MERCOSUL High Performance Liquid Chromatography (Diode Array Detection) (Ultra Violet) Instituto Mineiro de Agropecuária Minas Gerais Agriculture and Livestock Institute Instituto Nacional de Metrologia, Normalização e Qualidade Industrial Brazilian National Accreditation Body International Standardisation Organisation The Joint FAO/WHO Expert Committee on Food Additives Laboratório Nacional Agropecuário National Animal and Plant Laboratory Limit of Detection Limit of Quantification Ministério da Agricultura, Pecuária e Abastecimento Ministry of Agriculture, Livestock and Food Supply Mercado Comum do Sul. A Regional Trade Agreement between Brazil, Argentina, Uruguay, Venezuela, and Paraguay to promote inter alia, free trade. EC/ EU European Community / European Union ML Maximum Level FVO Food and Veterinary Office MRL Maximum Residue Limit GC/LC-MS- MS Gas / Liquid Chromatography-(tandem) Mass Spectrometry MRPL Minimum Required Performance Limit Group A, B Categories of substances listed in Annex I to Council Directive 96/23/EC: A1 Stilbenes; A2 Thyrostats A3 Steroids; A4 Zeranol A5 Beta-agonists A6 Substances listed in Annex IV to Council Regulation (EEC) No 2377/90 B1 Inhibitors (antimicrobials) B2a Anthelmintics; B2b Coccidiostats B2c Carbamates and pyrethroids B2d Sedatives; B2e NSAIDs B2f Others (e.g. corticosteroids) B3a Organochlorines including PCBs B3b Organophosphorus compounds B3c Chemical elements B3d Mycotoxins B3e Dyes; B3f Others NRCP RASFF SCR/DIPOA SDA SEDESA SEFAG National Residue Control Plan Rapid Alert System for Food and Feed Setor de Controle de Resíduos do DIPOA - Residue control section of DIPOA Secretaria de Defesa Agropecuaria Secretariat of Animal and Plant Protection Serviço de Saude Animal Service for the Protection of the Health of Livestock Serviço de Fiscalização Agropecuária State Service of Animal and Plant Inspection iii

SFA Superintendencia Federal de Agricultura, Pecuária e Abastecimento - Federal Superintendence of Agriculture, Livestock and Food Supply SOP Standard Operating Procedure SIF Serviço de Inspeção Federal Federal Inspection Service TLC Thin Layer Chromatography SIPAG Serviço de Inspeção de Produtos Agropecuários. Agriculture and Livestock Product Inspection Service VMP Veterinary Medicinal Product SISBOV Brazilian System of Identification and Certification of Origin for Cattle and Buffalo CONVENTIONS USED IN THE REPORT Bullet points marked thus indicate findings made by the mission team on the basis of observations on the spot or assessment of information received. iv

1. INTRODUCTION The mission took place in Brazil from 27 February to 8 March 2007. The mission team comprised two inspectors from the Food and Veterinary Office (FVO). The mission was undertaken as part of the FVO's planned mission programme, evaluating control systems and operational standards in this sector. Representatives from the competent authorities (CA) accompanied the inspection team during the whole mission. An opening meeting was held on 27 February 2007 with the central competent authority (CCA) - Ministry of Agriculture, Livestock and Food Supply (MAPA) - and other relevant CA. At this meeting, the objectives of, and itinerary for, the mission were confirmed by the inspection team and the control systems were described by the CCA. 2. OBJECTIVES AND SCOPE OF THE MISSION The objective of the mission was to evaluate the implementation of national measures, aimed at the control of residues and contaminants in live animals and animal products, including the controls on the distribution and use of veterinary medicinal products (VMPs) and feed additives, the use of which may give rise to residues in such products. The evaluation was based on the standards set out in Council Directive 96/23/EC 1 and other relevant Community legislation in this field, including legislation on the control and distribution of VMPs. The mission focused on the actions taken by the CCA in response to outcome of the previous residues mission (DG SANCO 7712/2005 MR Final) which took place in November 2005. Particular attention was paid to verifying CCA commitments to improve laboratory analytical capability and broaden the scope of residues testing across a number of commodities. COMPETENT AUTHORITY VISITS Competent Authorities LABORATORY VISITS Governmental laboratories Private Laboratories 4 FARMS OTHER SITES Comments Central 3 Opening, closing and a separate pre-closing clarification meeting with the Ministry of Agriculture, Livestock and Food Supply (MAPA) Regional 2 State MAPA Offices in the States of Goiás and Minas Gerais 3. LEGAL BASIS FOR THE MISSION LANAGRO Campinas, LANAGRO Minas Gerais carrying out 2 testing under the National Residues Control Plan All carrying out testing under the National Residues Control Plan and the Complementary Residues Testing Programme and located in São Paulo and Rio de Janeiro. 2 Cattle feedlots in the States of Goiás and Minas Gerais 1 Pharmacy for Veterinary Medicinal Products Cooperativa Agropecuria in Minas Gerais The mission was carried out under the general provisions of Community legislation and, in particular: 1 EU legal acts quoted in this report are listed in Annex I and refer, where applicable, to the last amended version. 1

Council Directive 96/23/EC on measures to monitor certain substances and residues thereof in live animals and animal products, and repealing Directives 85/358/EEC and 86/469/EEC and Decisions 89/187/EEC and 91/664/EEC, and in particular Article 29 (3). Article 46 of Regulation (EC) No. 882/2004 of the European Parliament and of the Council; Commission Decision 98/140/EC laying down certain detailed rules concerning on-the-spot checks carried out in the veterinary field by Commission experts in third countries. 4. BACKGROUND 4.1. COUNTRY STATUS IN RELATION TO SUBMISSION OF RESIDUES CONTROL PLANS Commission Decision 2004/432/EC (as last amended by Commission Decision 2007/115/EC), indicates that Brazil National Residues Control Plan (NRCP) is approved in accordance with Council Directive 96/23/EC for the following species/commodities: bovine, equine, poultry and aquaculture products. As a result of the outcome of the previous FVO residues mission (henceforth referred to as the 2005 mission), Brazil was delisted for honey and farmed game in March 2006 (Commission Decision 2006/208/EC) as no residues plans were in place for either commodity. At the request of the CCA, Brazil was delisted for milk and ovine/caprine casings in February 2007 (Commission Decision 2007/115/EC) and it was also delisted for swine at the same time for the reason that a beta-agonist is authorised for use in swine as a growth promoter and there is no split system in place guaranteeing that meat from treated animals would not (animal health conditions permitting) be exported to the EU in accordance with Article 11 of Council Directive 96/22/EC. 4.2. SUMMARY OF PREVIOUS MISSION RESULTS The Brazilian control system for residues of veterinary medicines and contaminants in food of animal origin was evaluated by the FVO in missions carried out in 1999, 2003 and 2005. The most recent residues missions to Brazil were undertaken in June 2003 - DG(SANCO)/9047/2003 - and November 2005 - DG(SANCO)/7712/2005 - the reports of which have been published on the Health and Consumer Protection Directorate General web site at: http://europa.eu.int/comm/food/fs/inspections/vi/reports/brazil/vi_rep_braz_9047-2003_en.pdf and http://www.ec.europa.eu/food/fvo/act_getpdf.cfm?pdf_id=5189 The missions identified significant deficiencies in the NRCP for a number of commodities. In both instances the Brazilian authorities subsequently submitted a detailed action plan to the Commission services intended to address the shortcomings identified. The 2005 mission concluded that many of the undertakings given by the CCA in response to the 2003 report had not been implemented and actions taken by the Commission included the delisting of honey and farmed game (see 4.1. above) and maintaining frequent correspondence with the CCA (including several face to face meetings) in relation to ensuring that the actions promised by the CCA on foot of the 2005 mission and published on: http://www.ec.europa.eu/food/fvo/act_getpdfannx.cfm?anx_id=4686 were implemented. The purpose of the present mission was to verify on-the-spot, how 2

the competent authority had implemented promised improvements to the residues control system since the 2005 mission. 4.3. RAPID ALERT SYSTEM FOR FOOD AND FEED (RASFF) NOTIFICATIONS FOR CONSIGNMENTS FROM BRAZIL CONCERNING RESIDUES In contrast to the position detailed in both the 2003 and 2005 missions, there have been no findings of residues of veterinary medicinal products from 2006 to date reported under the RASFF. 4.4. PRODUCTION AND TRADE INFORMATION Detailed information on the quantities of food commodities (of animal origin) produced in 2006 and the amounts exported to the EU were supplied by the CCA. Commodity National production Exported to the EU Bovine meat 24,615,358 animals 322,982 tonnes Poultry meat and products (incl. turkey) 3,931,828,171 birds 351, 479 tonnes Horse meat 154,594 animals 14,011 tonnes Aquaculture Shrimps 63,134 tonnes 31,350 tonnes Aquaculture finfish 179,746 tonnes 59 tonnes Honey 37,520 tonnes 1,103 tonnes* Eggs 26,536,242,398 eggs 116.8 tonnes** Live animals (including inter alia horses for breeding and production and sport horses) 77,393 animals * According to the CCA, exported before the ban on honey exports came into effect - March 2006. ** Dried Eggs = 84,000 kg and Dried yolk = 32,000 kg. The mission team noted that: exports of egg products have taken place to the EU even though Brazil is not listed in Commission Decision 2004/432/EC for this commodity (i.e. has an approved residues plan). This issue was raised during the 2005 mission. An economic chamber in one Member State had issued a letter to the exporting establishment on 9 December 2005 stating that the export of egg products from the one premises involved in Brazil was permitted and therefore this trade continued. (See remarks on this point in section 7 Closing meeting). 3

5. MAIN FINDINGS 5.1. NATIONAL RESIDUE CONTROL PLAN (NRCP) AND COMPLEMENTARY (RESIDUES TESTING) PLAN 5.1.1. Bodies involved Since the 2005 mission report there have been some changes in the names and responsibilities of some of the bodies involved in residues and veterinary medicines controls. These are listed in the organigramme below. MAPA Federal level DFIP DSA SDA DIPOA CCRC CGAL Functions Controls on VMPs and animal feeds Follow up of all VMP violations except hormones in cattle Live animal sampling and follow up of hormone violations in cattle Slaughterhouse sampling Sampling other than in slaughterhouses or on farm (e.g. honey and fish processors) Co-ordinating all residues testing and actions of MAPA sectors involved in the NRCP Lab co-ordination State Level SEFAG SEDESA SIPAG SIF SFA (Representation of MAPA at state level) 5.1.2. NRCP The process of planning the NRCP has been described in the 2005 mission report. The 2006 NRCP was elaborated by ordinance No 50 of 20 February 2006. The 2007 NRCP is about to be published. 2 5.1.3. Complementary residue testing programme for cattle, poultry, equidae, aquaculture (finfish and shrimp), pigs and honey In response to the outcome of the 2005 mission and, in response to FVO comments on the scope of the 2006 NRCP in May 2006 and following discussions with the Commission services in July 2006, the CCA presented the complementary plan to 2 In its response to the draft report, the Brazilian Competent Authority stated that the 2007 NRCP and the results of the 2006 MRCP were published on 30 March 2007. 4

the Commission services in August 2006 which entered into force by means of Memorandum SCR/DIPOA no. 178/2006, of August 15, 2006. The complementary plan is targeted on commodities which are currently, or which have the potential to be, exported to the EU. The CCA aims in the medium term to integrate this complementary plan into the NRCP. 5.1.4. Planning of the NRCP and the complementary plan The mission team noted that: relative to the 2005 mission, there have been no changes to the process by which the 2006 NRCP was planned. All relevant bodies are involved; in contrast to the situation for the 2005 NRCP, the 2006 NRCP which was submitted to the Commission services included detailed results of the 2005 NRCP; the 2006 NRCP covers all of the commodities for which Brazil is listed in the Annex to Commission Decision 2004/432/EC (approved residues plan) - : bovine animals (including live animal sampling), equidae, poultry, aquaculture (finfish and shrimp). In addition, the plan also covers swine, milk, eggs, and honey. The latter commodity was not included in the 2005 plan; in comparison with the testing requirements of Council Directive 96/23/EC, the 2006 NRCP (in common with the 2005 NRCP) did not cover many substance groups, several of which were seen as essential for monitoring. This was communicated to the CCA and in August 2006 a complementary residues testing programme (complementary plan see 5.1.3) has been progressively put in place for those commodities being exported to the EU (plus swine and honey) with the aim of addressing the gaps in the NRCP. The present situation is summarised in the following table. Comparison of the requirements of Council Directive 96/23/EC with the current Brazilian residues testing programmes for food of animal origin Group Bovine Equine Poultry Aquaculture Honey 96/23 NRCP Compl 96/23 NRCP Compl 96/23 NRCP Compl 96/23 NRCP Compl 96/23 NRCP Compl A1 Stilbenes Y Y Y N Y Y Y Y* N Y A2 Thyrostats Y N N Y N Y Y** N Y A3 Steroids Y Y Y Y N Y Y N Y Y* N N A4 Zeranol Y Y Y N Y Y Y A5 Beta-agonists Y Y Y Y N Y Y Y Y A6 Annex IV Y Y Y Y Y Y Y Y Y Y Y Y Y N Y B1 Antimicrobials Y Y Y Y Y Y Y Y Y Y N Y Y Y Y B2a Anthelmintics Y Y Y Y Y Y Y Y Y Y* N N B2b Coccidiostats Y Y Y Y N Y Y Y Y B2c Carb / pyreth Y N Y Y N Y Y N Y Y N Y B2d Sedatives Y N Y Y N Y B2e NSAIDs Y N Y Y N Y Y N Y B2f Other vet drugs Y B3a Organochlorines Y Y Y Y Y Y Y N Y Y N Y B3b Organophosphates Y N Y Y N Y Y N Y B3c Heavy Metals Y Y Y Y Y Y Y Y Y N Y B3d Mycotoxins Y N Y Y N Y Y N Y Y N Y B3e Dyes Y N Y Code: Not required by 96/23 Y = Yes; N = No * Only relevant for finfish, not shrimp ** Member States are not expected to analyse for thyrostats in poultry 5

taking both plans together, all of the substance groups seen as essential 3 for residues monitoring in order to be considered equivalent to Community requirements are now covered, with one exception Group A2 (thyrostats) is not currently tested for in bovine animals. 4 Substances regarded as highly desirable 2 for residues monitoring are also included; in common with the findings of the 2005 mission, the scope of some substance groups in the NRCP and the complementary plan is limited, in particular for the range of steroids and beta-agonists tested. Concerning residues of products used for tick control, whilst there are several compounds being analysed in each of the carbamate, pyrethroid and organophosphate groups, several important pharmacologically active substances which are authorised for use in food producing animals in Brazil are not presently included in the complementary plan e.g. cypermethrin, chlorfenvinphos, coumaphos, and amitraz. 5 However, for other substance groups, the scope of testing has been greatly increased (e.g. anticoccidials) and, with the implementation of the complementary plan many new substance groups are being covered for the first time; testing under the NRCP is carried out in both the four main LANAGRO governmental laboratories and in four private sector laboratories. All of the testing under the complementary plan is currently carried out in eight private sector laboratories, seven of which are located in Brazil with the remaining laboratory in Germany; this increase in the coverage of the entire Brazilian residue testing programme has been effected by a considerable increase in investment in governmental laboratory infrastructure, equipment, training of staff and utilisation of private laboratories to address the gaps in analytical expertise. 5.1.5. Implementation of the NRCP The implementation of the NRCP has been described in the 2005 mission report and the means by which this is effected has not differed significantly. All samples are taken by competent authority officials. Slaughterhouse samples are taken by SIF inspectors. On-farm urine sampling of cattle is undertaken by SEDESA. Other sampling (e.g. honey, fish etc) is carried out by SIPAG and Federal Inspectors located at State level sample milk from dairy farms and from milk processing establishments. The mission team noted that: in marked contrast to the situation observed in the 2005 mission, the 2006 NRCP was fully realised for the majority of analytes i.e. practically all samples foreseen to be taken were actually sampled and analysed as planned. For example, in the case of poultry, 7970 samples were foreseen to be taken. A total of 8213 analyses were carried out (3% more than required). For cattle the figures were 5533 and 5483 respectively (0.9% less than expected); 3 4 5 Description given on the DG SANCO webpage for residues controls in third countries: http://ec.europa.eu/comm/food/food/chemicalsafety/residues/third_countries_en.htm In its response to the draft report, the Brazilian Competent Authority stated that: the analytical method for thyrostats is expected to be fully validated by 25 May 2007 in one of the private laboratories carrying out analyses under the NRCP. the analytical methods for these substances are at an advanced stage in the validation process and that the FVO will be informed when samples are to be analysed for these substances. 6

in cases where it was not possible to carry out sampling in the timeframe given (e.g. because of no slaughtering in the establishment designated for sampling) or where samples arrived in a poor condition at the laboratory and were unsuitable for analysis, CCRC was informed and DIPOA reallocated these samples accordingly; with regard to on-farm sampling of live cattle (for residues of growth promoting hormones and beta-agonists), sampling is preferentially targeted on those farms with feedlot production systems rather than extensively grazed cattle. According to data provided by the CCA, feedlot production accounts for 0.55% of the cattle produced in Brazil; sampling is distributed evenly over the entire year with the exception of live animal sampling for cattle which takes place over the period from March December, which coincides with the period in which cattle would be reared in feedlots; 5.1.6. Implementation of the complementary plan The mission team noted that: sampling under the complementary plan is carried out as per the NRCP. The same personnel are involved and the same sampling forms etc are used. The target number of samples per substance group for 2006 was, with the exception of honey (see below), 60 samples per substance; as methods have been validated for the relevant matrices in each of the laboratories, sampling has been progressively introduced from June 2006 to date. Not all of the methods for individual analyte-matrix combinations are validated according to the standards required by the General Coordination of Laboratories (CGAL). According to data provided to the mission team before the mission the percentage of fully validated methods is 44% (143/328) though, on the spot in some of the private laboratories visited, it could be seen that there were more methods validated to the required standard which had not yet been formally communicated to CGAL. CGAL does not permit analyses to be carried out (under either of the residue testing programmes) before it has evaluated the data and approved the method. Consequently analyses for unapproved analyte-matrix combinations should not be carried out; in the case of honey, practically all of the methods have been validated 6 and are in place in the Brazilian laboratories. For this commodity, the number of samples was set at 30 samples per substance, taken only from the eleven establishments which had been exporting honey to the EU prior to the ban being implemented. To date 336 honey samples out of a planned 360 (93%) have been tested for all relevant substance groups. Samples were tested both in the private sector laboratories in Brazil and also in an internationally recognised laboratory for honey analysis in Germany. The results have been in agreement between the Brazilian and German laboratories and there have been no noncompliant results for this commodity; 6 In its response to the draft report, the Brazilian Competent Authority stated that all analytical methods for VMP residues in honey are now validated the outstanding substance erythromycin has now been completed. (Further details are in the action plan hyperlinked to this report). 7

for bovine animals 660 samples were foreseen to be taken (60 samples for each of 11 substance groups). To date there have been 305 analyses and no noncompliant results. In particular, Groups B2c (carbamates and pyrethroids) and B3b (organophosphates) have yet to be analysed. 7 Validation data on the methods for Group A5 (in particular, ractopamine), A6 (in particular nitroimidazoles), Group B2c (Carbamates and pyrethroids), Group B2e (nonsteroidal anti-inflammatory drugs), Group B3b (organophosphates) and Group B3d (mycotoxins), have not been forwarded to CGAL from the relevant laboratories. However, analyses for mycotoxins in this commodity have been carried out under the programme; for poultry, equine, aquaculture finfish and aquaculture shrimp there have been 592, 304, 252 and 125 analyses respectively and no non-compliant results for any of these commodities; for equine, validation data on the methods for Group A2, A3 (in particular boldenone), A6 (in particular chloramphenicol and nitrofurans), Group B2a, B2c (pyrethroids only), Group B2e (non-steroidal anti-inflammatory drugs) and Group B3d (mycotoxins), have not been forwarded to CGAL from the relevant laboratories. However, analyses for boldenone and mycotoxins have been carried out under the programme; for poultry, validation data on the methods for Group A2, A3 (in particular boldenone), A6 (in particular chloramphenicol though this is covered in the NRCP), Group B1 (confirmatory methods for several antimicrobials), B2a (in particular levamisole), B2c (Carbamates and pyrethroids) and Group B2e (nonsteroidal anti-inflammatory drugs) have not been forwarded to CGAL from the relevant laboratories. In this case no analyses for substances which are not currently validated have been carried out; for aquaculture, validation data on the methods for Group A6 (in particular chloramphenicol), Group B1 (confirmatory methods for several antimicrobials), B3a (organochlorines) and B3e (in particular malachite green) have not been forwarded to CGAL from the relevant laboratories. However, analyses for malachite green 8 have been carried out under the programme (see section on laboratories). 5.1.7. Supervision of implementation of NRCP and the complementary plan As reported in the 2005 mission, the CCRC is responsible for the overall supervision of the implementation of the NRCP and the complementary plan. The mission team noted that: new personnel appointments were made within the CCRC during early 2006; the new CCRC co-ordinator kept a weekly check on the progress of sampling and analysis and had devised a computer application to facilitate this activity, 7 8 In its response to the draft report, the Brazilian Competent Authority stated that: analytical methods for carbamates, pyrethroids and organophosphates in bovine have already been validated by one of the private laboratories and that CGAL is still evaluating these validation data. When complete the samples will be taken and analysed. (Further details are in the action plan hyperlinked to this report). that the analytical method for malachite green and leucomalachite green has now been correctly validated by the relevant private laboratory completed 10 May 2007. (Further details are in the action plan hyperlinked to this report). 8

identify the source of problems (if and when they occurred) and take appropriate corrective actions; in contrast to the findings of the previous mission, supervision of implementation was effective. 5.2. OTHER RESIDUES CONTROL PROGRAMMES 5.2.1. Nitrofurans monitoring programme for poultry meat The programme ran from February 2003 to December 2006. During this period a total of 82,620 samples of poultry meat were analysed for residues of four metabolites of nitrofurans (AOZ, AMOZ, AHD and SEM) in three accredited private laboratories (405 samples per week from all of the establishments exporting poultry to the EU). There were 86 non-compliant results (0.1%) overall. All of the results from 2005 and 2006 were compliant. On this basis the CCA wrote to the Commission Services in December 2006 proposing that the programme should be stopped and testing resources redirected into other areas. This was agreed by the Commission Services. Testing for residues of the nitrofuran metabolites is continuing under the NRCP. 5.3. FOLLOW-UP OF NON-COMPLIANT RESULTS Further to Memorandum CCRC/SDA Nº 03/2007, the follow up of non-compliant results has changed in 2007 whereby DSA (through SEDESA at state level) will continue to be responsible for carrying out on-farm investigations in the case of findings of hormone residues in cattle, however, DFIP (through SEFAG at state level) will be responsible for investigation on farms when findings of other residues of veterinary medicines are detected. In 2006, SEDESA was entirely responsible for this activity. The mission team noted that: DFIP has elaborated a new manual of procedures for dealing with noncompliant results. There were a total of 27 non-compliant results in the 2006 NRCP for commodities being exported to the EU. For cattle these included zeranol (3), anthelmintics (11), heavy metals (2), for equine, heavy metals (1), for aquaculture fish, heavy metals (3), for poultry, antibiotics (2), coccidiostats (2) and heavy metals (3). No residues of any Group A6 Annex IV substances (e.g. chloramphenicol, nitrofurans), Group A5 (beta-agonists), Group A3 (steroids) or Group A1 (stilbenes) were detected. To date in 2007, there has been one non-compliant result (ivermectin) in cattle. In the complementary programme there have been no non-compliant results. Several follow up files from non-compliant results in 2006 were examined at random by the mission team and it was observed that: in all cases the files were complete and actions had been carried out by the relevant services in accordance with national rules, however, the system of information exchange between the relevant services was rather bureaucratic and hindered rapid investigations on farm. The CCA considers that the new arrangements for follow-up in 2007 will expedite the rapid completion of investigations; the three zeranol non-compliant results were also tested for the main zeranol metabolite α-zeranol (taleranol) plus α- and β-zearalenol, zearalenone and 9

zearalanone. In each case the laboratory did not consider that the non-compliant results were due to contamination due to the absence of α- and β-zearalenol, zearalenone and zearalanone. In 2 cases on-farm investigations and follow-up samples were negative. In the last case, the follow up samples were also noncompliant and the farm is still blocked. The investigation of this farm is ongoing. 5.4. LABORATORIES 5.4.1. General description The role and responsibilities of the General Coordination of Laboratories (CGAL) was described in detail in the 2005 mission report. Four official MAPA laboratories are designated by CGAL for the analysis of NRCP samples: LANAGRO/MG in Pedro Leopoldo, LANAGRO/SP in Campinas, LANAGRO/RS in Porto Alegre and LANAGRO/PE in Recife. In addition, four private laboratories are approved by CGAL for carrying out (other) analyses under the NRCP. These four private laboratories plus, a further three (in Brazil) and one in a European Member State are solely carrying out testing under the complementary programme. Normative Instruction No 1 of 2007 requires inter alia that CGAL approved laboratories (i.e. private laboratories) are accredited to ISO 17025. Before private laboratories are permitted to use methods for either the NRCP or the complementary programme, CGAL has to evaluate the validation data and, if acceptable, the method is approved by way of publication in the Official Brazilian Government Gazette. The mission noted that: as was the case in the 2005 mission, none of the four LANAGRO laboratories for the NRCP are accredited according to ISO 17025. All of the private laboratories are accredited to this standard; there had been a plan for CGAL to apply to be certified as an accreditation body by a foreign organisation and that CGAL would accredit the LANAGRO laboratories by the end of 2006. However, this idea was abandoned and a plan to employ a consultant company to expedite the ISO accreditation process in each of the LANAGRO laboratories is at an advanced stage, awaiting final approval in the 2007 Federal budget; CGAL does not have a defined timetable for achieving ISO accreditation in the LANAGRO laboratories. However, CGAL has already contacted INMETRO (the Brazilian national accreditation body) and preliminary third party audits by state metrology institutes have been agreed for three of the LANAGRO laboratories as an initial preparation for an official INMETRO audit. LANAGRO MG has very recently been audited under this arrangement; CGAL has mandated that analytical methods used for the NRCP and the complementary programme must be validated in accordance with the main elements of Commission Decision 2002/657/EC; CGAL has audited the private laboratories involved in the NRCP and complementary programmes. The personnel involved in these audits have included LANAGRO chemists; 10

under the auspices of the Agência Nacional de Vigilância Sanitária (ANVISA) and MAPA, CGAL started a four year project in 2006 to deliver proficiency testing material to all laboratories in the governmental and private laboratory networks. The first of these proficiency test samples for residues produced by the United Kingdom Food Analysis Performance Assessment Scheme (FAPAS) will be delivered in 2007 and FAPAS will analyse the results under this project; in contrast to the position in 2005, there has been significant investment in training for LANAGRO personnel. Contacts have been made with leading European and Canadian laboratories. In Sept 2006, a Canadian expert gave a seminar on method validation in LANAGRO SP, CGAL personnel received training under the DG TRADE sponsored SARAF programme at the NRL in Nantes, France, in June 2006, twenty six laboratory staff received on the spot training on analytical methods and quality control in the UK at a National Reference Laboratory in December 2006, training with the Community Reference Laboratory in Berlin is being organised for the first semester in 2007, and staff have attended an international residues conference in Europe; in contrast to the position in 2005, the LANAGRO laboratories (in particular LANAGRO MG and SP) have received significant financial support for the purchase of modern analytical equipment capable of delivering state of the art residue analyses. CGAL is still in the process of deciding how new equipment will be distributed in the LANAGRO network in the future, in line with analytical testing capability. 5.4.2. On the spot visits in the LANAGRO laboratories Two LANAGRO laboratories were visited during the course of the mission LANAGRO/MG in Pedro Leopoldo and LANAGRO/SP in Campinas. 5.4.2.1. LANAGRO/MG in Pedro Leopoldo Under the NRCP, this laboratory is responsible for analysing Group A1, A3, A4, B1, B2a, and B3c in those commodities which are currently exported to the EU. This laboratory was visited during the 2005 mission. The mission team noted that: the laboratory is very well equipped with modern state of the art instruments and has adequate space; there is a quality management system in place in line with ISO 17025 requirements, including regular internal audits; there is an SOP for method validation including the essential elements of Commission Decision 2002/657/EC; in addition to the participation in the ANVISA/MAPA planned proficiency testing exercise, there is a comprehensive programme for participation in internationally recognised proficiency tests for 2007 covering residues of veterinary medicines and heavy metals; in 2006 the laboratory organised three rounds of national proficiency tests with some of the other LANAGRO laboratories for heavy metals and for four sulphonamides. 11

Each of the analytical methods used for the NRCP were examined in detail by the mission team. It was noted that: in each case, method SOPs were available and in a consistent format, however, in some cases these were incomplete as the full scope of analyses carried out in the method was not reflected in the SOP (e.g. avermectins); each SOP included a requirement for quality control checks in every batch; validation data were not presented in a uniform manner or always summarised whereby it was possible to clearly identify all of the key performance characteristics of an analytical method (e.g. recovery, repeatability, reproducibility); antimicrobial screening was carried out using a commercially available microbiological growth inhibition test. This had been comprehensively validated by the laboratory for seven antimicrobial substances; for the calculation of CCα and CCβ, for instrumental methods (e.g. MS methods) a spreadsheet had been constructed using the calibration curve approach advocated in ISO 11843; there were SOPs for several LC-MS-MS methods for the chemical confirmation of antibiotic residues. None of the methods, which had been taken from the scientific literature, had yet been validated according to Commission Decision 2002/657/EC (no CCα or CCβ) nor was there any data in the validation file on analytical recovery, repeatability or reproducibility. All are included in the validation plan for 2007 presented by the laboratory. The use of some compounds as internal standards (e.g. penicillin V in the penicillins method and sulphapyridine in the sulphonamide method) could be problematical as these substances may well be authorised for use in food producing animals; for hormones in bovine urine, a USDA (FSIS) method (GC-MS n ion trap) had been adapted covering Group A1 and A4. Deuterated diethylstilbestrol (d8) was used as an internal standard. Zeranol and its metabolites were included in the method, however, the recovery data had been generated some years previously using different analytical equipment and was therefore invalid. These old data were mixed up with new data generated using the new equipment i.e. CCα and CCβ were calculated using the new equipment; trenbolone (Group A3) was originally included in the method but due to unacceptable performance data in the validation process, was no longer tested for. This method does not therefore cover any Group A3 substances (steroids); an HPLC method for avermectins (abamectin, doramectin and ivermectin) was in place for cattle, horse, pig and poultry liver. Validation had only been carried out in bovine liver recovery, repeatability, reproducibility, LoD and LoQ had been calculated (not yet CCα or CCβ). [In contrast, CCα and CCβ had been calculated for the above analytes (plus eprinomectin) in milk]. 5.4.2.2. LANAGRO/SP in Campinas, São Paulo This laboratory was visited in both the 2003 and the 2005 missions. Under the NRCP the laboratory is responsible for analysing Group A1 (diethylstilbestrol) and Group A4 (zeranol) in bovine liver and Group B3a (organochlorine pesticides) in fat from cattle, equidae, poultry, swine and milk. The mission team noted that: 12

the laboratory had just received the CGAL organised exploratory audit from the metrology network but had not yet received an official report of the findings; the 2006 NRCP was largely realised by the laboratory all 60 bovine liver samples had been analysed and 94% of the organochlorine samples had been tested (1180/1257). In comparison to 2005, the number of analyses for which the laboratory was responsible had been reduced in order to facilitate the validation of methods and installation of new equipment four mass spectrometers and one HPLC; the staffing level in the residues laboratory is quite low 9 chemists/technicians. Whilst sufficient for the current NRCP commitments, the validation plan to validate 12 new methods in various matrices over the coming 18 24 months will be difficult to achieve, particularly in the mass spectrometry section where only two chemists are employed; there is no validation SOP as such but there are guidelines in place for validation. These differ from the model used in LANAGRO MG. there are SOPs in place for the two methods the laboratory is performing. The high resolution GC-MS method for diethylstilbestrol and zeranol in bovine liver was only finalised a few days prior to the mission. This method was examined in detail. The method which was an FSIS method was not fit for purpose. The data for validation (which was ongoing during the 2005 mission) presented showed that the method was not in statistical control during the validation phase (recoveries for DES ranged from 26 225%, outside the FSIS recommended range of 80-130%. There were no data collated on repeatability or reproducibility and only partial validation data for zeranol. CCα and CCβ had not been calculated. This partially validated and unreliable method had been used for routine analysis and a screening positive zeranol sample sent from a private laboratory had been confirmed as compliant using this method; the method for OC pesticides (excluding PCB cogeners) was functioning properly and was validated to the old standard i.e. LoD and LoQ had been determined. However, CCα and CCβ had not been calculated; a new method for chemical elements was being validated and CCα and CCβ had been calculated for lead; some experimental work on nitrofurans by LC-MS-MS was being carried out in the mass spectrometry section. Only standards in solution had been analysed to date. This was also the case when the laboratory was visited in 2003 and 2005. 5.4.2.3. Private sector laboratory A in Rio de Janeiro Under the NRCP, this laboratory is responsible for analysing Group A6 (nitrofurans in cattle, poultry, swine, shrimp and finfish) and, under the complementary programme, Groups A5 (ractopamine in cattle, poultry, equidae and swine), B1, B2b and B2f (carbadox and olaquindox), A4 and B3a in cattle, poultry and equidae and swine, all of which (with the exception of swine) are exported to the EU. This laboratory was also visited during the 2003 mission. The mission team noted that: the laboratory is accredited to ISO 17025 by INMETRO. The methods included in the complementary programme are not yet included in the scope of 13

accreditation. A quality control system is in place with regular calibration of equipment, control charts etc; the laboratory is well equipped with 2 LC-MS-MS solely dedicated for residue analysis carried out under both programmes. The laboratory also carries out inter alia, dioxin analysis with high resolution GC-High Resolution MS; the method for nitrofurans for poultry and the majority of methods in the complementary programme for poultry have comprehensive SOPs in place and have been validated to Commission Decision 2002/657/EC for this species ractopamine, lincomycin (in a multi method including tetracyclines and tylosin), ionophore coccidiostats (lasalocid, monensin, salinomycin, narasin and maduramycin) and the coccidiostats clopidol, robenidine, amprolium, nicarbazin, dicalzuril together with the antibacterial trimethoprim and the banned quinolone drugs carbadox and olaquindox in one multi method; not all of the methods above have yet been approved by CGAL (and published in the Official Brazilian Government Gazette) the laboratory has been approved for coccidiostats and nitrofurans and therefore only samples for these analytes have been sent to the laboratory for analysis under both the NRCP and complementary programme, in line with CGAL rules; the laboratory has internally fixed a deadline of 60 days in which to complete the validation of all other analyte-matrix combinations i.e. the extension of existing validated methods in poultry muscle to bovine and equine muscle (e.g. ractopamine in cattle and equine). CGAL have requested a full validation for the extension even though partial validation could be sufficient to demonstrate fitness for purpose in these other muscle matrices; staff stated that validation of the new methods had been carried out in poultry muscle first as the laboratory is familiar with this matrix (having been testing for nitrofurans in poultry meat since 2003) and that delays in validation for (the same) methods in equine were due to difficulties in getting horse meat from MAPA, which took more than 8 months to receive; the laboratory has internal turnaround times from sample receipt to reporting of between 3 days (for nitrofurans) and 15 days (for other analyses); CGAL audited the laboratory in May 2006 and a report of this was available; there is an SOP for validation which follows the main requirements of Commission Decision 2002/657/EC with the exception of between day reproducibility. It was noted that the range of acceptable recoveries (75-125%) was fixed, regardless of the target range of the analyte. This method had been followed for each of the methods studied by the mission team. Consequently CCα and CCβ determinations were based on a single day s data. Validation data for each of the validated methods were available in electronic format only; the ractopamine method had been validated for poultry and swine. There is no current Codex Alimentarius or Mercado Comum do Sul (MERCOSUL) Maximum Residue Limit (MRL) for ractopamine a proposal made by the World Health Organisation/Food and Agriculture Organisation Joint Expert Committee on Food Additives (JECFA) has been for 10 µg/kg in swine muscle. However, the laboratory had used an MRL figure of 40 µg/kg for swine for the calculation of CCα and CCβ for poultry. These were therefore 48 and 58 µg/kg respectively for poultry muscle. This approach is not acceptable for a substance which is not authorised for use in that species as CCα and CCβ should be as low 14

as possible. The current decision limit for poultry may be too high to effectively detect potential abuse of ractopamine in this species; in the field of VMP residues, laboratory staff have been trained in European national and Community reference laboratories. The laboratory has regularly participated in international proficiency tests including nitrofurans in poultry muscle (in 2004), tylosin and chloramphenicol in honey (in 2006) with acceptable results. The proficiency testing system currently being put in place by ANVISA will also be used by the laboratory. 5.4.2.4. Private sector laboratory B in Rio de Janeiro Under the NRCP, this laboratory is responsible for analysing for anabolic compounds in bovine urine (Group A1 (stilbenes), A3 (trenbolone), and A4 (zeranol) all analysed using one method) and for Group A6 (nitrofurans in poultry). Both methods have been approved by CGAL and notification of same was published in the Official Gazette in 2004 and 2006 respectively. Under the complementary programme the laboratory is responsible for: Group B1 (florfenicol), B2d (sedatives) and B3d (mycotoxins) in cattle; Group A1 (stilbenes), A3 (trenbolone and boldenone), A4 (zeranol), A5 (clenbuterol and salbutamol), B2d (sedatives) and B3d (mycotoxins) in equine; Group B1 (florfenicol and erythromycin) and B3e (malachite green) in aquaculture fin fish and; Group B1 (spiramycin) and B2d (sedatives levopromazine) in swine. The mission team noted that: the laboratory has extensive experience in the analysis of anabolic agents in doping control (human samples) and is an accredited International Olympic Committee laboratory. It has been accredited to ISO 17025 by INMETRO since 2002. With the exception of the methods for the anabolic agents, the methods included in the complementary programme are neither included in the scope of accreditation, nor are they approved by CGAL as the validation according to Commission Decision 2002/657/EC is still ongoing for some methods and some methods are not yet developed. Samples in the complementary plan for analysis other than for anabolics and malachite green have not yet been sent to the laboratory; in accordance with the requirements of ISO 17025 a comprehensive quality system was in place. A validation SOP was in place and this had been followed for validating the multi-residue anabolic method. Participation in proficiency tests is regular, though this has focussed on the main areas of work for the laboratory doping control in humans which is required for accreditation by the World Anti Doping Agency. To date no proficiency tests for residues of veterinary drugs in food have been carried out; the laboratory is equipped with a large number of state of the art (tandem) mass spectrometers (LC and GC). A batch of 13 LC and GC-MS-MS machines has just been purchased which, after being used for doping control in the forthcoming Pan American Games in July 2007, will be free to be employed on other projects including residues controls in food of animal origin. Thirteen personnel are currently dedicated to the residues testing programme for MAPA; the GC-MS method for anabolic agents was examined. The method which was validated to Commission Decision 2002/657/EC with CCα values of < 2 µg/kg for 14 compounds out of the 26 different anabolic steroids, beta agonists and stilbenes, though MAPA has only requested results for 5 compounds in total. 15