Polycyclic Aromatic Hydrocarbons in Food 1. Scientific Opinion of the Panel on Contaminants in the Food Chain. Adopted on 9 June 2008

Transcription

1 The EFSA Journal (2008) 724, Polycyclic Aromatic Hydrocarbons in Food 1 Scientific Opinion of the Panel on Contaminants in the Food Chain (Question N EFSA-Q ) Adopted on 9 June 2008 SCIENTIFIC PANEL MEMBERS Jan Alexander, Diane Benford, Andrew Cockburn, Jean-Pierre Cravedi, Eugenia Dogliotti, Alessandro Di Domenico, María Luisa Fernández-Cruz, Johanna Fink-Gremmels, Peter Fürst, Corrado Galli, Philippe Grandjean, Jadwiga Gzyl, Gerhard Heinemeyer, Niklas Johansson, Antonio Mutti, Josef Schlatter, Rolaf van Leeuwen, Carlos Van Peteghem, Philippe Verger. SUMMARY Polycyclic aromatic hydrocarbons (PAHs) constitute a large class of organic compounds that are composed of two or more fused aromatic rings. They are primarily formed by incomplete combustion or pyrolysis of organic matter and during various industrial processes. PAHs generally occur in complex mixtures which may consist of hundreds of compounds. Humans are exposed to PAHs by various pathways. While for non-smokers the major route of exposure is consumption of food, for smokers the contribution from smoking may be significant. Food can be contaminated from environmental sources, industrial food processing and from certain home cooking practices. In the past decade PAHs were evaluated by the International Programme on Chemical Safety (IPCS), the Scientific Committee on Food (SCF) and by the Joint FAO/WHO Expert Committee on Food Additives (JECFA). SCF concluded that 15 PAHs, namely benz[a]anthracene, benzo[b]fluoranthene, benzo[j]fluoranthene, benzo[k]fluoranthene, benzo[ghi]perylene, benzo[a]pyrene, chrysene, cyclopenta[cd]pyrene, dibenz[a,h]anthracene, dibenzo[a,e]pyrene, 1 For citation purposes: Scientific Opinion of the Panel on Contaminants in the Food Chain on a request from the European Commission on Polycyclic Aromatic Hydrocarbons in Food. The EFSA Journal (2008) 724, European Food Safety Authority, 2008

2 dibenzo[a,h]pyrene, dibenzo[a,i]pyrene, dibenzo[a,l]pyrene, indeno[1,2,3-cd]pyrene and 5- methylchrysene show clear evidence of mutagenicity/genotoxicity in somatic cells in experimental animals in vivo and with the exception of benzo[ghi]perylene have also shown clear carcinogenic effects in various types of bioassays in experimental animals. Thus, SCF reasoned that these compounds may be regarded as potentially genotoxic and carcinogenic to humans and therefore represent a priority group in the assessment of the risk of long-term adverse health effects following dietary intake of PAHs. SCF suggested to use benzo[a]pyrene as a marker of occurrence and effect of the carcinogenic PAHs in food, based on examinations of PAH profiles in food and on evaluation of a carcinogenicity study of two coal tar mixtures in mice. Using the assessments of IPCS and SCF as starting points and taking into account newer studies, the JECFA re-evaluated PAHs in Overall, the JECFA concluded that 13 PAHs are clearly genotoxic and carcinogenic. Except benzo[ghi]perylene and cyclopenta[cd]pyrene the compounds were the same as those stated by SCF. The JECFA also concluded that benzo[a]pyrene could be used as a marker of exposure to, and effect of, the 13 genotoxic and carcinogenic PAHs. In addition, the JECFA recommended to include benzo[c]fluorene as a further compound into future analyses as data on its occurrence in food are still scarce but rat studies indicate that benzo[c]fluorene may contribute to the formation of lung tumours after oral exposure to coal tar. Following a recommendation on the further investigation into levels of PAHs in certain foods (2005/108/EC) 2, eighteen Member States submitted almost 10,000 results for PAH levels in different food commodities. An evaluation of these data performed by EFSA in June 2007 and updated in June 2008 demonstrated that benzo[a]pyrene could be detected in about 50% of the samples. However, in about 30% of all the samples other carcinogenic and genotoxic PAHs were detected despite testing negative for benzo[a]pyrene. Of the individual PAHs, chrysene was most commonly found in the samples negative for benzo[a]pyrene with the highest level of 242 µg/kg. In view of these findings, the Commission requested a full review of the 2002 SCF opinion on PAHs. The EFSA Panel on Contaminants in the Food Chain (CONTAM Panel) reviewed the available data on occurrence and toxicity of PAHs. As no new toxicological data could be identified that would lead to the inclusion of further compounds into the list of priority PAHs, the CONTAM Panel decided to cover the 15 PAHs identified by SCF in 2002 and additionally benzo[c]fluorene as suggested by the JECFA in 2005 in the present opinion. Special attention was paid to those eight carcinogenic and genotoxic PAHs that were measured in the coal tar mixtures used in the carcinogenicity studies, which provided the basis of the SCF and JECFA risk assessments. 2 OJ L 34, , p.43 The EFSA Journal (2008) 724, 2-114

3 The CONTAM Panel explored whether a toxic equivalency factor (TEF) approach in the risk characterisation of the PAH mixtures in food could be applied and concluded that the TEF approach is not scientifically valid because of the lack of data from oral carcinogenicity studies on individual PAHs, their different modes of action and the evidence of poor predictivity of the carcinogenic potency of PAH mixtures based on the currently proposed TEF values. Therefore the CONTAM Panel concluded that the risk characterisation should be based upon the PAHs for which oral carcinogenicity data were available, i.e. for benzo[a]pyrene and the other PAHs that were measured in the two coal tar mixtures used in the carcinogenicity studies of Culp et al. (1998): benz[a]anthracene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[ghi]perylene, chrysene, dibenz[a,h]anthracene and indeno[1,2,3-cd]pyrene. The CONTAM Panel concluded that these eight PAHs (PAH8), either individually or in a combination, are currently the only possible indicators of the carcinogenic potency of PAHs in food. In total, results from 9714 PAH analyses in 33 food categories/subcategories were evaluated. As in about 30% of the samples analysed for all 15 priority PAHs as recommended by SCF other carcinogenic and genotoxic PAHs were detected despite testing negative for benzo[a]pyrene, individual compounds were grouped and summed in order to check whether their sums would better reflect the occurrence of carcinogenic and genotoxic PAHs in different food categories. The selection of the individual PAHs was based on the frequency of their results above the limit of detection (LOD). Besides the sum of the above mentioned eight PAHs (PAH8), the sum of benzo[a]pyrene, chrysene, benz[a]anthracene and benzo[b]fluoranthene (PAH4) as well as the sum of benzo[a]pyrene and chrysene (PAH2) were calculated. The correlation between PAH2 and PAH4 or PAH8 was 0.92 and between PAH4 and PAH8 was Of samples negative for PAH2, 26% and 18% identified concentrations above the LOD for at least one other PAH for samples tested for all PAH15 or all PAH8, respectively. The frequency varied between 2% and 9% for the individual PAHs or PAH combinations. Of samples negative for PAH4, 14% and 6% identified concentrations above the LOD for at least one other PAH for samples tested for all 15 PAHs or all PAH8, respectively. The frequency varied between 1% and 6% for the individual PAHs or PAH combinations. Overall, the Panel concluded that PAH4 and PAH8 were better indicators of the occurrence of PAHs than PAH2. For different food categories and subcategories, the data on PAH8, PAH4 and PAH2 were then used for the exposure calculation as well as the estimation of margins of exposure (MOEs) based on the bench mark dose lower confidence limit for a 10% increase in the number of tumour bearing animals compared to control animals (BMDL 10 ). The median dietary exposure across European countries was calculated both for mean and high dietary consumers and varied between 235 ng/day (3.9 ng/kg b.w. per day) and 389 ng/day (6.5 ng/kg b.w. per day) respectively for benzo[a]pyrene alone, 641 ng/day (10.7 ng/kg b.w. per day) and 1077 ng/day (18.0 ng/kg b.w. per day) respectively for PAH2, 1168 ng/day (19.5 ng/kg b.w. The EFSA Journal (2008) 724, 3-114

4 per day) and 2068 ng/day (34.5 ng/kg b.w. per day) respectively for PAH4 and 1729 ng/day (28.8 ng/kg b.w. per day) and 3078 ng/day (51.3 ng/kg b.w. per day) respectively for PAH8. The two highest contributors to the dietary exposure were cereals and cereal products, and sea food and sea food products. The CONTAM Panel used a MOE approach based on dietary exposure for average and high level consumers to benzo[a]pyrene, PAH2, PAH4 and PAH8, respectively and their corresponding BMDL 10 values derived from the two coal tar mixtures that were used in the carcinogenicity studies of Culp et al. (1998). The resulting MOEs for average consumers were 17,900 for benzo[a]pyrene, 15,900 for PAH2, 17,500 for PAH4 and 17,000 for PAH8. For high level consumers, the respective MOEs were 10,800, 9,500, 9,900 and 9,600. These MOEs indicate a low concern for consumer health at the average estimated dietary exposures. This applies to the full range of estimates of average exposures across EU Member States ( ng/kg b.w. per day, MOEs: 16,300-22,600 for benzo[a]pyrene alone and ng/kg b.w. per day, MOEs: 13,800-20,800 for PAH8). However, for high level consumers the MOEs are close to or less than 10,000, which as proposed by the EFSA Scientific Committee indicates a potential concern for consumer health and a possible need for risk management action. Comparison of the MOEs calculated for benzo[a]pyrene, PAH2, PAH4 and PAH8, indicates that PAH2, PAH4 and PAH8 can be used as alternatives to benzo[a]pyrene alone as markers of the carcinogenicity of the genotoxic and carcinogenic PAHs, and would be equally effective. The CONTAM Panel concluded that benzo[a]pyrene is not a suitable indicator for the occurrence of PAHs in food. Based on the currently available data relating to occurrence and toxicity, the CONTAM Panel concluded that PAH4 and PAH8 are the most suitable indicators of PAHs in food, with PAH8 not providing much added value compared to PAH4. Keywords: Polycyclic aromatic hydrocarbons (PAHs), food, occurrence, indicators, exposure, risk assessment, benchmark dose lower confidence limit (BMDL), margin of exposure (MOE), toxic equivalency factor (TEF) The EFSA Journal (2008) 724, 4-114

5 TABLE OF CONTENTS SCIENTIFIC PANEL MEMBERS... 1 SUMMARY... 1 BACKGROUND AS PROVIDED BY THE EUROPEAN COMMISSION... 7 TERMS OF REFERENCE AS PROVIDED BY THE EUROPEAN COMMISSION... 9 ACKNOWLEDGEMENT... 9 ASSESSMENT Introduction Legislation Sampling and methods of analysis Sampling Methods of analysis Sources and environmental fate Formation and production Environmental fate Sources of food contamination Occurrence and patterns of PAHs in food Factors influencing the levels of PAHs in food Commercial processing techniques Home cooking and other small scale cooking practices Suitable indicators for occurrence of total PAHs Food consumption Human exposure assessment Dietary intakes Contributions of different food groups to PAH exposure Mean dietary exposure to PAHs Dietary exposure to PAHs for high consumers Sensitivity analysis of dietary exposure Estimates of non-dietary exposure to PAHs Ambient air Occupational exposure Smoking Hazard identification and characterisation Summary of key data Toxicokinetics Absorption Distribution Metabolism Excretion Biomarkers of exposure Biomarkers of effect Toxicological studies Acute and short-term toxicity Reproductive toxicity Immunotoxicity The EFSA Journal (2008) 724, 5-114

6 Carcinogenicity Genotoxicity Observations in humans Dose-response modelling Benchmark dose modelling Benchmark dose calculations for benzo[a]pyrene Benchmark dose calculations for PAH Benchmark dose calculations for PAH Benchmark dose calculations for PAH Use of a TEF concept in the risk assessment of mixtures of carcinogenic PAHs Risk characterisation Suitable indicators for occurrence and toxicity of genotoxic and carcinogenic PAHs Uncertainty analysis Assessment objectives Exposure scenario Exposure model Model input (parameters) Other uncertainties CONCLUSIONS RECOMMENDATIONS (INCL. KNOWLEDGE/DATA GAPS) REFERENCES LIST OF ABBREVIATIONS AND ACRONYMS The EFSA Journal (2008) 724, 6-114

7 BACKGROUND AS PROVIDED BY THE EUROPEAN COMMISSION Polycyclic aromatic hydrocarbons (PAHs) form a class of diverse organic compounds, each of them containing two or more aromatic rings. Hundreds of different such compounds may be formed and released during a variety of combustion and pyrolysis processes and thus the natural and anthropogenic sources of PAHs in the environment are numerous. PAH compounds are emitted from the processing of coal, crude oil, petroleum, and natural gas, from production of aluminium, iron and steel, from heating in power plants and homes (oil, gas, charcoal-fired stoves, wood stoves), burning of refuse, wood fires, and from motor vehicle exhausts. Humans can be exposed to PAHs through different routes. For the general population, the major routes of exposure are from food and inhaled air, while in smokers, the contributions from smoking and food may be of a similar magnitude. Food can be contaminated by environmental PAHs that are present in air, soil or water, by industrial food processing methods (e.g. heating, drying and smoking processes) and by home food preparation (e.g. grilling and roasting processes). The Scientific Committee on Food (SCF) reviewed the presence and toxicity of PAHs in food and issued an opinion on 4 December SCF concluded that a number of PAHs are genotoxic carcinogens and recommended that exposure to PAHs should be as low as reasonably achievable. Fifteen substances were identified as a priority due to their potential genotoxicity and/or carcinogenicity in humans. SCF also concluded that benzo[a]pyrene may be used as a marker of occurrence and effect of the carcinogenic PAHs in food, but stressed that data collection on the whole PAH profile should continue in order to be able to evaluate the contamination of food commodities and any future change in the PAH profile. The Joint FAO/WHO Expert Committee on Food Additives (JECFA) performed a risk assessment on PAHs in 2005 in which it estimated Margins of Exposure (MOE) for PAHs. Based on these Margins of Exposure the JECFA concluded that PAHs were of low concern for human health. The JECFA also identified an additional substance, benzo[c]fluorene, as a priority substance. In the framework of Council Directive 93/5/EEC, a specific SCOOP task on data collection for PAH has been performed in On the basis of SCF opinion and the SCOOP data, the Commission introduced maximum levels for benzo[a]pyrene for some commodities that are 3 Opinion of the Scientific Committee on Food on the risks to human health of polycyclic aromatic hydrocarbons in food expressed on 4 December 2002, available at: 4 Report on SCOOP task "Collection of occurrence data on polycyclic aromatic hydrocarbons in food, October 2004, available at: 12_final_report_pah_en.pdf. The SCOOP task was carried out in the framework of scientific cooperation with Member States under Council Directive 93/5/EEC, OJ L 52, , p The EFSA Journal (2008) 724, 7-114

8 significant for human exposure and/or in which high PAH levels were found (e.g. oils and fats, smoked meats and smoked meat products, smoked fish and smoked fish products, muscle meat of fish other than smoked, crustaceans, cephalopods, bivalve molluscs and infant foods). The current Regulation setting maximum levels for PAHs in foodstuffs is Commission Regulation (EC) No. 1881/ For some foodstuffs high levels were found in the SCOOP report, but data were inconclusive as to the levels that were reasonably achievable (e.g. dried fruits, food supplements and cocoa butter). The Commission therefore asked Member States to monitor PAHs (and in particular the 15 priority substances identified by SCF) in Commission Recommendation 2005/108/EC 6. EFSA collected the data submitted in the framework of this monitoring recommendation and reported the findings in the recent Report "Findings of the EFSA Data collection on Polycyclic Aromatic Hydrocarbons in Food" of 29 June The EFSA report stated that the conclusion made by SCF that benzo[a]pyrene is a good indicator for PAH occurrence could not be demonstrated by the recent monitoring data from the Member States. Moreover, in October 2005 EFSA adopted the Margin of Exposure (MOE) approach 7 for substances which are both genotoxic and carcinogenic. In order to reflect these changes, it is appropriate to request a full review of the 2002 SCF opinion. Within this general review, a special focus should be given to the question whether or not benzo[a]pyrene can be maintained as a marker for both occurrence and toxicity of the most relevant PAHs. The relevance of the 15 priority PAHs as identified by SCF and the additional PAH identified by the JECFA should be confirmed in the light of any new scientific data that may have become available since the SCF opinion of OJ L 364, , p. 5 6 Commission Recommendation of 4 February 2005 on the further investigation into the levels of polycyclic aromatic hydrocarbons in certain foods, OJ L 34, , p The EFSA Journal (2005) 282, 1-31 The EFSA Journal (2008) 724, 8-114

9 TERMS OF REFERENCE AS PROVIDED BY THE EUROPEAN COMMISSION In accordance with Article 29 (1) (a) of Regulation (EC) No 178/2002 the European Commission asks the European Food Safety Authority to review the scientific opinion on PAH given by the Scientific Committee on Foods (SCF) of December The updated opinion should take into account the new occurrence data that were collected by EFSA and reported in the "Report on Findings of the EFSA Data collection on Polycyclic Aromatic Hydrocarbons in Food" of 29 June 2007 and should contain an updated exposure assessment on basis of these data, any other relevant scientific information that may have become available since the SCF opinion in 2002, including any new toxicological studies, the margin of exposure approach (MOE) as adopted by the EFSA Scientific Committee in the Opinion related to substances which are both genotoxic and carcinogenic. In the updated risk assessment the MOE approach should be used, if appropriate. Within the general review, the following more specific questions should also be addressed: Can benzo[a]pyrene still be considered as a suitable marker for both occurrence and carcinogenic effects for the most relevant PAHs? In the event that benzo[a]pyrene can not be maintained as a marker, could other suitable markers or concepts (e.g. TEF concept) be recommended for the occurrence as well as toxicity of the most relevant PAHs to be applied in European monitoring and official controls in order to protect human health? Which are the food commodities in the different Member States that contribute most to the exposure to the most relevant PAHs and which specific groups of the population are most exposed? ACKNOWLEDGEMENT EFSA wishes to thank the working group members Diane Benford, Jean-Pierre Cravedi, Peter Fürst, Niklas Johansson, John Christian Larsen, Dieter Schrenk, Rolaf van Leeuwen and Philippe Verger. The EFSA Journal (2008) 724, 9-114

10 ASSESSMENT 1. Introduction Polycyclic aromatic hydrocarbons (PAHs) constitute a large class of organic compounds that are composed of two or more fused aromatic rings. They solely consist of carbon and hydrogen and do not contain heteroatoms. PAHs are primarily formed by incomplete combustion or pyrolysis of organic matter and during various industrial processes. Consequently, the natural and anthropogenic sources in the environment are numerous. PAHs generally occur in complex mixtures which may consist of hundreds of compounds. The composition of these mixtures varies with the generating process. Humans are exposed to PAHs by various pathways. While for non-smokers the major route of exposure is consumption of food, for smokers the contribution from smoking may be significant. Food can be contaminated from environmental sources, industrial food processing and from certain home food preparation. In the past decade PAHs were evaluated by the International Programme on Chemical Safety (IPCS) (WHO/IPCS, 1998), the Scientific Committee on Food (SCF) (EC, 2002) and by the Joint FAO/WHO Expert Committee on Food Additives (JECFA) (FAO/WHO, 2005). SCF concluded that 15 out of the 33 PAHs which were considered in their assessment, namely benz[a]anthracene, benzo[b]fluoranthene, benzo[j]fluoranthene, benzo[k]fluoranthene, benzo[ghi]perylene, benzo[a]pyrene, chrysene, cyclopenta[cd]pyrene, dibenz[a,h]anthracene, dibenzo[a,e]pyrene, dibenzo[a,h]pyrene, dibenzo[a,i]pyrene, dibenzo[a,l]pyrene, indeno[1,2,3- cd]pyrene and 5-methylchrysene show clear evidence of mutagenicity/genotoxicity in somatic cells in experimental animals in vivo and with the exception of benzo[ghi]perylene, have also shown clear carcinogenic effects in various types of bioassays in experimental animals. Furthermore, SCF concluded that these compounds may be regarded as potentially genotoxic and carcinogenic to humans and thus represent a priority group in the assessment of the risk of longterm adverse health effects following dietary intake of PAHs. SCF estimated a maximum daily intake of benzo[a]pyrene from food of approximately 6-8 ng/kg b.w. per day for a person weighing 70 kg. It suggested to use benzo[a]pyrene as a marker of occurrence and effect of the carcinogenic PAHs in food, based on examinations of PAH profiles in food and on evaluation of a carcinogenicity study of coal tar mixtures in mice (Culp et al., 1998). Based on this latter study, a conservative assumption would imply that the carcinogenic potency of total PAHs in foods would be 10 times of that contributed by benzo[a]pyrene alone. SCF finally stressed that though it considered benzo[a]pyrene as a marker of carcinogenic PAHs in food, chemical analyses should continue to collect data on other PAHs in order to be able to evaluate the contamination of food commodities and any future change in the PAH profile. The EFSA Journal (2008) 724,

11 Using the risk assessments of IPCS and SCF as starting points and taking into account newer studies, the JECFA re-evaluated PAHs in Overall, the JECFA concluded that 13 PAHs are clearly genotoxic and carcinogenic. Except benzo[ghi]perylene and cyclopenta[cd]pyrene the compounds are the same as those stated by SCF. The JECFA also decided to apply a surrogate approach to the evaluation, in which benzo[a]pyrene was used as a marker of exposure to, and effect of, the 13 genotoxic and carcinogenic PAHs. Comparisons of representative mean and high level intakes with the BMDL 10 8 equivalent to 100 µg of benzo[a]pyrene/kg body weight per day indicated margins of exposure (MOE) of 25,000 and 10,000, respectively. Based on these MOEs, the JECFA concluded that the estimated dietary intakes of PAHs were of low concern for human health. Like the SCF the JECFA also recommended that efforts should be made to collect data on concentrations in the major food groups for those PAHs which were clearly identified as being genotoxic and carcinogenic. In addition, the JECFA recommended to include benzo[c]fluorene as a further compound into the analysis as data on its occurrence in food are still scarce but levels of benzo[c]fluorene-derived adducts were much higher than those of benzo[a]pyrene-derived adducts in the lung of rats fed with coal tar, which might indicate that benzo[c]fluorene may contribute to the formation of lung tumours after oral exposure to coal tar. Following the suggestion of SCF the Commission issued a recommendation on the further investigation into levels of polycyclic aromatic hydrocarbons in certain foods (2005/108/EC) 9. These investigations should especially include those PAHs considered to be carcinogenic by SCF. It was also recommended that Member States should investigate the production and processing methods used for food oils, fats as well as smoking and drying foods. Meanwhile eighteen Member States have already submitted approximately 10,000 results for PAH levels in food. An evaluation of these data by EFSA, published in June 2007 and updated in June 2008 showed that samples belonging to those food categories that are covered by Commission Regulation (EC) No 1881/2006 exceeded the respective maximum levels for benzo[a]pyrene in 6.4% of cases. Certain PAHs, such as chrysene were found in some food samples that were tested negative for benzo[a]pyrene (chrysene levels up to 242 µg/kg). In these cases benzo[a]pyrene can not be an indicator of PAH contamination. In view of these findings as well as the MOE approach adopted by EFSA in 2005 (EFSA, 2005a) for substances which are both genotoxic and carcinogenic, the Commission requested a full review of the 2002 SCF opinion on PAHs (see terms of reference). The EFSA Panel on Contaminants in the Food Chain (CONTAM Panel) reviewed the available data on occurrence and especially toxicology of PAHs. As no new toxicological data could be identified that would lead to the inclusion of further compounds into the list of priority PAHs, the CONTAM Panel decided to cover the 15 PAHs identified by SCF in 2002 and additionally 8 BMDL 10 : 95% lower confidence limit of the benchmark dose for a 10% increase in tumor bearing animals 9 OJ L 34, , p.43 The EFSA Journal (2008) 724,

12 benzo[c]fluorene as suggested by the JECFA in 2005 in the present opinion. The 16 compounds are given in Table 1 along with their molecular weight, CAS number and structure. Special attention was paid to the eight carcinogenic and genotoxic PAHs that were measured in the coal tar mixtures used in the carcinogenicity studies of Culp et al. (1998), which provided the basis of the SCF and JECFA risk assessments. These eight PAHs are underlined in Table 1. Table 1: Polycyclic aromatic hydrocarbons considered in the present opinion. Compound Abbr. MW CAS Structure Benz[a]anthracene BaA Benzo[b]fluoranthene BbFA Benzo[j]fluoranthene BjFA Benzo[k]fluoranthene BkFA Benzo[ghi]perylene BghiP Benzo[a]pyrene BaP The EFSA Journal (2008) 724,

13 Chrysene CHR Cyclopenta[cd]pyrene CPP Dibenz[a,h]anthracene DBahA Dibenzo[a,e]pyrene DBaeP Dibenzo[a,h]pyrene DBahP Dibenzo[a,i]pyrene DBaiP The EFSA Journal (2008) 724,

14 Dibenzo[a,l]pyrene DBalP Indeno[1,2,3-cd]pyrene IP methylchrysene MCH CH 3 Benzo[c]fluorene BcFL Abbr.: Abbreviation, MW: Molecular weight, CAS: Chemical Abstract Service Number 2. Legislation In view of disparities caused by different maximum levels (ML) for PAHs in food in several Member States, the Commission set harmonized ML for benzo[a]pyrene for the first time in 2005 by Commission Regulation (EC) No 208/ amending Regulation (EC) No 466/ as regards PAHs. Benzo[a]pyrene was chosen because SCF concluded in its opinion of 4 December 2002 that this compound can be used as a marker for the occurrence and effects of carcinogenic PAHs in food. Moreover, the data on occurrence and relative proportions of other carcinogenic PAHs than benzo[a]pyrene in food were considered insufficient by the Commission for setting further ML. The current MLs are laid down in the Annex, Section 6 of Commission Regulation (EC) No 1881/2006 setting maximum levels for certain contaminants in foodstuffs 12. Maximum levels are especially set for foodstuffs containing fats and oils and foods where smoking and 10 OJ L 34, , p OJ L 77, , p OJ L 364, , p. 5 The EFSA Journal (2008) 724,

15 drying processes or environmental pollution might cause high levels of contamination. The lowest MLs are set for food for infants and young children. All MLs are expressed as µg/kg wet weight (Table 2) unless more specific footnote applies. Table 2: Maximum levels (MLs) for benzo[a]pyrene as laid down in Regulation (EC) No. 1881/2006. See the original Regulation for further definitions and explanations of individual food commodities. Foodstuff Oils and fats (excluding cocoa butter) intended for direct human consumption or use as an ingredient in foods ML (µg/kg wet weight) 2.0 Smoked meats and smoked meat products 5.0 Muscle meat of smoked fish and smoked fishery products, excluding bivalve molluscs. The maximum level applies to smoked crustaceans, excluding the brown meat of crab and excluding head and thorax meat of lobster and similar large crustaceans (Nephropidae and Palinuridae). 5.0 Muscle meat of fish, other than smoked fish 2.0 Crustaceans, cephalopods, other than smoked. The maximum level applies to crustaceans, excluding the brown meat of crab and excluding head and thorax meat of lobster and similar large crustaceans (Nephropidae and Palinuridae) 5.0 Bivalve molluscs 10.0 Processed cereal-based foods and baby foods for infants and young children* Infant formulae and follow-on formulae, including infant milk and follow-on milk* Dietary foods for special medical purposes intended specifically for infants* 1.0 * The maximum level applies to the product as sold. In the past 20 years the traditional smoking process of food has been replaced more and more by the application of smoke flavourings. Because smoke flavourings are produced from smoke which is subjected to fractionation and purification processes, their use is generally considered to be of less health concern than the traditional smoking process. In this context, already in 1988 the The EFSA Journal (2008) 724,

16 Council of the European Communities set a ML of 0.03 µg/kg for benzo[a]pyrene in foodstuffs as a result of the use of flavourings by Council Directive 88/388/EEC on the approximation of the laws of the Member States relating to flavourings for use in foodstuffs and to source materials for their production 13. This directive defines a smoke flavouring as a smoke extract used in traditional foodstuffs smoking processes. Directive 88/388/EEC also provides for the adoption of appropriate provisions concerning source materials used for the production of smoke flavourings and reaction conditions under which they are prepared. These provisions are laid down in Regulation (EC) No 2065/2003 of the European Parliament and of the Council on smoke flavourings used or intended for use in or on foods 14. This Regulation lays down a Community procedure for the: (a) evaluation and authorisation of primary smoke condensates and primary tar fractions for use as such in or on foods or in the production of derived smoke flavourings for use in or on foods, (b) establishment of a list of primary smoke condensates and primary tar fractions authorised to the exclusion of all others in the Community and their conditions of use in or on foods. To this end Regulation (EC) No 2065/2003 sets MLs of 10 µg/kg for benzo[a]pyrene and 20 µg/kg for benz[a]anthracene in water-based primary smoke condensates as well as fractions obtained after physical processing from water-insoluble high density tar phases. Requirements for several PAHs in drinking water are set by Council Directive 98/83/EC on the quality of water intended for human consumption 15. This Directive stipulates that Member States shall set limit values of 0.01 µg/l for benzo[a]pyrene and 0.1 µg/l for the sum of benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[ghi]perylene and indeno[1,2,3-cd]pyrene. Commission Directive 95/45/EC laying down specific purity criteria concerning colours for use in foodstuffs 16 and Commission Directive 96/77/EC laying down specific purity criteria on food additives other than colours and sweeteners 17 both provide maximum levels for PAHs as impurities in some food additives without referring to specific PAH. 13 OJ L 184, , p OJ L 309, , p OJ L 330, , p OJ L 226, , p OJ L 339, , p. 1 The EFSA Journal (2008) 724,

17 3. Sampling and methods of analysis 3.1 Sampling As sampling, sample preparation and analytical procedures play an important role for a reliable determination of PAHs in foodstuffs, harmonized general criteria were at first set in the European Union (EU) in 2005 by Commission Directive 2005/10/EC laying down the sampling methods and the methods of analysis for the official control of the levels of benzo[a]pyrene in foodstuffs 18 simultaneously with the first setting of MLs for this compound. The present provisions are set in Commission Regulation (EC) No. 333/2007 laying down the methods of sampling and analysis for the official control of the levels of lead, cadmium, mercury, inorganic tin, 3- monochloropropane-1,2-diol (3-MCPD) and benzo[a]pyrene in foodstuffs 19. Besides definitions and general provisions this Regulation stipulates detailed requirements for methods of sampling for the different food commodities. No requirements exist for the number of samples that have to be analysed. The requirements for sampling methods differ especially depending on type and weight of lot, in order to obtain samples that are representative for the respective lot. Each lot which is to be examined shall be sampled separately. Large lots shall be subdivided into defined sub lots which have to be sampled separately. In any case, as far as possible, incremental samples shall be taken at various places throughout the lot or sub lot. By combining the incremental samples an aggregate sample is formed which after homogenization represents the base material for enforcement, defence and reference purposes. In the course of sampling and preparation of the samples, precautions shall be taken to avoid any changes in composition of the sample. As a specific and detailed requirement it is laid down that the analyst shall ensure that samples do not become contaminated during sample preparation. Containers shall be rinsed with high purity acetone or hexane before use to minimise the risk of contamination. Wherever possible, apparatus and equipment coming into contact with the sample shall be made of inert materials such as aluminium, glass or polished stainless steel. Plastics such as polypropylene or polytetrafluoroethylene (PTFE) shall be avoided because the PAHs can adsorb onto these materials. Losses of PAHs may also occur if the sample is exposed to light and high temperatures during the sample collection and storing, or PAHs react with other matrix substances during a long-term storage. Exposure to tobacco smoke may increase the PAH levels in the sample (FAO/WHO, 2006). 18 OJ L 34, , p OJ L 88, , p. 29 The EFSA Journal (2008) 724,

18 3.2 Methods of analysis Commission Regulation (EC) No 333/2007 also contains strict requirements with which the methods of analysis have to comply in order to ensure that control laboratories use procedures with comparable levels of performance. The Regulation follows the criteria approach. This means that no prescribed fixed official methods have to be followed but laboratories can use each method of analysis, provided it can be demonstrated in a traceable manner that they strictly fulfil the analytical requirements laid down in the respective legislation. As a general requirement, methods for benzo[a]pyrene analysis used for food control purposes must comply with the provisions of points 1 and 2 of Annex III (characterisation of methods of analysis) to Regulation (EC) No 882/2004 of the European Parliament and of the Council of on official controls performed to ensure the verification of compliance with feed and food law, animal health and animal welfare rules 20. While Regulation (EC) No 882/2004 contains the general provisions, the specific requirements for the official control of benzo[a]pyrene in foodstuffs are laid down in Commission Regulation (EC) No 333/2007. Annex, Table 7 of the latter one sets performance criteria for applicability, limits of detection (LOD) and quantification (LOQ), precision, recovery and specificity. The methods used for the determination should be applicable to those food specified in Regulation (EC) No 1881/2006. Requirements for LOD and LOQ are given as less than 0.3 µg/kg and less than 0.9 µg/kg, respectively. The recovery should be %. Concerning precision, it is required that the HORRAT r 21and HORRAT R 22 values are less than 2. The requirement for specificity is given as free from matrix or spectral interferences, verification of positive detection. Finally, the Commission Regulation (EC) No 333/2007 sets requirements for the reporting of results and the assessment of compliance of the lot or sub lots. For this, the analytical result corrected for recovery shall be used for checking compliance. The analytical result must be reported as x ± U whereby x is the analytical result and U is the expanded measurement uncertainty, using a coverage factor of 2 which gives a level of confidence of approximately 95%. The lot or sublot is accepted if the analytical result of the laboratory sample does not exceed the respective maximum level as laid down in Regulation (EC) No 1881/2006 taking into account the expanded measurement uncertainty and correction of the result for recovery. 20 OJ L 191, , p HORRAT r: : The observed relative standard deviation calculated from results generated under repeatability conditions (RSD r ) divided by the RSD r value estimated from the Horwitz equation (using the assumption that the repeatability r = 0.66R (reproducibility). The Horwitz equation is a generalised precision equation which has been found to be independent of analyte and matrix but solely dependent on concentration for most routine methods of analysis. 22 HORRAT R : The observed relative standard deviation calculated from results generated under reproducibility conditions (RSD R ) divided by the RSD R value calculated from the Horwitz equation. The EFSA Journal (2008) 724,

19 The extraction method used for extracting PAHs from food samples greatly depends on the nature of the food matrix. Saponification followed by liquid-liquid extraction and extraction with organic solvent are the most often used methods for solid food samples (e.g. meat, fish and their products), while liquid-liquid extraction is often used for liquid samples (e.g. vegetable oils). For solid food matrices automated extraction techniques, such as pressurised liquid extraction (PLE) and supercritical fluid extraction (SFE), are also applied, but less frequently. Column chromatography, solid phase extraction (SPE) and gel permeation chromatography (GPC) are the main sample purification techniques used for isolating PAHs from interfering matrix substances. Recently especially the automatic GPC technique has been applied for cleaning up PAH sample extracts (Fontcuberta et al., 2006; Llobet et al., 2006; Fromberg et al., 2007; Reinik et al., 2007). Nowadays the two main analytical techniques for determining PAHs in foods are high performance liquid chromatography (HPLC) coupled to a fluorescence detector (FLD) and gas chromatography-mass spectrometry (GC-MS). Both of these methods are sufficiently sensitive for determining PAH concentrations usually found in foods. Earlier HPLC with an ultraviolet (UV) or a photo-diode array (PDA) detector and GC with a flame ionisation detector (FID) were also methods often applied but today they are not state of the art anymore due to their poorer selectivity and sensitivity. Programmed temperature vaporization (PTV) injection used in the GC-methods allows injection of high sample volumes and hence low detection limits are achieved for PAHs. On-line methods simplify the analytical procedure of the PAH analysis because several separate manual sample preparation steps can be omitted. In the on-line HPLCmethod the sample extract is injected into a donor-acceptor complex chromatographic column which is coupled on-line to an analytical column in the HPLC-FLD system (van der Wielen et al., 2006; ISO, 2007). The GC-MS methods have become popular methods for analysing PAHs in foods. This is due to the selectivity of the MS-detector, the use of mass spectrum data for reliable confirmation of PAHs, and the possibility to use isotope labelled PAHs as internal standards. Single quadrupole MS is the main detection technique applied but the use of tandem mass spectrometry is increasing because it gives more specific mass fragments (daughter ions) and hence improves specificity and sensitivity of the MS-methods (Varlet et al., 2007; Veyrand et al., 2007). LODs and recoveries for PAHs are compound, food matrix and method dependent. The LODs for PAHs typically range from 0.1 to 1 µg/kg. However, for some PAHs lower and higher LODs, even up to 20 µg/kg, are determined (EFSA, 2007). Due to very specific tandem mass spectrometry, LODs lower than 0.1 µg/kg can be obtained by GC-MS/MS methods (Varlet et al., 2007). The recoveries for PAHs in different food samples are mostly over 70% with both HPLC and GC methods but for some PAHs recoveries are lower (FAO/WHO, 2006). A number of PAH compounds, such as benzofluoranthene isomers, have been reported to be difficult to separate chromatographically (FAO/WHO, 2006; Rose et al., 2007). In the first interlaboratory test organised by the Community Reference Laboratory for PAHs (CRL PAH) on the The EFSA Journal (2008) 724,

20 determination of the 15 SCF priority PAHs and benzo[c]fluorene in solvent, the participating National Reference Laboratories (NRLs) reported difficulties to analyse cyclopenta[cd]pyrene and some of the dibenzopyrenes (dibenzo[a,h]pyrene and dibenzo[a,i]pyrene) (EC-DG-JRC- IRMM/CRL, 2007). The second inter-laboratory comparison test organised by the CRL PAH for the consortium of NRLs was in that respect much more promising. Nearly all NRLs were able to determine all 15 SCF priority PAHs and benzo[c]fluorene both in solvent solution and in spiked olive oil. The overall result of the measurements of the 15 SCF priority PAHs and benzo[c]fluorene in the edible oil was satisfactory. All together 345 analysis results for individual PAHs were reported to the CRL PAH. About 76% of them were rated with z-scores < ±1. The majority of the 29 unsatisfactory results were reported by 4 laboratories only. However, scrutinising their results revealed that mainly instrument calibration was responsible for the deviations, which underpins the importance of proper standard preparation (EC-DG-JRC- IRMM/CRL, 2008). Interferences coming from non-target PAHs such as triphenylene that may interfere with chrysene, should also be considered. For these reasons care must be taken when the results for PAHs are calculated. Isotope labelled PAHs (deuterium-labelled and 13 C-labelled PAHs) are commonly used as internal standards to ensure the reliability of the final results. The 13 C-labelled PAHs are preferred in most of the GC-MS methods due to the instability of the deuterated standards (Fromberg et al., 2007; Veyrand et al., 2007). Although a number of 13 C-labelled PAH standards are available on the market at the moment there is still a need for 13 C-labelled analogues for various PAHs such as benzo[j]fluoranthene, dibenzo[a,h]pyrene, dibenzo[a,l]pyrene cyclopenta[cd]pyrene, 5-methylchrysene and benzo[c]fluorene. The lack of the 13 C-labelled standards may partly explain the high variation of the results reported for some of the PAH compounds in foods (Rose et al., 2007). Some certified reference materials (CRMs) for PAHs are available on the market and they are used for method validation studies to assess the performance of the method. In addition, the performance of the analytical methods is tested in the proficiency tests organised for PAHs. However, only a limited number of CRMs containing only some of the PAH compounds are on the market today and the few proficiency tests, which are organised presently, often cover only a narrow range of PAHs in some food products 23. Only recently proficiency tests for the full range of 15 SCF priority PAHs and benzo[c]fluorene (olive oil) have been provided 23. The lack of CRMs and proficiency tests are clearly limitations when the method performance for the analytical methods for analysis of PAHs in foods is assessed and The EFSA Journal (2008) 724,

21 4. Sources and environmental fate Data on the sources, fate and degradation of PAHs are extensively covered in a review made by IPCS (WHO/IPCS, 1998). PAHs have low solubility in water but are readily soluble in organic solvents or organic acids. Thus, in aqueous environments, PAHs are generally found adsorbed on particulates and on humic matter, or dissolved in any oily contaminant that may be present in water, sediment and soil. The solubility of PAHs in water is inversely proportional to the number of rings the PAH molecule contains. PAHs are solids at room temperature. Since PAHs tend to have low vapour pressure, they are usually adsorbed on particulate matter in the atmosphere. The vapour pressure of PAH is inversely proportional to the number of rings contained and thus almost all five-ring PAH compounds are particulate bound, while three-ring PAHs are also present as vapour in the atmosphere. 4.1 Formation and production There is no known use for the carcinogenic and genotoxic PAHs considered in this opinion except as research chemicals. A plethora of different PAHs may be formed and released during a variety of combustion and pyrolysis processes. Thus the natural and anthropogenic sources of PAHs in the environment are numerous. So far about 500 PAHs have been detected in ambient air. The emission of PAHs during industrial production and processing in developed countries are not thought to be important in comparison with the release of PAHs from incomplete combustion processes, since closed systems and recycling processes are usually used (WHO/IPCS, 1998). The primary natural sources of airborne PAHs are forest fires and volcanoes. The most important stationary anthropogenic sources include residential burning of wood, oil, gas and charcoal as well as industrial power generation, incineration, production of aluminium, iron and steel, petroleum catalytic cracking and production of asphalt, coal tar and coke (EC, 2002). Stationary sources account for approximately 80% of total annual PAH emissions. The most important mobile sources are vehicular exhausts from gasoline and diesel-powered engines (ATSDR, 1995). In combustion processes the formation of PAHs is reduced when combustion is more thoroughly performed but this will increase the formation of nitrogen oxides. Limited data are available on emission factors of PAHs, and the available data are often reported in different ways which means that comparison of data for verification purposes is difficult. The reasons for this are usually: many of the reported emissions of PAHs only give a figure for total PAHs, without indicating which PAH compounds are included in the total PAHs, The EFSA Journal (2008) 724,

22 where emissions of individual PAHs are given, there is a lack of consistency between reports on which PAHs are included in the measurements taken, most of the reported emissions of individual PAH only give data for one or two compounds (usually including benzo[a]pyrene). Different main sources exhibit different PAH profiles regarding their PAH emissions. This is exemplified in Tables 3 and 4 where the estimations for ratios of benzo[b]fluoranthene, benzo[k]fluoranthene and indeno[1,2,3-cd]pyrene in relation to benzo[a]pyrene are given. Table 3: PAH profiles for main stationary sources, estimated in a ratio to benzo[a]pyrene. PAH Coal combustion (industrial and domestic) Wood combustion (industrial and domestic) Natural fires/ agricultural biomass burning Anode baking Benzo[b]fluoranthene BbFA + BkFA Benzo[k]fluoranthene Benzo[a]pyrene Indeno[1,2,3-cd]pyrene Profiles in the above table taken from Table 4: PAH profiles for main mobile sources, estimated in a ratio to benzo[a]pyrene. PAH Passenger cars, conventional Passenger cars, closed loop catalyst Passenger cars, diesel, direct Passenger cars, diesel, indirect Heavy duty vehicles injection injection Benzo[b]fluoranthene Benzo[k]fluoranthene Benzo[a]pyrene Indeno[1,2,3-cd]pyrene Profiles in the above table taken from As seen in Tables 3 and 4 the PAH profiles vary considerably between the stationary sources investigated. Whereas benzo[a]pyrene is of major importance in coal combustion activities, it is less dominating in the anode baking process. On the other hand, the variation in PAH profiles for mobile sources is less pronounced with the exception of high relative emissions of benzo[b]fluoranthene and benzo[k]fluoranthene from heavy duty vehicles. The EFSA Journal (2008) 724,