Flavouring Group Evaluation 82 Revision 2 (FGE.82Rev2): Consideration of epoxides evaluated by JECFA Authors: EFSA Panel on Food Additives and Flavourings (FAF), Laurence Castle, Monica Andreassen, Gabriele Aquilina, Maria Bastos, Polly Boon, Biagio Fallico, Rex FitzGerald, María José Frutos Fernández, Bettina Grasl‐Kraupp, Ursula Gundert‐Remy, Rainer Gürtler, Eric Houdeau, Marcin Kurek, Henriqueta Louro, Patricia Morales, Sabina Passamonti, Gisela Degen, Karl‐Heinz Engel, Maria Carfí, Ana Criado, Borana Dino, Carla Martino Published: 2026 ABSTRACT AbstractThe Panel on Food Additives and Flavourings (FAF) of the European Food Safety Authority was requested to consider evaluations of flavouring substances assessed since 2000 by the Joint FAO/WHO Expert Committee on Food Additives (JECFA), and to decide whether further evaluation is necessary, as laid down in Commission Regulation (EC) No. 1565/2000. The present consideration concerns a group of five epoxides evaluated by JECFA at the 65th meeting. This revision of FGE.82 is made due to new information on stereoisomeric composition, toxicity studies, uses and use levels for beta‐ionone epoxide [FL‐no: 07.170]. In addition, new data on uses and use levels for the substances [FL‐no: 16.015, 16.018, 16.040 and 16.043] have been provided and considered for the estimation of exposure (modified theoretical added maximum daily intake (mTAMDI) approach). The Panel considered that the information on the stereochemical composition of [FL‐no: 07.170] is sufficient. The Panel did not agree with the application of the Procedure for the epoxides as performed by JECFA, because in the absence of experimental data, the extent of detoxification of these epoxides remains uncertain. Therefore, [FL‐no: 07.170] was assessed via the B‐side of the Procedure. In the 90‐day toxicity study on [FL‐no: 07.170], substance‐related findings were observed in adrenal glands of male rats. Based on this observation, the Panel identified a no observed adverse effect level (NOAEL) of 39 mg/kg bw per day and concluded that for [FL‐no: 07.170] there is ‘no safety concern at estimated levels of intake as flavouring substance’, when based on the maximised survey‐derived daily intake (MSDI) approach. Based on new information on uses and use levels for [FL‐no: 07.170, 16.015, 16.018, 16.040 and 16.043], the mTAMDI exposure estimates are above the toxicological threshold of concern (TTC) for their structural class III. The Panel concluded that more reliable data on uses and use levels would be needed for these five substances. INTRODUCTION The present revision of FGE.82 (FGE.82Rev2) is made due to newly submitted data for beta‐ionone epoxide [FL‐no: 07.170]: information on stereoisomeric composition, toxicity studies, uses and use levels. In addition, new information on uses and use levels has been provided for the substances [FL‐no: 16.015, 16.018, 16.040 and 16.043] and considered for the estimation of exposure (modified Theoretical Added Maximum Daily Intake (mTAMDI) approach). Background and Terms of Reference as provided by the requestor Background The use of flavourings in food is regulated under Regulation (EC) No 1334/2008 of the European Parliament and of the Council of 16 December 2008 1 on flavourings and certain food ingredients with flavouring properties for use in and on foods. On the basis of Article 9(a) of this Regulation, an evaluation and approval are required for flavouring substances. The Union list of flavourings and source materials was established by Commission Implementing Regulation (EC) No 872/2012. 2 The list includes flavouring substances for which the scientific evaluation should be completed in accordance with Commission Regulation (EC) No 1565/2000. 3 In the scientific opinion on Flavouring group evaluation 82 revision 1 (FGE.82Rev1) adopted on 21 May 2014, the EFSA Panel on Food Contact Materials, Enzymes, Flavourings and processing Aids (CEF) concluded that the evaluation of beta‐ionone epoxide [FL‐no: 07.170] cannot be finalised because a NOAEL from a 90‐day study is not available and information on its stereoisomeric composition are missing. In March 2025, the European Flavour Association submitted to the Commission the missing data. Terms of Reference The European Commission requests the European Food Safety Authority (EFSA) to evaluate this new information and to proceed to the full evaluation of the flavouring substance [FL‐no: 07.170], in accordance with Commission Regulation (EC) No 1565/2000.3 Interpretation of the Terms of Reference In addition to assess the new data submitted for [FL‐no: 07.170], new information on uses and use levels provided for the substances [FL‐no: 16.015, 16.018, 16.040 and 16.043] (Documentation provided to EFSA No. 3) are considered in the present revision to estimate the exposure based on the mTAMDI approach. History of the evaluation and presentation of the substances in the present FGE FGE.82 At its 65th meeting, the Joint FAO/WHO Expert Committee on Food Additives (JECFA) evaluated a group of nine flavouring substances belonging to the class of epoxides (JECFA, 2006a, 2006b). One substance evaluated by JECFA, trans‐carvone‐5,6‐oxide (JECFA no. 1572), was not included in the EU Register 4 at the time of EFSA's assessment and was therefore not evaluated further by EFSA, while the other eight substances [FL‐no: 07.170, 16.015, 16.018, 16.043, 16.040, 16.044, 16.051, 16.071] were included in the EU Register and therefore considered by EFSA. Owing to the presence of an α, β‐unsaturated carbonyl moiety, which constitutes a structural alert for genotoxicity, three substances from this group, two α, β‐unsaturated ketones (β‐ionone epoxide [FL‐no: 07.170] and piperitenone oxide [FL‐no: 16.044]) and one α, β‐unsaturated aldehyde (4,5‐epoxydec‐2(trans)‐enal [FL‐no: 16.071]), were assessed separately for their genotoxic potential within dedicated EFSA genotoxicity flavouring group evaluations (FGE.210Rev1, FGE.213 and FGE.226Rev1, respectively). Piperitenone oxide [FL‐no: 16.044] was evaluated in FGE.213, where the CEF Panel requested additional genotoxicity data (EFSA, 2009a). Industry informed that [FL‐no: 16.044] was no longer of interest for use as a flavouring substance in Europe, and no further data were submitted (Documentation provided to EFSA No. 2). Therefore, the substance [FL‐no: 16.044] was not included in the Union List.2 4,5‐Epoxydec‐2(trans)‐enal [FL‐no: 16.071] was evaluated in FGE.226Rev1, where the CEF Panel concluded that [FL‐no: 16.071] does raise a safety concern with respect to genotoxicity and, therefore, it was not evaluated according to the Procedure (which is described in Appendix A) (EFSA CEF Panel, 2017). The remaining epoxide flavouring substances, i.e. ethyl methylphenylglycidate [FL‐no: 16.015], ethyl 3‐phenyl‐2,3‐epoxypropionate [FL‐no: 16.018], ethyl 2,3‐epoxy‐3‐methyl‐3‐p‐tolylpropionate [FL‐no: 16.040], β‐caryophyllene epoxide [FL‐no: 16.043] and epoxy oxophorone [FL‐no: 16.051], were evaluated by the EFSA Panel on Food Additives, Flavourings, Processing Aids and Materials in Contact with Food (AFC Panel) in FGE.82 (EFSA, 2009b). Among these five substances, adequate specifications, including complete purity criteria and identity tests, were available only for β‐caryophyllene epoxide [FL‐no: 16.043]. Information on stereoisomeric composition was lacking for [FL‐no: 16.015, 16.018, 16.040 and 16.051], and data on solubility in water and ethanol were missing for ethyl 3‐phenyl‐2,3‐epoxypropionate [FL‐no: 16.018]. The AFC Panel considered that epoxy oxophorone [FL‐no: 16.051] should not be evaluated through the Procedure because of concerns regarding genotoxicity and concluded that adequate genotoxicity data would be required. For ethyl methylphenylglycidate [FL‐no: 16.015] and ethyl 3‐phenyl‐2,3‐epoxypropionate [FL‐no: 16.018], the AFC Panel considered that the available in vitro and in vivo studies indicated a genotoxic potential. However, for ethyl methylphenylglycidate [FL‐no: 16.015], a carcinogenicity study allowed derivation of a no observed adverse effect level (NOAEL) of 0.1% in the diet (35 mg/kg bw per day for males and 60 mg/kg bw per day for females). The AFC Panel considered that the negative outcome of this carcinogenicity study overruled the positive genotoxicity findings for ethyl methylphenylglycidate [FL‐no: 16.015] and also for the structurally related substances ethyl 3‐phenyl‐2,3‐epoxypropionate [FL‐no: 16.018] and ethyl 2,3‐epoxy‐3‐methyl‐3‐p‐tolylpropionate [FL‐no: 16.040]. 5 For ethyl methylphenylglycidate [FL‐no: 16.015], ethyl 3‐phenyl‐2,3‐epoxypropionate [FL‐no: 16.018], ethyl 2,3‐epoxy‐3‐methyl‐3‐ptolylpropionate [FL‐no: 16.040] and beta‐caryophyllene epoxide [FL‐no: 16.043], the AFC Panel concluded that the available data do not preclude their evaluation through the Procedure. Accordingly, ethyl methylphenylglycidate [FL‐no: 16.015], ethyl 3‐phenyl‐2,3‐epoxypropionate [FL‐no: 16.018], ethyl 2,3‐epoxy‐3‐methyl‐3‐p‐tolylpropionate [FL‐no: 16.040] and β‐caryophyllene epoxide [FL‐no: 16.043] were evaluated through the Procedure. Contrary to JECFA, the AFC Panel considered that these substances could not be predicted to be metabolised to innocuous products and therefore these substances were evaluated via the B‐side of the Procedure. For β‐caryophyllene epoxide [FL‐no: 16.043], no toxicity studies were available from which a NOAEL could be identified (see Appendix A) for the substance or for structurally related substances and additional toxicity data were therefore required. For the four substances evaluated through the B‐side of the Procedure, the AFC Panel expressed reservations due to missing information on stereoisomerism for three substances [FL‐no: 16.015, 16.018, 16.040] and the need for additional toxicity data for one substance [FL‐no: 16.043]. In addition, for these substances [FL‐no: 16.015, 16.018, 16.040 and 16.043], use levels would be required to calculate the mTAMDIs in order to identify those flavouring substances that need a more refined exposure assessment and to finalise their evaluation. FGE.82Rev1 After the publication of FGE.82, industry informed that epoxy oxophorone [FL‐no: 16.051] is no longer of interest for use as flavouring substance in Europe (Documentation provided to EFSA No. 2), and therefore the data requested by the AFC Panel in FGE.82 were not provided. Accordingly, this substance [FL‐no: 16.051] was not included in the Union List.2 The α, β‐unsaturated epoxide, beta‐ionone epoxide [FL‐no: 07.170] was evaluated in FGE.210Rev1 (EFSA CEF Panel, 2014b), where the substance was considered not to be of concern with respect to genotoxicity. Beta‐ionone epoxide [FL‐no: 07.170] was therefore included in FGE.82Rev1 (EFSA CEF Panel, 2014a) to complete the evaluation through the Procedure. Therefore, in FGE.82Rev1, the CEF Panel considered five flavouring substances evaluated by JECFA [FL‐no: 07.170, 16.015, 16.018, 16.040 and 16.043]. New information on missing stereoisomeric composition for [FL‐no: 16.015, 16.018 and 16.040] was provided (Documentation provided to EFSA No. 6) and considered in FGE.82Rev1. The CEF Panel considered that adequate specifications including complete purity criteria and identity tests were available for four JECFA‐evaluated substances [FL‐no: 16.015, 16.018, 16.040 and 16.043], but for [FL‐no: 07.170] information on stereoisomeric composition was not adequate. Additional toxicity data (a 14‐day dose range‐finding study and a 90‐day dietary study) were provided for beta‐caryophyllene epoxide [FL‐no: 16.043]. Based on the NOAEL derived from the 90‐day study and the estimated exposure, the CEF Panel considered that the margin of safety 6 when used as a flavouring substance was adequate. The CEF Panel considered that four of the five substances [FL‐no: 16.015, 16.018, 16.040 and 16.043] evaluated via the B‐side of the Procedure scheme were of no safety concern at the estimated levels of intake, when based on the MSDI approach. The evaluation of beta‐ionone epoxide [FL‐no: 07.170] could not be finalised because a 90‐day study was not available. For [FL‐no: 07.170 and 16.043] data on use levels were provided and considered in FGE.82Rev1. For the other three substances [FL‐no: 16.015, 16.018 and 16.040] use levels would be required to calculate the mTAMDIs in order to identify those flavouring substances that need more refined exposure assessment and to finalise their evaluation. Table 1 provides a compilation of the three opinions on FGE.82. TABLE 1References to the opinions on FGE.82.FGEReferenceSubstances a FGE.82EFSA (2009b)5FGE.82Rev1EFSA CEF Panel (2014a)5FGE.82Rev2Current opinion5 a Substances considered in: FGE.82: [FL‐no: 16.015, 16.018, 16.040, 16.043 and 16.051]; FGE.82Rev1: [FL‐no: 07.170, 16.015, 16.018, 16.040 and 16.043]; FGE.82Rev2: [FL‐no: 07.170, 16.015, 16.018, 16.040 and 16.043]. References to the opinions on FGE.82. Substances considered in: FGE.82: [FL‐no: 16.015, 16.018, 16.040, 16.043 and 16.051]; FGE.82Rev1: [FL‐no: 07.170, 16.015, 16.018, 16.040 and 16.043]; FGE.82Rev2: [FL‐no: 07.170, 16.015, 16.018, 16.040 and 16.043]. A summary of the history of the evaluation of the substances in FGE.82 and revisions is presented in Figure 1. Summary of the history of evaluation of the substances in FGE.82 and revisions. DATA AND METHODOLOGIES Data In the present revision (FGE.82Rev2) the following data were considered: −a 14‐day dietary toxicity/palatability study, a 90‐day toxicity study in rats, information on stereoisomeric composition and updated information on uses and use levels for beta‐ionone epoxide [FL‐no: 07.170] (Documentation provided to EFSA No. 1, 7 and 8).−new information on uses and use levels for four substances [FL‐no: 16.015, 16.018, 16.040 and 16.043] has been provided (Documentation provided to EFSA No. 3). −a 14‐day dietary toxicity/palatability study, a 90‐day toxicity study in rats, information on stereoisomeric composition and updated information on uses and use levels for beta‐ionone epoxide [FL‐no: 07.170] (Documentation provided to EFSA No. 1, 7 and 8). a 14‐day dietary toxicity/palatability study, a 90‐day toxicity study in rats, information on stereoisomeric composition and updated information on uses and use levels for beta‐ionone epoxide [FL‐no: 07.170] (Documentation provided to EFSA No. 1, 7 and 8). −new information on uses and use levels for four substances [FL‐no: 16.015, 16.018, 16.040 and 16.043] has been provided (Documentation provided to EFSA No. 3). new information on uses and use levels for four substances [FL‐no: 16.015, 16.018, 16.040 and 16.043] has been provided (Documentation provided to EFSA No. 3). Methodologies This opinion was prepared following the principles described in the EFSA Guidance on transparency with regard to scientific aspects of risk assessment (EFSA, 2009c) and following the relevant existing Guidelines from the EFSA Scientific Committee. As requested in the EC mandate, the assessment strategy applied is in line with the evaluation programme of flavouring substances, as laid down in Commission Regulation (EC) No 1565/2000,3 which is based on the Opinion on a Programme for the Evaluation of Flavouring substances of the Scientific Committee on Food (SCF, 1999). The approach for safety evaluation of chemically defined flavouring substances as referred to in Commission Regulation (EC) No 1565/2000,3 named the ‘Procedure’, is described in Appendix A. The approach used for calculation of the intake of the flavouring substances is described in Appendix A (point a ‘Intake’) and in Appendix C (Section C.2 ‘mTAMDI calculation’). ASSESSMENT The present revision of FGE.82 (FGE.82Rev2) addresses the evaluation of β‐ionone epoxide [FL‐no: 07.170] for which a 14‐day dietary toxicity/palatability study, a 90‐day toxicity study, updated information on uses and use levels and data on stereoisomerism have been provided (Documentation provided to EFSA No. 1, 7 and 8). After the publication of FGE.82Rev1 in 2014, new information on uses and use levels has been provided for four substances [FL‐no: 16.015, 16.018, 16.040 and 16.043] (Documentation provided to EFSA No. 3). Accordingly, the corresponding mTAMDI estimates for these four substances are included in the present opinion. Therefore, FGE.82Rev2 includes five substances, of which four were already considered in FGE.82Rev1 as of no safety concern when based on the MSDI approach. For the sake of completeness, information on specifications, intake and evaluation status of all five substances, being already available in previous assessments, is presented in the respective Appendices B, C and E. Specifications JECFA status JECFA specifications are available for all five substances [FL‐no: 07.170, 16.015, 16.018, 16.040 and 16.043] (JECFA, 2005). All these five substances have one or more chiral centres. For β‐caryophyllene epoxide [FL‐no: 16.043], JECFA ‘revised the melting point to 55–63°C and the assay minimum to 95% (sum of isomers) from data on eight lots of commercial product. Specifications for the isomeric compositions were also established: 84%–89% 1R,4R,6R,10S (CAS No. 1139‐30‐6), 7%–9% 1R,4R,6S,10S (CAS No. 60594‐22‐1), 0.3–2% 1R,4S,6S,10S (CAS No. 103475‐43‐0) and 1%–2% humulene‐1,2‐epoxide (CAS No. 19888‐34‐7) (JECFA, 2020). EFSA considerations The chemical name, structure and structural class (Cramer et al., 1978) of the flavouring substance [FL‐no: 07.170] considered in FGE.82Rev2 are reported in Table 2. Specifications including complete purity criteria and identity are available for four substances [FL‐no: 16.015, 16.018, 16.040 and 16.043]. Information about the stereoisomerism for one substance [FL‐no: 07.170] has been provided. According to the information provided by industry, beta‐ionone epoxide [FL‐no: 07.170] is a racemic mixture of 50%–70% (E)‐ and 30%–50% (Z)‐isomers (Documentation provided to EFSA No. 1). Since [FL‐no: 07.170] has two chiral centres with two configurations each, it may occur as a mixture of four stereoisomers. Specifications for all five substances in FGE.82Rev2 are summarised in Appendix B, Table B.1, based on data reported in FGE.82Rev1, information considered by JECFA (JECFA, 2020) and the additional information provided for [FL‐no: 07.170] (Documentation provided to EFSA No. 1). The available specifications are considered adequate for all five substances. TABLE 2Chemical name, structure and structural class (Cramer et al., 1978) of the flavouring substance under evaluation in FGE.82Rev2. FL‐no JECFA no Structural formula b Chemical nameStructural class a 07.170 1571 beta‐Ionone epoxideIIIAbbreviations: FL‐no, FLAVIS number; JECFA, The Joint FAO/WHO Expert Committee on Food Additives; JECFA no, JECFA number. a Determined with OECD Toolbox (version 4.8 available at https://qsartoolbox.org/). b Only the (E) isomer is shown. Chemical name, structure and structural class (Cramer et al., 1978) of the flavouring substance under evaluation in FGE.82Rev2. FL‐no JECFA no 07.170 1571 Abbreviations: FL‐no, FLAVIS number; JECFA, The Joint FAO/WHO Expert Committee on Food Additives; JECFA no, JECFA number. Determined with OECD Toolbox (version 4.8 available at https://qsartoolbox.org/). Only the (E) isomer is shown. Estimation of intake JECFA status For all five substances evaluated through the JECFA Procedure, production volume data in the EU were available and allowed the calculation of MSDI values (JECFA, 2006a, 2006b). Since beta‐ionone epoxide [FL‐no: 07.170] was proposed for use as a flavouring substance, an anticipated volume of production was available to JECFA at that time. The conclusions of JECFA on the safety of [FL‐no: 07.170] were conditional pending the submission of actual poundage data (JECFA, 2006a, 2006b). The actual production volume for [FL‐no: 07.170] was subsequently submitted and considered by JECFA at the 69th meeting (JECFA, 2009). Based on this new information, JECFA confirmed the assessment of [FL‐no: 07.170] as of no safety concern (JECFA, 2009). EFSA considerations EU production volumes are available for all substances [FL‐no: 07.170, 16.015, 16.018, 16.040 and 16.043]. Information on uses and use levels has been provided for each of the five substances (Documentation provided to EFSA No. 1 and 3) and considered in the present revision, FGE.82Rev2 (see Table C.1, Appendix C). The MSDI figures and mTAMDI intake estimates for the flavouring substances in FGE.82Rev2 are shown in Table C.4, Appendix C. Natural occurrence Data on natural occurrence were reported by JECFA (2006a, 2006b) for [FL‐no: 07.170] (JECFA no. 1571) and [FL‐no: 16.043] (JECFA no. 1575) and were also provided by industry for [FL‐no: 07.170] (Documentation provided to EFSA No.1). This information is included in Appendix C. Biological and toxicological data ADME data JECFA status JECFA (2006a, 2006b) considered that epoxides are ‘readily and adequately’ detoxified in animals via two pathways, hydrolysis in the gastrointestinal tract or in other tissues resulting in diols, or via reaction of epoxides with glutathione, and concluded that ‘all the flavouring agents in this group can be predicted to be metabolised to innocuous products’. Therefore, the substances were evaluated via the A‐side of the Procedure. EFSA considerations The Panel considered that beta‐ionone epoxide as well as the other epoxides in this FGE can be metabolised to non‐toxic products. However, in the absence of experimental data, the extent to which these epoxides are detoxified remains uncertain. For this reason, the Panel cannot confirm that the flavouring substances in this group can be predicted to be completely metabolised to innocuous products. Therefore, these substances are evaluated through the Procedure via the B‐side (Appendix A) as in the previous opinions on FGE.82 (EFSA, 2009b; EFSA CEF Panel, 2014a). Genotoxicity studies JECFA status JECFA (2006a) assessed the genotoxicity of this group of epoxides (including [FL‐no: 07.170] based on genotoxicity studies available for [FL‐no: 16.015, 16.018 and 16.043]): ‘The genotoxic potential of glycidate and alicyclic epoxides was studied in several standard assays in bacteria and mammalian cells in vitro. In the Ames assay for reverse mutation, both caryophyllene oxide 7 and ethyl methylphenylglycidate 8 gave unequivocally negative results with and without metabolic activation in a series of standard S. Typhimurium tester strains. Ethyl 3‐phenylglycidate 9 gave positive results only in S. Typhimurium strains TA100 and TA98 and mainly in the presence of metabolic activation; however, a digest of ethyl 3‐phenylglycidate had no mutagenic potential in S. Typhimurium TA100 and TA98 with or without metabolic activation. Negative results were reported with cis‐methyl epoxycinnamate, a structurally related glycidic ester, in the standard battery of S. Typhimurium tester strains, but trans‐methyl epoxycinnamate was mutagenic in TA1535, TA1537, TA1538 and TA100, only in the absence of metabolic activation. Cyclopentane and cyclohexane oxide induced reverse mutation in S. Typhimurium strains TA1535 and TA100. Although cycloaliphatic epoxides of larger ring size were generally less mutagenic than the smaller‐ring epoxides, marked cytotoxicity was seen. Uniformly negative results were obtained in the SOS Chromotest in E. coli. Although clastogenic activity was reported in mammalian cell lines with most glycidate and alicyclic epoxides, increased sister chromatid exchange frequencies were observed with ethyl methylphenylglycidate only in the absence of metabolic activation. Likewise, chromosomal aberrations were reported in Chinese hamster ovary cells incubated with ethyl methylphenylglycidate or cyclohexane oxide, but the results with ethyl methylphenylglycidate were equivocal when S9 activation was incorporated into the assay. No unscheduled DNA synthesis was found with a series of epoxyalkanes and alicyclic epoxides. In vivo, ethyl methylphenylglycidate had slight potential to induce sex‐linked recessive lethal mutations in Drosophila, but ethyl 3‐phenylglycidate showed no activity. Ethyl 3‐phenylglycidate and ethyl methylphenylglycidate did not induce micronucleus formation in mice given single intraperitoneal doses. Epoxides are naturally occurring substances that are also added to food as flavouring agents. The principal epoxide used in this way is ethyl methylphenylglycidate. Studies on the metabolism of glycidate and alicyclic (including terpene) epoxides indicate that these compounds are readily and adequately detoxicated in animals via two pathways, GSH conjugation and hydrolysis in the gastrointestinal tract or other tissues, followed by glucuronic acid or sulphate conjugation in the liver. Although glycidate and alicyclic epoxides had some genotoxic potential in standard assays in vitro, the results of assays for genotoxicity in mammals in vivo were negative. Furthermore, a number of long‐term studies with dietary administration provided no evidence of carcinogenic potential. Several long‐term studies with repeated doses of ethyl methylphenylglycidate showed no carcinogenicity even at intake levels that were orders of magnitude higher than the intake of epoxides added as flavouring agents. The NOEL in the 2‐year bioassay of ethyl methylphenylglycidate was 35 mg/kg bw per day. This intake level is > 1100 times the daily per capita intake (‘eaters only’) of 0.031 mg/kg bw per day from use of ethyl methylphenylglycidate as a flavouring agent.' JECFA (2006a) concluded that: ‘The known pathways of metabolic detoxication, the lack of evidence of carcinogenicity in long‐term feeding studies and the lack of genotoxic potential in vivo indicate that it is unlikely that epoxides pose a significant genotoxic risk to humans under the conditions of their use as flavouring agents’. EFSA considerations The genotoxic potential of [FL‐no: 07.170], an α, β‐unsaturated epoxide, was assessed in FGE.210Rev1 (EFSA CEF Panel, 2014b) where the concern was ruled out. In FGE.210Rev1 (EFSA CEF Panel, 2014b), the CEF Panel assessed in vitro genotoxicity studies for [FL‐no: 07.170]: ‘beta‐Ionone epoxide did not induce any significant increase in bacterial mutation when evaluated in five different S. Typhimurium strains and an E. coli strain, either in the presence or absence of S9 metabolic activation in two independent studies. Beta‐Ionone epoxide also did not increase mutation frequencies when tested in a tk mutation assay using mouse lymphoma cells either in the presence or absence of S9 metabolic activation. No in vitro assay for chromosomal aberration is available, but the mouse lymphoma assay is a test that is able to detect the chemical potential to induce structural chromosomal aberrations. The lack of an in vitro micronucleus assay is not consistent with the current EFSA guideline (EFSA Scientific Committee, 2011), it is however consistent with the genotoxicity test strategy for substances belonging to subgroups of FGE.19 (EFSA, 2008) applicable at the time when the scientific opinion on FGE.210 was adopted (EFSA, 2009d). The Panel concluded that the data submitted for beta‐ionone epoxide are sufficient in the light of data available for structurally related substances’. (EFSA CEF Panel, 2014b). In FGE.210Rev1, the CEF Panel concluded that based on the experimental data on the representative substances alpha‐ionone [FL‐no: 07.007] and beta‐ionone epoxide [FL‐no: 07.170], the concern with respect to genotoxicity could be ruled out. Therefore, for the flavouring substance [FL‐no: 07.170] the safety evaluation through the Procedure can be performed. Toxicity studies JECFA status Toxicological data for the flavouring substance [FL‐no: 07.170] were not considered by JECFA, because JECFA concluded on the safety of this substance at step A3 of the Procedure (JECFA, 2006a, 2006b). EFSA considerations Since [FL‐no: 07.170] is evaluated via the B‐side of the Procedure, the CEF Panel requested toxicity data for this substance in FGE.82Rev1 (EFSA CEF Panel, 2014a). The new data provided are described below and summarised in Appendix D. Toxicity studies on beta‐ionone epoxide [FL‐no: 07.170] 14‐day dose range‐finding study A 14‐day dose range‐finding study (Documentation provided to EFSA No. 7) was performed with beta‐ionone epoxide (purity 95.6%) to support dose selection for the subsequent 90‐day oral toxicity study. The study was not GLP compliant. Four groups of seven to 8 weeks old CRL Sprague–Dawley® CD® IGS rats (5/sex/group) were fed with a diet containing beta‐ionone epoxide at concentrations of 0, 1200, 2400 and 4800 mg/kg that would correspond to intended doses of 0, 100, 200 and 400 mg/kg bw per day for 14 days. The study authors calculated the mean nominal overall daily intake to be 0, 106, 214 and 419 mg/kg bw per day for males and 0, 103, 211 and 421 mg/kg bw per day for females. To evaluate the stability of beta‐ionone epoxide in diet, samples of the diet were collected on days 0, 4, 7 and 10. Due to marked variability of analytical results, the nominal overall intake could not be confirmed. The test substance was considered stable over the course of the study under the conditions of storage, but stability in the diet could not be assessed due to large variability in the measured concentrations. No mortality occurred and no test substance‐related clinical observations were observed. A statistically significant lower body weight (p < 0.05) was observed in females of the low dose group compared to controls on day 7. At the low dose, females had significantly lower (p < 0.01) mean body weight gain on days 0–7 compared to controls. A slight, but statistically significant decrease in food consumption was observed in males at the highest dose on days 0–3, and in females at the low and mid dose on days 3–7 and 10–14. All animals in the study were subjected to gross necropsy; there were no macroscopic alterations in any of the rats. The study authors concluded that treatment with beta‐ionone epoxide was well tolerated at doses up to 400 mg/kg bw per day (the highest dose tested). 90‐day toxicity study Beta‐ionone epoxide (purity 95.6%) was tested in a 90‐day repeated dose toxicity study in rats (Documentation provided to EFSA No. 8) according to OECD TG 408 (OECD, 1998) and in compliance with GLP. Four groups of seven to 8 weeks old CRL Sprague–Dawley® CD® IGS rats (10/sex/group) were fed a diet containing beta‐ionone epoxide aimed to provide doses of 0, 20, 40 or 80 mg/kg bw per day. Based on food consumption data (collected throughout the study), the actual mean doses were 0, 19, 39 and 77 mg/kg bw per day for males and 0, 20, 39 and 78 mg/kg bw per day for females. Stability of the test substance in the diet was confirmed by analysis, and variability of test substance concentrations in the diet was considered to be within an acceptable margin of variation. The study authors reported that selection of dose levels was based on the 14‐day palatability/toxicity study. However, the highest tolerated dose of 400 mg/kg bw per day in the 14‐day study was not tested in the 90‐day study. In the respective publication of this study (Lu et al., 2021), the authors justified that the dose range selected for the 90‐day study was limited by test substance availability as the annual production volume for this substance is low. No mortality occurred and no test substance‐related clinical signs were observed. No effects on ophthalmological parameters, body weights and food consumption were observed. No changes in organ weights or organ weights relative to terminal body and brain weights were observed in male rats fed a diet with beta‐ionone epoxide. In the females of the high dose group, a significant increase in absolute liver weight, liver‐to‐body weight ratio (p < 0.05) and liver‐to‐brain weight ratio was observed. Without any correlating hepatic histopathological findings, these increases are interpreted to be non‐adverse. Haematology, clinical chemistry and urine analysis were carried out in blood and urine samples collected on days 51 and 86 from all animals. Blood samples were also collected prior to necropsy for coagulation determinations on day 94/95. The Panel noted some statistically significant changes in clinical parameters (prothrombin time, absolute large unstained cell count, absolute reticulocyte count, aspartate aminotransferase, bilirubin, calcium, chloride); however, these were considered non‐adverse as they were within historical control ranges, showed no biological significance, were not supported by corresponding histopathological findings and/or were not dose‐related. Oestrus cycle was also evaluated in all groups on 6–7 and 12–13 weeks. The mean cycle length and the number of cycles were not altered at any dose. No alterations on spermatozoid motility, epididymal sperm count, homogenisation‐resistant sperm count or percent abnormal sperm were observed. Macroscopic findings include a small right adrenal gland in one male of the mid‐dose group and a kidney cyst in one male of the high dose group, which had no microscopic correlation in both cases. Uterus was noted as fluid‐filled at necroscopy in 8/10, 3/10, 6/10 and 4/10 females of the control group and of the low‐, mid‐ and high dose, respectively. Fluid‐filled uteri usually corresponded microscopically to luminal dilation of the uterus that was attributable by the study authors to variations in the oestrus cycle. Additionally, one female of the control group presented agenesis of the left ovary, noted macroscopically and microscopically, which was considered by the study authors to be not related to administration of beta‐ionone epoxide. There were treatment‐related adrenal gland findings in 8/10 males of the high dose group. The adrenal glands exhibited minimal to moderate diffuse cortical vacuolation, which was limited to the zona fasciculata. The vacuoles were clear, variably sized and discrete. An increased incidence and severity of diffuse cortical vacuolation was observed in 8 high‐dose males (5 rats with minimal severity, 1 mild, 2 moderate) when compared to controls (3/10 males, minimal severity). Although this alteration was considered to be test substance‐related, the study authors concluded that it was not adverse due to its non‐degenerative nature (Documentation provided to EFSA No. 8). The Panel considered that vacuolation of the zona fasciculata has been described as a background finding in rats (McInnes, 2012). However, it could also be the result of chemically‐induced impairment of steroid synthesis leading to lipid accumulation (Brändli‐Baiocco et al., 2018). Despite the lack of degenerative changes (i.e. degeneration/necrosis), the Panel considered the adrenal findings, at the high dose, as adverse based on the severity (moderate) that is above the threshold of concern (i.e. minimal to mild). In the absence of data from functional assays demonstrating the preserved functionality of adrenal cortex this concern could not be dismissed. Thus, based on the substance‐related findings observed in adrenal glands of males of the high dose group that were considered adverse, the Panel identified the (actual) dose of 39 mg/kg bw per day as the NOAEL in this study. Application of the procedure Application of the procedure by JECFA According to JECFA (JECFA, 2006a, 2006b) all five substances belong to structural class III, using the decision tree approach presented by Cramer et al. (1978). JECFA considered that all five substances can be predicted to be metabolised to innocuous products (step 2 of the Procedure), so they were evaluated along the A‐side of the Procedure. For three substances [FL‐no: 07.170, 16.040 and 16.043], the estimated intakes performed by JECFA were below the toxicological threshold of concern (TTC) (90 μg/person per day) for structural class III (step A3 of the Procedure). For two substances [FL‐no: 16.015 and 16.018], the estimated intakes were above the TTC for structural class III, and they do not occur endogenously in humans (step A4 of the Procedure). Therefore, the evaluation proceeded to step A5, where a long‐term study (Dunnington et al., 1981) for ethyl methylphenylglycidate [FL‐no: 16.015] was considered by JECFA. From this study, a NOAEL of 35 mg/kg bw per day was derived, which provides a margin of safety of more than 8000 for [FL‐no: 16.015] and more than 17,000 times the estimated intake of the related substance, [FL‐no: 16.018]. In conclusion, JECFA evaluated all five substances to be of no safety concern at the estimated levels of intake as flavouring substances based on the MSDI approach. The summary evaluations by JECFA of the five epoxides are presented in Appendix E. EFSA considerations In FGE.82Rev1 (EFSA CEF Panel, 2014a), the CEF Panel did not agree with the application of the Procedure for the epoxides as performed by JECFA (see Section 3.3.1). The five substances [FL‐no: 07.170, 16.015, 16.018, 16.040 and 16.043] were evaluated via the B‐side of the Procedure scheme (Appendix A). Taking into account the absorption, distribution, metabolism, excretion (ADME) considerations (Section 3.3.1), the Panel agreed with the CEF Panel and assessed [FL‐no: 07.170] via the B‐side of the Procedure scheme (Appendix A): Step 1. The Panel agreed with JECFA with respect to the allocation of the candidate flavouring substance [FL‐no: 07.170] in structural class III. Step 2. The Panel did not agree with JECFA that [FL‐no: 07.170] can be predicted to be completely metabolised to innocuous products. Therefore, the evaluation was performed via the B‐side of the Procedure. Step B3. The estimated MSDI value of 0.4 μg/capita per day, calculated by the Panel, is below the TTC for class III (Table C.4, Appendix C). Step B4. Based on a NOAEL of 39 mg/kg bw per day identified from a 90‐day toxicity study on [FL‐no: 07.170] (see Section 3.3.3) and the estimated MSDI value, an adequate margin of exposure of almost 6 million is calculated. 10 Thus, the Panel concluded that beta‐ionone epoxide [FL‐no: 07.170] is of no safety concern when based on the MSDI approach. In addition, industry provided updated use levels for [FL‐no: 07.170], resulting in an estimation of mTAMDI of 840 μg/person per day (Appendix C). Industry also provided new information on uses and use levels for the other structural class III substances [FL‐no: 16.015, 16.018, 16.040 and 16.043] included in FGE82, allowing the estimation of the mTAMDI for each substance of 710, 630, 880 and 1200 μg/person per day, respectively. All mTAMDI values are above the TTC for structural class III substances of 90 μg/person per day. Therefore, the Panel concluded that more reliable data on uses and use levels are needed for [FL‐no: 07.170, 16.015, 16.018, 16.040 and 16.043]. When these data become available, the assessment for these flavouring substances may be updated accordingly and expanded if necessary (e.g. request of additional toxicity data). DISCUSSION The present revision of FGE.82 (FGE.82Rev2) is based on new toxicological data, information on stereoisomeric composition and updated information on uses and use levels for beta‐ionone epoxide [FL‐no: 07.170]. In addition, new information on uses and use levels for four flavouring substances [FL‐no: 16.015, 16.018, 16.040 and 16.043] has been provided and considered for the estimation of their intake with the MSDI and mTAMDI approach. The new information provided on the stereochemical composition of [FL‐no: 07.170] is considered sufficient by the Panel. Taking into account the ADME considerations (Section 3.3.1), the Panel did not agree with the application of the Procedure for the epoxides as performed by JECFA. Since the Panel cannot confirm that the flavouring substances in this group can be predicted to be completely metabolised to innocuous products, beta‐ionone epoxide [FL‐no: 07.170] was assessed via the B‐side of the Procedure, in line with the assessment performed by the CEF Panel for the other four epoxides in the previous revisions (EFSA, 2009b; EFSA CEF Panel, 2014a). The Panel identified the dose of 39 mg/kg bw per day as the NOAEL in the 90‐day study on [FL‐no: 07.170] based on the findings observed in adrenal glands in males of the high dose group suggesting steroid synthesis impairment with lipid accumulation. Based on the NOAEL and on the estimated MSDI, an adequate margin of exposure (of almost 6 million) was calculated. Therefore, the Panel considered that [FL‐no: 07.170] does not pose a safety concern when used as flavouring substance at estimated levels of intake, when based on the MSDI approach. For all five substances [FL‐no: 07.170, 16.015, 16.018, 16.040, 16.043] in FGE.82Rev2, mTAMDI estimates exceeded the TTC of 90 μg/person per day for structural class III substances. For these substances, more reliable data on uses and use levels should be provided to refine the exposure assessment to finalise their safety evaluation. The Panel considered that the available specifications for the five substances in FGE.82Rev2 are adequate and that the conclusions in FGE.82Rev2 can be applied to the materials of commerce. CONCLUSIONS The Panel evaluated beta‐ionone epoxide [FL‐no: 07.170] through the Procedure and concluded in agreement with JECFA that there is ‘no safety concern at estimated levels of intake as flavouring substance’, when based on the MSDI approach. For the other four substances included in FGE.82Rev2, [FL‐no: 16.015, 16.018, 16.040, 16.043] the same conclusion, i.e. ‘no safety concern at estimated levels of intake as flavouring substances’, when based on the MSDI approach, was already drawn in the previous opinions (FGE.82 and FGE.82Rev1). However, for all the five substances [FL‐no: 07.170, 16.015, 16.018, 16.040, 16.043] in FGE.82Rev2, mTAMDI estimates exceed the TTC of 90 μg/person per day for structural class III substances. Therefore, for these substances, more reliable data on uses and use levels would be needed to refine the exposure assessments and to finalise their safety evaluation. RECOMMENDATION The stereochemical information available for [FL‐no: 07.170, 16.015, 16.018, 16.040, 16.043] and presented in Table B.1 (Appendix B) should be included into the Union List. DOCUMENTATION AS PROVIDED TO EFSA Dossier on beta‐ionone epoxide [FL‐no: 07.170] submitted by EFFA. August 2025.DG SANCO (Directorate General for Health and Consumer Affairs), 2012. Information from DG SANCO 07/022012, concerning a list of 85 substances in the Register and 15 substances not in the Register which are non‐supported substances and one list of 30 substances in the Register for which no data have been submitted or which are duplicates. FLAVIS.2.23rev1.DG SANCO (Directorate General for Health and Consumer Affairs), 2014. Information from DG SANCO concerning a list of use levels for 123 JECFA evaluated substances allocated to structural class III. 16.09.2014EFFA (European Flavour Association), 2002. Letter from EFFA to Danish Veterinary and Food Administration. Dated 31 October 2002. Re.: Second group of questions. FLAVIS/8.26.EFFA (European Flavour Association), 2009. Supplement list of EU‐only Footnote‐10 materials for Commission. Unpublished communication submitted by EFFA to the FLAVIS secretariat. 14 December 2009.EFFA (European Flavour Association), 2013. E‐mail from EFFA to FLAVIS Secretariat, Danish Food Institute, Technical University of Denmark, dated 24 June 2013. Information on substances in FGE.82Rev1 [FL‐no: 16.015, 16.018 and 16.040]. FLAVIS/8.218.■■■■■, 2015. Beta‐epoxy ionone: a 14‐day dietary toxicity/palatability study in rats. Study No. 40134. August 2015. Submitted by EFFA.■■■■■, 2016. Beta‐epoxy ionone: a 90‐day dietary study in rats. Study No. 40135. February 2016. Submitted by EFFA. Dossier on beta‐ionone epoxide [FL‐no: 07.170] submitted by EFFA. August 2025. Dossier on beta‐ionone epoxide [FL‐no: 07.170] submitted by EFFA. August 2025. DG SANCO (Directorate General for Health and Consumer Affairs), 2012. Information from DG SANCO 07/022012, concerning a list of 85 substances in the Register and 15 substances not in the Register which are non‐supported substances and one list of 30 substances in the Register for which no data have been submitted or which are duplicates. FLAVIS.2.23rev1. DG SANCO (Directorate General for Health and Consumer Affairs), 2012. Information from DG SANCO 07/022012, concerning a list of 85 substances in the Register and 15 substances not in the Register which are non‐supported substances and one list of 30 substances in the Register for which no data have been submitted or which are duplicates. FLAVIS.2.23rev1. DG SANCO (Directorate General for Health and Consumer Affairs), 2014. Information from DG SANCO concerning a list of use levels for 123 JECFA evaluated substances allocated to structural class III. 16.09.2014 DG SANCO (Directorate General for Health and Consumer Affairs), 2014. Information from DG SANCO concerning a list of use levels for 123 JECFA evaluated substances allocated to structural class III. 16.09.2014 EFFA (European Flavour Association), 2002. Letter from EFFA to Danish Veterinary and Food Administration. Dated 31 October 2002. Re.: Second group of questions. FLAVIS/8.26. EFFA (European Flavour Association), 2002. Letter from EFFA to Danish Veterinary and Food Administration. Dated 31 October 2002. Re.: Second group of questions. FLAVIS/8.26. EFFA (European Flavour Association), 2009. Supplement list of EU‐only Footnote‐10 materials for Commission. Unpublished communication submitted by EFFA to the FLAVIS secretariat. 14 December 2009. EFFA (European Flavour Association), 2009. Supplement list of EU‐only Footnote‐10 materials for Commission. Unpublished communication submitted by EFFA to the FLAVIS secretariat. 14 December 2009. EFFA (European Flavour Association), 2013. E‐mail from EFFA to FLAVIS Secretariat, Danish Food Institute, Technical University of Denmark, dated 24 June 2013. Information on substances in FGE.82Rev1 [FL‐no: 16.015, 16.018 and 16.040]. FLAVIS/8.218. EFFA (European Flavour Association), 2013. E‐mail from EFFA to FLAVIS Secretariat, Danish Food Institute, Technical University of Denmark, dated 24 June 2013. Information on substances in FGE.82Rev1 [FL‐no: 16.015, 16.018 and 16.040]. FLAVIS/8.218. ■■■■■, 2015. Beta‐epoxy ionone: a 14‐day dietary toxicity/palatability study in rats. Study No. 40134. August 2015. Submitted by EFFA. ■■■■■, 2015. Beta‐epoxy ionone: a 14‐day dietary toxicity/palatability study in rats. Study No. 40134. August 2015. Submitted by EFFA. ■■■■■, 2016. Beta‐epoxy ionone: a 90‐day dietary study in rats. Study No. 40135. February 2016. Submitted by EFFA. ■■■■■, 2016. Beta‐epoxy ionone: a 90‐day dietary study in rats. Study No. 40135. February 2016. Submitted by EFFA. ABBREVIATIONS absorption, distribution, metabolism, excretion Panel on Food Additives, Flavourings, Processing Aids and Materials in Contact with Food body weight Chemical Abstract Service Panel on Food Contact Materials, Enzymes, Flavourings and Processing Aids Council of Europe European Flavour Association female Panel on Food Additives and Flavourings Food and Agriculture Organisation Flavour and Extract Manufacturer Association Flavouring Group Evaluation Flavour Information System (database) Good Laboratory Practice identity infrared spectroscopy The Joint FAO/WHO Expert Committee on Food Additives male mass spectrometry maximised survey‐derived daily intake modified theoretical added maximum daily intake nuclear magnetic resonance number no observed adverse effect level Organisation for Economic Co‐operation and Development (quantitative) structure–activity relationship Scientific Committee Scientific Committee on Food toxicological threshold of concern United States of America World Health Organization REQUESTOR European Commission QUESTION NUMBER EFSA‐Q‐2025‐00494 COPYRIGHT FOR NON‐EFSA CONTENT EFSA may include images or other content for which it does not hold copyright. In such cases, EFSA indicates the copyright holder and users should seek permission to reproduce the content from the original source. PANEL MEMBERS Laurence Castle, Monica Andreassen, Gabriele Aquilina, Maria Bastos, Polly Boon, Biagio Fallico, Rex FitzGerald, María José Frutos Fernández, Bettina Grasl‐Kraupp, Ursula Gundert‐Remy, Rainer Gürtler, Eric Houdeau, Marcin Kurek, Henriqueta Louro, Patricia Morales, Sabina Passamonti LEGAL NOTICE The published scientific output includes information items that have been redacted, in accordance with the confidentiality requests submitted by the data provider, pending a decision thereon by EFSA. Redactions may be subject to review once a decision on such confidentiality requests is taken by EFSA.