Recent developments in the risk assessment of chemicals in food and their potential impact on the safety assessment of substances used in food contact materials

Tabs

Article
Panel on Food Contact Materials, Enzymes, Flavourings and Processing Aids
EFSA Journal
EFSA Journal 2016;14(1):4357 [28 pp.].
doi
10.2903/j.efsa.2016.4357
Panel members at the time of adoption
Claudia Bolognesi, Laurence Castle, Jean-Pierre Cravedi, Karl-Heinz Engel, Paul Fowler, Roland Franz, Konrad Grob, Rainer Gürtler, Trine Husøy, Sirpa Kärenlampi, Wim Mennes, Maria Rosaria Milana, André Penninks, Vittorio Silano, Andrew Smith, Maria de Fátima Tavares Poças, Christina Tlustos, Detlef Wölfle, Holger Zorn and Corina-Aurelia Zugravu.
Acknowledgements

The Panel wishes to thank the members of the Working Group on Food Contact Materials: Claudia Bolognesi, Laurence Castle, Jean-Pierre Cravedi, Roland Franz, Konrad Grob, Martine Kolf-Clauw, Eugenia Lampi, Maria Rosaria Milana, Maria de Fátima Tavares Poças, Kettil Svensson and Detlef Wölfe. The CEF Panel also wishes to thank the former members of the CEF Panel, Ricardo Crebelli, Jean Claude Lhuguenot, Catherine Leclercq and Iona Pratt, and EFSA staff, Eric Barthélémy and Dimitrios Spyropoulos, for the preparatory work on this scientific opinion.

Contact
Type
Opinion of the Scientific Committee/Scientific Panel
On request from
EFSA
Question Number
EFSA-Q-2011-00107
Adopted
2 December 2015
Published
28 January 2016
Affiliation
European Food Safety Authority (EFSA), Parma, Italy
Download Article (642.68 KB)
Abstract

This Opinion describes recent developments in the safety assessment of chemicals in food and explores their potential impact on EFSA evaluation of food contact materials (FCM). It is not intended to be a guidance document. The draft opinion was subject to a public consultation and this final Opinion takes into account the scientific comments received. The Opinion will provide the European Commission with the scientific basis for a discussion among risk managers on possible implications for risk management. One major area to revisit is the estimation of consumer exposure. Four food consumption categories could be set. They are approximately 9, 5, 3 and 1.2 times higher than the current SCF default scenario, i.e. 17 g/kg bw per day, and so using them would afford a higher level of protection, particularly for infants and toddlers. Special exposure scenarios might be used if consumption were lower. The amount of toxicity data needed should be related to the expected human exposure. The tiered approach of the SCF is updated. For substances used in FCM, genotoxicity testing is always required, even if their migration leads to a low exposure. Beyond this, three threshold levels of human exposure, namely 1.5, 30 and 80 μg/kg bw per day, are proposed as triggers for the requirement for additional toxicity data. Regarding the identification and evaluation of migrating substances, experience has shown that more focus is needed on the finished materials and articles. Considering the non-intentionally added substances (NIAS), such as impurities of the substance along with reaction and degradation products including oligomers, the same approach as is used for authorised substances could, in principle, be applied for their toxicological assessment, as the same degree of safety should be warranted for all migrating substances. However, non-testing methods could have increased importance for the assessment of genotoxicity of NIAS. 

Summary

In accordance with Regulation (EC) No 1935/2004 on materials and articles intended to come into contact with food, the European Food Safety Authority (EFSA) Panel on Food Contact Materials, Enzymes, Flavourings and Processing Aids (CEF Panel) evaluates the safety of certain substances prior to their authorisation for use in food contact materials (FCM) plastics. The current guidelines on this risk assessment process and the corresponding data requirements from applicants date back to the Scientific Committee on Food (SCF) guidelines from 2001. In the light of new developments in science and regulation, along with the experience gained since 2001 from the safety evaluation of hundreds of substances, it is appropriate to revisit the scientific underpinnings of the SCF guidelines published back in 2001 with a view to possibly updating them.

This Opinion is an outcome of a self-tasking activity by the CEF Panel. It describes the recent developments in the risk assessment of chemicals in food and explores their potential impact on EFSA evaluation of FCM substances. The draft of this opinion was published for a 3-month public consultation and was then modified in the light of the scientific comments received. EFSA technical report on that consultation process lists the comments received and provides a response to those comments, and it has been published as an accompanying document to this final, adopted Opinion. This Opinion will provide the European Commission (EC) with the scientific basis for a discussion among risk managers on possible implications for risk management. It is intended that, in turn, the EC will provide feedback for EFSA to prepare updated guidelines for data requirements for the safety assessment of a substance to be used in FCM.

One major area revisited is the estimation of consumer exposure. For most substances used in FCM, human exposure data were not readily available in the past. For this reason, the SCF used the assumption that a person may consume daily up to 1 kg of food in contact with 6 dm² of the relevant FCM. Now that EFSA’s Comprehensive European Food Consumption Database is available, based on the 95th percentile value for the highest European Union (EU) country and using the default water consumption figures set by the World Health Organization (WHO) for infants, four food group categories could be set. For category 1, FCM intended for contact with water and foodstuffs such as reconstituted infant milk formula, the age group with the highest consumption is ‘Infants’, with a consumption figure of 150 g/kg body weight (bw) per day. For category 2, in which contact with category 1 is excluded, but contact with milk, milk products and other non-alcoholic drinks is intended, then the age group with the highest consumption is ‘Toddlers’, with a value of 80 g/kg bw per day. For category 3, in which contact with food categories 1 and 2 are excluded but contact is with foods specifically intended for infants and toddlers, the age group with the highest consumption is ‘Toddlers’, with a value of 50 g/kg bw per day. For category 4, in which the FCM is intended for contact with foods other than those covered by categories 1, 2 and 3, the age group with the highest consumption is ‘Toddlers’, with a value of 20 g/kg bw per day. The food consumption values for these four categories are approximately 9, 5, 3 and 1.2 times higher than the current SCF default model, i.e. 17 g/kg bw per day (1 kg food consumed by an adult weighting 60 kg bw), and so using them would afford a higher level of protection, especially for infants and toddlers. Under certain conditions, special exposure scenarios might be used if consumption were lower.

Regarding the identification and evaluation of all substances that migrate, experience gained over the years has shown that more focus is needed on the finished materials and articles, including the manufacturing process used. Substances used in the manufacture of plastic materials or articles may contain impurities originating from their manufacturing. Moreover, during manufacturing and use, reaction and degradation products can be formed, of which oligomers can be the dominant class. These substances have become known as non-intentionally added substances (NIAS) and are referred to as such in Commission regulations. Whether their presence is intentional or not, it is necessary to evaluate the safety of all migrating substances and not just of the starting substances – for example the monomers or additives alone – and the guidelines should be updated to account more fully for this more comprehensive approach. This change towards the finished FCM and its use calls for an adjustment of the present system of listing substances in order to render transparent what has been evaluated.

In the case of testing for migration using food simulants, new rules are provided in Regulation (EU) No 10/2011. Similarly, the use of mathematical migration models has developed significantly in recent years, including proper validation for some of the most common types of plastics.

The amount of toxicity data needed should be related to the expected human exposure level, in accordance with the principle that the higher the exposure, the greater the amount of data required. Considering human exposure to determine the data needed may allow more efficient use of resources and contribute to reducing the use of experimental animals without loss in the safety assessment. In this Opinion, the tiered approach recommended by the SCF in 2001 is updated based on scientific progress. It focuses on the evaluation of substances used for the manufacture of plastic FCM, but it is, in principle, also applicable to those used in non-plastic FCM and those substances that are not specifically regulated but are assessed by the users.

For the safety assessment of substances used in FCM, genotoxicity testing is always required, even if exposure is low. Beyond this, three threshold levels of human exposure, namely 1.5, 30 and 80 μg/kg bw per day, are proposed as triggers for the requirement of additional toxicity data. The first level, 1.5 μg/kg bw per day, is intended to be a general threshold for the investigation of potential toxic effects other than genotoxicity. In case a substance can be classified in Cramer class I (the less toxic class, i.e. the substance has a simple chemical structure and can be anticipated to be metabolised to innocuous products, suggesting low oral toxicity), a second level of 30 µg/kg bw per day could be set instead of 1.5 μg/kg bw per day as the threshold for the investigation of repeated-dose toxicity. A third exposure threshold is proposed as a trigger for additional toxicity studies beyond the core set of general toxicity data. This threshold is defined as 80 μg/kg bw per day, consistent with previous SCF guidelines. The Panel considers that exposure above this level would approach that observed for food additives and that it would, therefore, be appropriate to require a corresponding toxicological data set.

The EFSA Scientific Committee recommendations on genotoxicity testing strategies call for two tests: (i) a bacterial reverse mutation assay; and (ii) an in vitro mammalian cell micronucleus test. This combination of tests fulfils the basic requirements to cover the three genetic endpoints with the minimum number of tests: the bacterial reverse mutation assay covers gene mutations and the in vitro micronucleus test covers both structural and numerical chromosome aberrations. The following tests in vivo would be suitable to follow-up for substances positive in the in vitro base set: (i) the in vivo micronucleus test; (ii) the in vivo Comet assay; and (iii) the transgenic rodent gene mutation assay.

Studies of subchronic toxicity generally provide sufficient information to establish the main toxicological profile of a substance, providing information on the target organs and tissues affected, on the nature and severity of the effects induced, and on the dose–response relationships. Chronic toxicity and carcinogenicity studies may reveal effects not evident in subchronic studies, or may confirm effects observed in subchronic studies, at the same or perhaps lower doses. Subchronic and chronic toxicity studies should allow the determination of the point of departure for safety assessment.

New testing strategies were recently developed to enhance the toxicological information from short-term and reproductive toxicity studies on potential effects on the endocrine, nervous and immune systems. Consequently, these improved study designs should be incorporated into the recommended toxicological test methods and study protocols.

Other updated test protocols are also described and discussed with respect to their applicability in any updating of the FCM guidelines, specifically protocols to test subchronic toxicity, prenatal developmental toxicity, chronic toxicity, toxicokinetics, endocrine disruption, neurotoxic potential, developmental effects on behaviour and neurotoxicity, and, finally, immunotoxic and immunomodulatory effects.

Read-across may also be used in the hazard characterisation of all migrating substances. The read-across approach contributes to the reduction in animal testing and resources.

FCM are one sector for potential use of nanotechnology and nanomaterials. The specific properties of nanomaterials may affect their toxicokinetic and toxicology profiles. The Panel recognised that the availability of data to cope with some of the six cases identified may depend on the specific properties of the nanomaterials and on the likely impact of the matrix in which they are dispersed.

Considering the NIAS, the same approach as that used for authorised substances should be applied for their toxicological assessment, as the same degree of safety should be warranted for all migrating substances. However, non-testing methods could have increased importance for the assessment of genotoxicity of NIAS.

Keywords
food contact materials, plastics, substances, safety assessment, migration, exposure, toxicological evaluation
Print on demand
Number of Pages
28