The EFSA Journal (2006) 372, 1-63

Transcription

1 The EFSA Journal (006) 37, 63 pinion of the Scientific Panel on Food Additives, Flavourings, Processing Aids and Materials in contact with Food (AFC) on a request from the Commission related to Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 (Commission Regulation (EC) o 6/000 of 8 July 000) QUESTI EFSAQ00367 Adopted on 4 May 006 SUMMARY The Scientific Panel on Food Additives, Flavourings, Processing Aids and Materials in Contact with Food is asked to advise the Commission on the implications for human health of chemically defined flavouring substances used in or on foodstuffs in the Member States. In particular, the Scientific Panel is asked to evaluate 3 flavouring substances in the Flavouring Group Evaluation FGE.4, using the procedure as referred to in the Commission Regulation EC o 6/000. These 3 flavouring substances belong to chemical group 8, Annex I of the Commission Regulation EC o 6/000. The present Flavouring Group Evaluation deals with 3 pyridine, pyrrole, indole and quinoline derivatives. one of the 3 candidate substances can exist as geometrical or optical isomers. Twentyfive of the flavouring substances are classified into structural class II and six are classified into structural class III. Twentyeight of the flavouring substances in the present group have been reported to occur naturally in a wide range of food items. In its evaluation, the Scientific Panel as a default used the Maximised Surveyderived Daily Intakes (MSDIs) approach to estimate the per capita intakes of the flavouring substances in Europe. However, when the Scientific Panel examined the information provided by the European flavouring industry on the use levels in various foods, it appeared obvious that the MSDI approach in a number of cases would grossly underestimate the intake by regular consumers of products flavoured at the use level reported by the industry, especially in those cases where the annual production values were reported to be small. In consequence, the Scientific Panel had reservations about the data on use and use levels provided and the intake estimates obtained by the MSDI approach. In the absence of more precise information that would enable the Scientific Panel to make a more realistic estimate of the intakes of the flavouring substances, the Scientific Panel has decided also to perform an estimate of the daily intakes per person using a modified Theoretical Added Maximum Daily Intake (mtamdi) approach based on the normal use levels reported by industry. In those cases where the mtamdi approach indicated that the intake of a flavouring substance might exceed its corresponding threshold of concern, the Scientific Panel decided not to carry out a formal safety assessment using the Procedure. In these cases the Scientific Panel requires more precise data on use and use levels.

2 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 According to the default MSDI approach, the 3 flavouring substances in this group have intakes in Europe from 0.00 to 0.73 microgram/capita/day which are all below the thresholds of concern for both structural class II (40 microgram/person/day) and structural class III (90 microgram/person/day) substances. Data on genotoxicity of the flavouring substances are limited and the genotoxicity could not be adequately assessed. The data available, however, give rise to some concern of a genotoxic potential of the quinoline derivatives: methylquinoline [FLno: 4.38], 4methylquinoline [FLno: 4.00] and 4butylquinoline [FLno: 4.094]. The Scientific Panel, therefore, concluded that the Procedure could not be applied to these three substances until adequate in vivo data become available. The genotoxicity data available for the remaining candidate substances do not preclude their evaluation through the Procedure. Two of the 8 flavouring substances ethyl nicotinate [FLno: 4.0] and isopropyl nicotinate [FLno: 4.0] in this flavouring group, which are evaluated through the Procedure, are expected to be metabolised to innocuous products. For the remaining 6 flavouring substances it cannot be anticipated that they will be metabolised to innocuous products. It was noted that where toxicity data were available they were consistent with the conclusions in the present flavouring group evaluation using the Procedure. It was considered that on the basis of the default MSDI approach 3 of the 8 flavouring substances evaluated through the Procedure would not give rise to safety concerns at the estimated levels of intake arising from their use as flavouring substances. For the remaining five candidate substances acetylfurfurylpyrrole [FLno: 3.00], methylpyrrole [FLno: 4.03], acetyl methylpyrrole [FLno: 4.08],,dimethylpyrrole [FLno: 4.07] and pyrrolecarbaldehyde [FLno: 4.4], a AEL could not be derived for the substance or a structurally related substance. Accordingly, additional data are required for these five flavouring substances. When the estimated intakes were based on the mtamdi they ranged from 380 to 60 microgram/person/day for the flavouring substances from structural class II. The intake was above the threshold of concern of 40 microgram/person/day for one of these candidate substances. The mtamdi estimates of the three flavouring substances, assigned to structural class III and evaluated using the Procedure, are between 0 and 380 microgram/person/day, which are all above the threshold of concern for structural class III of 90 microgram/person/day. Thus for four of the 8 flavouring substances evaluated through the Procedure the intakes, estimated on the basis of the mtamdi, exceed the relevant threshold for their structural class, to which the flavouring substance has been assigned. Therefore, for these four substances [FLno: 3.00, 4.03, and 4.3] more reliable exposure data are required. n the basis of such additional data, these flavouring substances should be reconsidered along the steps of the Procedure. Following this procedure additional toxicological data might become necessary. In order to determine whether this evaluation could be applied to the materials of commerce, it is necessary to consider the available specifications. Adequate specifications including complete purity criteria and identity tests for the materials of commerce have been provided for all 3 flavouring substances and these would present no safety concern at the levels of intake estimated on the basis of the MSDI approach. KEYWRDS Flavourings, pyridine, pyrrole, indole, quinoline, safety.

3 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 TABLE F CTETS Summary... Keywords... Background... 4 Terms of Reference... 4 Assessment Presentation of the Substances in the Flavouring Group Evaluation Description Stereoisomers atural ccurrence in Food.... Specifications Intake Data Estimated Daily per Capita Intake (MSDI Approach) Intake Estimated on the Basis of the Modified TAMDI (mtamdi) Absorption, Distribution, Metabolism and Elimination Application of the Procedure for the Safety Evaluation of Flavouring Substances Comparison of the Intake Estimations Based on the MSDI Approach and the mtamdi Approach Considerations of Combined Intakes from Use as Flavouring Substances Toxicity Acute Toxicity Subacute, Subchronic, Chronic and Carcinogenicity Studies Developmental / Reproductive Toxicity Studies Genotoxicity Studies Conclusions... Table : Specification Summary of the Substances in the Flavouring Group Evaluation 4... Table a: Summary of Safety Evaluation Applying the Procedure (Based on Intakes Calculated by the MSDI Approach)... Table b: Evaluation Status of Hydrolysis Products of Candidate Esters... Table 3: Supporting Substances Summary... Annex I: Procedure for the Safety Evaluation... Annex II: Use Levels / mtamdi... Annex III: Metabolism... Annex IV: Toxicity... References:... Scientific Panel Members... Acknowledgement... 3

4 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 BACKGRUD Regulation (EC) o 3/96 of the European Parliament and the Council (EC, 996) lays down a procedure for the establishment of a list of flavouring substances, the use of which will be authorised to the exclusion of all others in the EU. In application of that Regulation, a register of flavouring substances used in or on foodstuffs in the Member States was adopted by Commission Decision 999/7/EC (EC, 999a), as last amended by Commission Decision 00/389/EC (EC, 00). Each flavouring substance is attributed a FLAVISnumber (FLnumber) and all substances are divided into 34 chemical groups. Substances within a group should have some metabolic and biological behaviour in common. Substances which are listed in the register are to be evaluated according to the evaluation programme laid down in Commission Regulation (EC) o 6/000 (EC, 000) which is broadly based on the opinion of the Scientific Committee on Food (SCF, 999). For the submission of data by the manufacturer, deadlines have been established by Commission Regulation (EC) o 6/00 (EC, 00b). After the completion of the evaluation programme the positive list of flavouring substances for use in or on foods in the EU shall be adopted (Article () of Regulation (EC) o 3/96) (EC, 996). TERMS F REFERECE EFSA is requested to carry out a risk assessment on flavouring substances prior to their authorisation and inclusion in a positive list according to Commission Regulation (EC) o 6/000 (EC, 000). ASSESSMET. Presentation of the Substances in the Flavouring Group Evaluation 4.. Description The present Flavouring Group Evaluation, using the procedure as referred to in the Commission Regulation EC o 6/000 (EC, 000) (The Procedure shown in schematic form in Annex I), deals with 3 derivatives of pyridine, pyrrole, indole and quinoline from chemical group 8, Annex I of Commission Regulation (EC) o 6/000 (EC, 000). The 3 flavouring substances have one heterocyclic atom in a five or six membered ring structure. Twentysix of the compounds are monocyclic, with either a five membered ring (pyrroles) [FLno: 3.00, 4.03, 4.08, 4.07 and 4.4] or a six membered ring (pyridines) [FLno: 4.089, 4.09, 4.093, 4.03, 4.04, 4.0, 4.06, 4.0, 4., 4.6, 4.7, 4.8, 4.0, 4.4, 4., 4.34, 4.3, 4.36, 4.40, 4.43 and 4.0]. The remaining five candidate substances are aromatic bicyclic structures consisting of either two six membered rings (quinolines) [FLno: 4.00, and 4.38] or one six and one five membered ring (indoles) [FLno: and 4.3]. The substances are substituted with simple alkyl, alcohol, aldehyde, ketone or ester moities, except for one which possesses a furfuryl group as substituent. The 3 flavouring substances under consideration, as well as their chemical names, FLAVIS (FL), Chemical Abstract Service (CAS), Council of Europe (CoE) and Flavor and Extract Manufactures Association (FEMA) numbers, structure and specifications, are listed in Table. 4

5 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 The 3 flavouring substances (candidate substances) are closely related structurally to 9 flavouring substances (supporting substances) evaluated at the 63 rd JECFA meeting (JECFA, 006a) in the group of Pyridine, pyrrole and quinoline derivatives. Two other structurally related flavouring substances (8methylquinoline and indole3carbinol) are also included as supporting substances in this evaluation. The candidate substances under consideration in the present evaluation are listed in Tables and a and the supporting substances are listed in Table 3. The hydrolysis products of the two esters [FLno: 4.0 and 4.0] and the amide [FLno: 4.088] contained in the present FGE as well as their evaluation status are listed in Table b... Stereoisomers one of the 3 candidate substances can exist as geometrical or optical isomers..3. atural ccurrence in Food Twentyeight out of 3 flavouring substances have been reported to occur in one or more of the following food items: tea, coffee, cocoa, fruits, almond, asparagus, cheese, pork, chicken, shrimps, beer, whisky, malt, popcorn and peppermint oil. Quantitative data on the natural occurrence in food have been reported for 6 of the 8 candidate substances. These reports include among others: Acetylfurfurylpyrrole [FLno: 3.00]: Up to. mg/kg in coffee. 4Methylquinoline [FLno: 4.00]: Up to 0. mg/kg in peppermint oil. Methylpyrrole [FLno: 4.03]: mg/kg in chicken, 0. mg/kg in cocoa, up to. mg/kg in coffee, 0.3 mg/kg in tamarind. Acetylmethylpyrrole [FLno: 4.08]: 0.3 mg/kg in coffee, 0.0 mg/kg in malt.,3dimethylpyridine [FLno: 4.03]: mg/kg in pork, mg/kg in tea.,4dimethylpyridine [FLno: 4.04]: Trace amounts in chicken, up to mg/kg in pork, 0.00 mg/kg in shrimps. 3,4Dimethylpyridine [FLno: 4.0]: Up to mg/kg in whisky. 3,Dimethylpyridine [FLno: 4.06]: Up to 0. mg/kg in tea. Ethyl nicotinate [FLno: 4.0]: Up to 0.06 mg/kg in apricot, up to 0.0 mg/kg in arctic bramble, up to.4 mg/kg in beer, up to 0.3 mg/kg in papaya. Ethylpyridine [FLno: 4.]: Trace amounts in asparagus, mg/kg in pork. 4Ethylpyridine [FLno: 4.6]: mg/kg in pork. Methylpyridine [FLno: 4.34]: 0.08 mg/kg in cheese, mg/kg in chicken, mg/kg in pork, up to 0.0 mg/kg in shrimps, 0.00 mg/kg in whisky. 3Methylpyridine [FLno: 4.3]: mg/kg in beer,.3 mg/kg in coffee and up to mg/kg in whisky. 4Methylpyridine [FLno: 4.36]: mg/kg in pork, trace amounts in shrimps. Pyrrolecarbaldehyde [FLno: 4.4]: 8 mg/kg in almond, 0.06 mg/kg in asparagus, up to.7 mg/kg in coffee, 0.8 mg/kg in malt, up to 0.3 mg/kg in tea.,4,6trimethylpyridine [FLno: 4.0]: Up to 0. mg/kg in tea. Three of the 3 substances have not been reported to occur naturally in any food items according to T (T, 000). These are: acetylindole [FLno: 4.088], isopropyl nicotinate [FLno: 4.0] and 4isopropylpyridine [FLno: 4.].

6 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37. Specifications Purity criteria for the 3 candidate substances have been provided by the Flavouring Industry (EFFA, 004l) (Table ). Judged against the requirements in Annex II of Commission Regulation EC o 6/000 (EC, 000), the information is adequate for all of the candidate substances (see section. and Table ). 3. Intake Data Annual production volumes of the flavouring substances as surveyed by the Industry can be used to calculate the Maximized SurveyDerived Daily Intake (MSDI) by assuming that the production figure only represents 60 % of the use in food due to underreporting and that 0 % of the total EU population are consumers (SCF, 999). However, the Panel noted that due to yeartoyear variability in production volumes, to uncertainties in the underreporting correction factor and to uncertainties in the percentage of consumers, the reliability of intake estimates on the basis of the MSDIapproach is difficult to assess. The Panel also noted that in contrast to the generally low per capita intake figures estimated on the basis of this MSDIapproach, in some cases the regular consumption of products flavoured at use levels reported by the Flavour Industry in the submissions would result in much higher intakes. In such cases, the human exposure thresholds below which exposures are not considered to present a safety concern might be exceeded. Considering that the MSDI model may underestimate the intake of flavouring substances by certain groups of consumers, the SCF recommended also taking into account the results of other intake assessments (SCF, 999). ne of the alternatives is the Theoretical Added Maximum Daily Intake (TAMDI)approach which is calculated on the basis of standard portions and upper use levels (SCF, 99) for flavourable beverages and foods in general, with exceptional levels for particular foods. This method is regarded as a conservative estimate of the actual intake in most consumers because it is based on the assumption that the consumer regularly eats and drinks several food products containing the same flavouring substance at the upper use level. ne option to modify the TAMDIapproach is to base the calculation on normal rather than upper use levels of the flavouring substances. This modified approach is less conservative (e.g., it may underestimate the intake of consumers being loyal to products flavoured at the maximum use levels reported (EC, 000). However, it is considered as a suitable tool to screen and prioritise the flavouring substances according to the need for refined intake data (EFSA, 004a). 3.. Estimated Daily per Capita Intake (MSDI Approach) The Maximised SurveyDerived Daily Intake (MSDI (SCF, 999)) data are derived from surveys on annual production volumes in Europe. These surveys were conducted in 99 by the International rganization of the Flavour Industry, in which flavour manufacturers reported the total amount of each flavouring substance incorporated into food sold in the EU during the previous year (IFI, 99). The intake approach does not consider the possible natural occurrence in food. 6

7 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 Average per capita intake (MSDI) is estimated on the assumption that the amount added to food is consumed by 0 % of the population (Eurostat, 998). This is derived for candidate substances from estimates of annual volume of production provided by Industry and incorporates a correction factor of 0.6 to allow for incomplete reporting (60%) in the Industry surveys (SCF, 999). In the present Flavouring Group Evaluation (FGE.4) the total annual production volume of the 3 candidate substances from use as flavouring substances in Europe was reported to be 8 kg (EFFA, 004m). For of the 9 supporting substances the total annual volume of production in Europe is approximately 990 kg (JECFA, 006a). The annual volumes of production in Europe for four of the supporting substances [FLno: 4.04, 4.08, 4.09 and 4.64] were not reported. n the basis of the annual volumes of production reported for the 3 candidate substances, the daily per capita intakes for each of these flavourings have been estimated (Table a). More than 0% of the total annual volume of production for the candidate substances (EFFA, 004m) is accounted for by the following three flavourings: 4methylpyridine [FLno: 4.36], methylpyrrole [FLno: 4.03] and methylpyridine [FLno: 4.34]. The estimated daily per capita intakes of these candidate substances from use as flavouring substances are 0.73, 0.3 and 0. microgram, respectively. The daily per capita intakes for each of the remaining substances are less than 0. microgram (Table a). 3.. Intake Estimated on the Basis of the Modified TAMDI (mtamdi) The method for calculation of modified Theoretical Added Maximum Daily Intake (mtamdi) values is based on the approach used by SCF up to 99 (SCF, 99). The assumption is that a person may consume a certain amount of flavourable foods and beverages per day. For the present evaluation of the 3 candidate substances, information on food categories and normal and maximum use levels,3 were submitted by the Flavour Industry (EFFA, 004l). The 3 candidate substances are used in flavoured food products divided into the food categories, outlined in Annex III of the Commission Regulation 6/000 (EC, 000), as shown in Table 3.. For the present calculation of mtamdi, the reported normal use levels were used. In the case where different use levels were reported for different food categories the highest reported normal use level was used. EU figure 37 millions (Eurostat, 998). This figure relates to EU population at the time for which production data are available, and is consistent (comparable) with evaluations conducted prior to the enlargement of the EU. o production data are available for the enlarged EU ormal use is defined as the average of reported usages and maximum use is defined as the 9th percentile of reported usages (EFFA, 00i) 3 The normal and maximum use levels in different food categories (EC, 000) have been extrapolated from figures derived from model flavouring substances (EFFA, 004e). 7

8 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 Table 3. Use of Candidate Substances Food category Description Category Dairy products, excluding products of category All 3 Category Fats and oils, and fat emulsions (type waterinoil) All 3 Category 3 Edible ices, including sherbet and sorbet All 3 Category 4. Processed fruits All 3 Category 4. Processed vegetables (incl. mushrooms & fungi, roots & tubers, pulses and legumes), and nuts & seeds Flavourings used one Category Confectionery All 3 Category 6 Cereals and cereal products, incl. flours & starches from roots & tubers, pulses & legumes, excluding bakery All 3 Category 7 Bakery wares All 3 Category 8 Meat and meat products, including poultry and game All 3 Category 9 Fish and fish products, including molluscs, crustaceans and echinoderms All 3 except [FLno: 4.] Category 0 Eggs and egg products nly [FLno: 4.] Category Sweeteners, including honey one Category Salts, spices, soups, sauces, salads, protein products etc. All 3 Category 3 Foodstuffs intended for particular nutritional uses. All 3 Category 4. onalcoholic ("soft") beverages, excl. dairy products All 3 Category 4. Alcoholic beverages, incl. alcoholfree and lowalcoholic counterparts one Category Readytoeat savouries All 3 Category 6 Composite foods (e.g. casseroles, meat pies, mincemeat) foods that could not be placed in categories According to the Flavour Industry the normal use levels for the 3 candidate substances are in the range of 0. mg/kg food, and the maximum use levels are in the range of mg/kg (EFFA, 004l). The mtamdi values for the candidate substances from structural class II (see section ) range from 380 to 60 microgram/person/day. For the six candidate substances from structural class III (see section ) the mtamdi range from 0 to 380 microgram/capita/day. For detailed information on use levels and intake estimations based on the mtamdi approach, see Section 6 and Annex II. All 3 4. Absorption, Distribution, Metabolism and Elimination The candidate substances are divided into six subgroups with more closely related structural properties within the subgroups as shown in Table 4.. Table 4. Candidate substances divided into groups of related chemical structures Subgroup Candidate substance FLno Chemical group,dimethylpyrrole 4.07 Pyrrolecarbaldehyde 4.4 Pyrroles Acetylmethylpyrrole 4.08 Acetylfurfurylpyrrole 3.00 Methylpyrrole 4.03 substituted pyrroles 3 Methylindole 4.3 Acetylindole Indoles 4 Methylpyridine 4.34 Pyridines with alkyl, hydroxyl 8

9 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 Table 4. Candidate substances divided into groups of related chemical structures Subgroup Candidate substance FLno Chemical group 3Methylpyridine 4.3 4Methylpyridine 4.36 Ethylpyridine 4. 4Ethylpyridine 4.6 3Propylpyridine 4.43 Isopropylpyridine 4.4 4Isopropylpyridine 4. Butylpyridine Butylpyridine Pentylpyridine 4.40 Hexylpyridine 4.7,3Dimethylpyridine 4.03,4Dimethylpyridine ,4Dimethylpyridine 4.0 3,Dimethylpyridine 4.06,4,6Trimethylpyridine 4.0 Hydroxypyridine 4.8 4Acetylpyridine Ethyl nicotinate 4.0 Isopropyl nicotinate 4.0 Pyridines with ester groups Methylquinoline Methylquinoline 4.00 Quinolines 4Butylquinoline Subgroup : The group contains three pyrrol derivatives. For pyrrolecarbaldehyde [FLno: 4.4], which is not substituted on both C and C, hydroxylation in one of these positions might occur probably through an epoxide intermediate. The candidate substances with alkyl substitutions in both these positions,,dimethylpyrrole [FLno: 4.07] or acetyl methylpyrrole [FLno: 4.08], are more likely hydroxylated on the alkyl groups. Further oxidation of the aldehyde and acetyl groups in pyrrolecarbaldehyde and acetylmethylpyrrole is also expected. According to (Damani & Crooks, 98) pyrroles are likely substrates for enzymatic oxidation at. Subgroup : The group contains two substituted pyrrole derivatives. For methylpyrrole [FLno: 4.03], which is not substituted on both C and C, hydroxylation in one of these positions might occur probably through an epoxide intermediate. o metabolism study is available for the candidate substance acetylfurfurylpyrrole [Flno: 3.00]. From general knowledge, it is anticipated that the acetyl group of acetylfurfurylpyrrole may be hydroxylated at the terminal C and reduced in the carbonyl group. From the knowledge of the metabolism of pyrrole and furane compounds (Dalvie et al., 00) epoxidation of the double bonds of the furan and pyrrole groups may be anticipated. Subgroup 3: The group contains two indole derivatives. o data are available on the hydrolysis of acetylindole [FLno: 4.088] or structurally related substances. Amides are known to be hydrolysed but less rapidly than esters. A rapid hydrolysis of acetylindole in the gastrointestinal tract is not anticipated although hydrolysis by amidases in the tissues might be expected. The candidate substance methylindole [FLno: 4.3] will be metabolised through two main 9

10 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 pathways. xidation of the methyl group and the ring structure is anticipated, resulting in introduction of hydroxyl groups which are conjugated with glucuronic acid or sulphate. Epoxidation of the double bonds on the pyrrole ring is also expected, leading to oxindole compounds which can be further hydroxylated and conjugated with glucuronic acid or sulphate. The supporting substance 3methylindole is anticipated to be metabolised to the reactive metabolite 3methyleneindolenine which is conjugated with glutathione. Since a methyleneindolenine is not expected to be formed from methylindole, conjugation with glutathione is unlikely for this substance. Subgroup 4: The group contains 9 pyridine derivatives substituted with alkyl, hydroxyl or acetyl groups. Two main metabolism pathways can be predicted for the alkyl substituted pyridines. The main pathway is hydroxylation of the alkyl groups, with additional oxidation to the corresponding carboxyl compounds. The carboxyl metabolites are excreted in the urine as such or conjugated with glycine. Alternatively the candidate pyridines may be oxidised on the nitrogen atom, and excreted in the urine without conjugation. Hydroxylation of the pyridine ring through epoxide intermediates cannot be excluded, although not shown to be a major pathway of the supporting substances, 3 ethylpyridine,,3dimethylpyridine,,6dimethylpyridine, 3acetylpyridine and 3hydroxypyridine. However, this is probably dependent on the positions of the substitutions. From the metabolism studies on the supporting substances 3acetylpyridine and 3hydroxypyridine the candidate pyridines, 4acetylpyridine [FLno: 4.089] and hydroxypyridine [FLno: 4.8], are suggested to be metabolised through three pathways. The major metabolism pathways are reduction of keto groups and oxidation of the alkyl side groups resulting in alcohol and carboxylic acid sidegroups, which in some cases were further metabolised by conjugation with glycine and excreted in the urine. Some minor pathways include oxidation and methylation, and ring hydroxylation of the carbons of the ring structure. Subgroup : The group contains two esters of nicotinic acid: Ethyl nicotinate [FLno: 4.0] and isopropyl nicotinate [FLno: 4.0], which are expected to be hydrolysed in vivo to nicotinic acid and ethanol, and to nicotinic acid and isopropanol respectively. Subgroup 6: The group contains three quinoline derivatives. The candidate substances methylquinoline [FLno: 4.38], 4methylquinoline [FLno: 4.00] and 4butylquinoline [FLno: 4.094] are expected to be metabolised by hydroxylation of the ring structure in both rings. Some diol metabolites may be formed through epoxide intermediates. Hydroxylation of the alkyl groups is expected to be another main metabolic pathway. Additionally, the quinolines may be oxidised at the heteroatom leading to oxide metabolites. Two of the 3 candidate substances [FLno: 4.0 and 4.0] in this flavouring group evalution are expected to be metabolised to innocuous products. For the remaining 9 candidate substances it cannot be anticipated that they will be metabolised to innocuous products. For two of these candidate substances [FLno: 3.00 and 4.03] no metabolism studies were found for either the candidate substances or for the substances supporting it. For the remaining 7 candidate substances of this flavouring group [FLno: 4.00, 4.08, 4.088, 4.089, 4.09, 4.093, 4.094, 4.03, 4.04, 4.0, 4.06, 4.07, 4., 4.6, 4.7, 4.8, 4.4, 4., 4.3, 4.34, 4.3, 4.36, 4.38, 4.40, 4.43, 4.4 and 4.0] the Panel would expect metabolites such as oxides or epoxides, which cannot be anticipated to be innocuous products. A more detailed description of the metabolism is given in Annex III. 0

11 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37. Application of the Procedure for the Safety Evaluation of Flavouring Substances The application of the Procedure is based on intakes estimated on the basis of the MSDI approach. Where, the mtamdi approach indicates that the intake of a flavouring substance might exceed its corresponding threshold of concern, a formal safety assessment is not carried out using the Procedure. In these cases the Panel requires more precise data on use and use levels. For comparison of the intake estimations based on the MSDI approach and the mtamdi approach, see Section 6. For two of the candidate substances, methylquinoline [FLno: 4.38] and 4methylquinoline [FLno: 4.00] there is indication of a genotoxic potential in vitro. Therefore, in the absence of further genotoxicity data, the Panel concluded that the Procedure could neither be applied to these two substances, nor to the structurally related quinoline derivative, 4butylquinoline [FLno: 4.094]. For the safety evaluation of the remaining 8 candidate substances the Procedure as outlined in Annex I was applied, based on the MSDI approach. The stepwise evaluations of the 8 substances are summarised in Table. Step Twentyfive of the 8 candidate substances for which the Procedure could be applied are classified into structural class II and three candidate substances into structural class III [FLno: 3.00, and 4.3] according to the decision tree approach by Cramer et al. (Cramer et al., 978). Step Ethyl nicotinate [FLno: 4.0] and isopropyl nicotinate [FLno: 4.0] are expected to be hydrolysed in vivo to nicotinic acid and ethanol, and to nicotinic acid and isopropanol respectively. icotinic acid is a vitamin and accordingly the evaluation of the two candidate substances proceeds via the Aside of the Procedure. For two of the remaining 6 candidate substances [FLno: 3.00 and 4.03] no metabolism studies were found for either the candidate substances or the supporting substances. For the remaining 4 candidate substances of this flavouring group [FLno: 4.08, 4.088, 4.089, 4.09, 4.093, 4.03, 4.04, 4.0, 4.06, 4.07, 4., 4.6, 4.7, 4.8, 4.4, 4., 4.3, 4.34, 4.3, 4.36, 4.40, 4.43, 4.4 and 4.0] metabolites such as oxides or epoxides are expected. Therefore, for these 6 candidate substances it cannot be anticipated that they will be metabolised to innocuous products and accordingly they proceed via the Bside of the Procedure. Step A3 The two candidate substances evaluated along the Aside of the Procedure are both classified in structural class II. The substances have estimated daily per capita intakes from use as flavouring substances of 0.00 and 0.03 microgram. These intakes are below the threshold of concern of 40 microgram/person/day for structural class II and accordingly ethyl nicotinate [FLno: 4.0] and isopropyl nicotinate [FLno: 4.0] are not expected to be of safety concern at the estimated level of intake. Tolerable upper intake levels for nicotinic acid and nicotinamide are 0 and 900 mg/person/day, respectively (SCF, 00f).

12 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 Step B3 The levels of intake of the candidate substances classified in structural class II were estimated to be between 0.00 and 0.73 microgram/capita/day which are below the threshold of concern of 40 microgram/person/day. The levels of intake of the candidate substances classified in structural class III were estimated to be between 0.00 and 0.09 microgram/capita/day which are below the threshold of concern of 90 microgram/person/day. Therefore these 6 candidate substances proceed to step B4. Step B4 Subgroup and : For the five candidate substances: acetylfurfurylpyrrole [FLno: 3.00], methylpyrrole [FLno: 4.03], acetylmethylpyrrole [FLno: 4.08],,dimethylpyrrole [FLno: 4.07] and pyrrolecarbaldehyde [FLno: 4.4] a AEL could not be derived for the substance or a structurally related substance. Accordingly, additional toxicological data are required for these five substances. Subgroup 3: In an oral 37 weeks feeding study in rats on indole3carbinole, a substance structurally related to the two indole derivatives in this FGE, a AEL of 0 mg/kg/day could be derived. The combined estimated daily per capita intake of microgram for acetylindole [FLno: 4.088] and methylindole [FLno: 4.3] corresponds to 0.04 ng/kg bw/day at a body weight of 60 kg. Thus a margin of safety of.3 x 0 9 can be calculated. Acetylindole and methylindole are accordingly not expected to be of safety concern at the estimated level of intake. Subgroup 4: A 90 days oral feeding study in rats is available for acetylpyridine. The AEL derived is 37 mg/kg/day. The MSDI values for the 9 pyridine derivatives in this FGE are between 0.0 and 0. microgram/capita/day. The combined estimated daily per capita intake of these 9 derivatives is. microgram corresponding to 0.0 microgram/kg bw/day. Thus a margin of safety of. x 0 6 can be calculated using the AEL of 37 mg/kg/day. The 9 pyridine derivatives in this flavouring group are accordingly not expected to be of safety concern at the estimated level of intake. 6. Comparison of the Intake Estimations Based on the MSDI Approach and the mtamdi Approach The estimated intakes for the candidate substances in structural class II based on the mtamdi range from 380 to 60 microgram/person/day. For one substance [FLno: 4.03] the mtamdi is above the threshold of concern of 40 microgram/person/day. The mtamdi estimations of the three substances [FLno: 3.00, and 4.3] which are assigned to structural class III and are evaluated using the Procedure are between 0 and 380 microgram/person/day. These intake estimates are above the threshold of concern for structural class III substances of 90 microgram/person/day for all three substances. For comparison of the MSDI and mtamdivalues see Table 6.. Thus for four candidate substances [FLno: 3.00, 4.03, and 4.3] further information is required. This would include more reliable intake data and then, if required, additional toxicological data.

13 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 Table 6. Estimated intakes based on the MSDI approach and the mtamdi approach FLno EU Register name MSDI (µg/capita/day) mtamdi (µg/person/day) Structural class Threshold of concern (µg/person/day) 4.0 Ethyl nicotinate Class II Isopropyl nicotinate Class II Methylpyrrole Class II Acetylmethylpyrrole Class II Acetylpyridine Class II Butylpyridine Class II Butylpyridine Class II ,3Dimethylpyridine Class II ,4Dimethylpyridine Class II ,4Dimethylpyridine Class II ,Dimethylpyridine Class II ,Dimethylpyrrole Class II Ethylpyridine Class II Ethylpyridine Class II Hexylpyridine Class II Hydroxypyridine Class II Isopropylpyridine Class II Isopropylpyridine Class II Methylpyridine Class II Methylpyridine Class II Methylpyridine Class II Pentylpyridine Class II Propylpyridine Class II Pyrrolecarbaldehyde Class II ,4,6Trimethylpyridine Class II Acetylfurfurylpyrrole Class III Acetylindole Class III Methylindole Class III Methylquinoline 0. 0 Class III Butylquinoline Class III Methylquinoline Class III Considerations of Combined Intakes from Use as Flavouring Substances Because of structural similarities of candidate and supporting substances, it can be anticipated that many of the flavourings are metabolised through the same metabolic pathways and that the metabolites may affect the same target organs. Further, in case of combined exposure to structurally related flavourings, the pathways could be overloaded. Therefore, combined intake should be considered. As flavourings not included in this Flavouring Group Evaluation may also be metabolised through the same pathways, the combined intake estimates presented here are only preliminary. Currently, the combined intake estimates are only based on MSDI exposure estimates, although it is recognised that this may lead to underestimation of exposure. After completion of all FGEs, this issue should be readdressed. The total estimated combined daily per capita intake of structurally related flavourings is estimated by summing the MSDI for individual substances. n the basis of the reported annual production volumes in Europe (EFFA, 004m), the combined estimated daily per capita intake as flavourings of the candidate substances assigned to structural class II is microgram, which does not exceed the threshold of concern for a compound belonging to structural class II of 40 microgram/person/day. The combined daily per capita intake as 3

14 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 flavourings of the three candidate substances assigned to structural class III and evaluated using the Procedure is 0. microgram, which does not exceed the threshold of concern for a compound belonging to structural class III of 90 microgram/person/day. The 8 candidate substances, to which the Procedure has been applied, are structurally related to 9 supporting substances evaluated by JECFA at its 63 th meeting (JECFA, 006a). Based on reported production volumes, European per capita intakes (MSDI) could be estimated for 6 of the 9 supporting substances. Production volumes in Europe were not reported for three of the supporting substances [FLno: 4.04, 4.09 and 4.64]. The total combined intakes of the candidate and supporting substances are approximately 90 and 4 microgram/capita/day for structural class II and III, respectively, which do not exceed the thresholds of concern for a compound belonging to structural class II of 40 microgram/person/day and structural class III of 90 microgram/person/day, respectively. 8. Toxicity 8.. Acute Toxicity Data are available for seven candidate substances and for eleven structurally related substances evaluated by JECFA (JECFA, 006a). The LD 0 values range from 60 to 300 mg/kg bw. The acute toxicity data are summarised in Annex IV, Table IV Subacute, Subchronic, Chronic and Carcinogenicity Studies o data are available on subacute or subchronic oral toxicity on the candidate substances. Data on subacute or subchronic oral toxicity are available for six supporting substances, [FLno: 4.007, 4.047, 4.038, 3.34 and 4.04] and the structurally related indole3carbinol. For the candidate substance, 4methylquinoline [FLno: 4.00] data on liver carcinogenicity are available. For the supporting substances [FLno: 4.007, 4.00, 4.04 and 4.063] and 8methylquinoline there are data on chronic toxicity or carcinogenicity. 4Methylquinoline The candidate substance 4methylquinoline [FLno: 4.00], was tested for carcinogenicity in the liver of mice and rats. Mice were given 0. mg 4methylquinoline divided into three ip injections on day, 8 and after birth and the liver was examined for adenomas after one year. Liver adenomas were induced in 3 of 8 (8%) male mice, but not in female mice (=9). Male and female rats (= 6, 0) were given subcutaneous injections of 9 mg/kg bw first week after birth, followed by 4 mg/kg bw in weeks 7 and 9 mg/kg bw in week 8. The rats were examined for liver adenomas after 78 weeks. o increase in the number of liver adenomas was observed after exposure to 4methylquinoline (LaVoie et al., 988). The Panel concluded that these studies were not considered adequate for the evaluation of the oral toxicity of 4methylquinoline. Indole3carbinol In two studies, indole3carbinole was administered to male rats (8 animals/group) in the feed for 7 or 37 weeks at doses of 0 and 0 mg/kg bw/day. All tissues and gross lesions were subjected to histological examination and body and liver weights were measured. o differences between the groups of treated and control animals were detected. The only dose of 0 mg/kg bw has been taken as a AEL (Tanaka et al., 990; Tanaka et al., 99). 4

15 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 Acetylpyridine Acetylpyridine was administered to male and female rats (0 animals/sex/group) by gavage for 9 days at doses of 0, 37, 0, 330 and 000 mg/kg bw/day. Urine analysis, haematological and histological examinations were performed. Liver enlargement was observed in the two highest dosed groups. Slight anaemia was observed in females dosed 0 mg/kg and in males dosed 330 mg/kg. The AEL derived was 37 mg/kg bw/day (Til & van der Meulen, 97). Data are summarised in Annex IV, Table IV Developmental / Reproductive Toxicity Studies There are no data available on developmental or reproductive toxicity for the candidate or the supporting substances Genotoxicity Studies Genotoxicity data were provided for 0 of the 3 candidate substances: In in vitro studies on the candidate substances, methylpyrrole [FLno: 4.03], methylindole [FLno: 4.3], methylpyridine [FLno: 4.34], 3methylpyridine [FLno: 4.3], 4 methylpyridine [FLno: 4.36], 3,dimethylpyridine [FLno: 4.06] and 4acetylpyridine [FLno: 4.089] in doses up to 0000 microgram/plate, with and without metabolic activation, did not cause reverse mutations in various strains of (Table IV.4). Methylquinoline [FLno: 4.38] was reported weakly positive in two out of three studies in TA 00 with metabolic activation. The three studies were considered valid. egative results in single dose studies with methylquinoline were reported in TA 98, 00, 3, 37, 38 and TM677. The concentrations tested were not reported and thus, these results cannot be considered valid. Three valid studies on 4methylquinoline [FLno: 4.00] are available; two with positive results in the Ames test in strain TA 00 with metabolic activation and one with positive result in strain TA98 also with metabolic activation. This substance was also able to induce UDS in cultured rat hepatocytes. Studies on induction of aneuploidy in S. cerevisiae D6.M are available on five candidate substances. Methylpyrrolidine gave negative result while methylpyridine,,4dimethylpyridine,,6 dimethylpyridine and 4acetylpyridine gave positive results. The positive results were obtained at high doses inhibiting the growth of the yeast. Furthermore, fungal systems for measuring aneuploidy have little relevance compared to the mammalian system. o in vivo studies on genotoxicity of the candidate substances were available. Genotoxicity tests are available for nine supporting substances: In several valid studies, 6methylquinoline was consistently tested positive in the presence of S9 in several strains of, while negative results were obtained without metabolic activation. Acetylpyrrol (methyl pyrrolyl ketone) was positive in TA98 without metabolic activation at the two highest concentrations. egative results were obtained at the lowest concentration as well as with metabolic activation. This study is considered of limited relevance. Pyrrole, indole, 3methylindole (skatole), ethylpyridine and acetylpyridine were negative in bacterial mutation assays. In vivo data are available for two supporting substances.

16 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 3Methylindole (skatole) [FLno: 4.004] was reported negative in the micronucleus test in mice. The validity of this study, however, cannot be evaluated, as only an abstract is available. 6 Methylquinoline is reported negative in a test for gene mutations performed in Drosophila. 6 Methylquinoline is also reported negative in the micronucleus test in mice. The study design does, however, not meet current guidelines (PCE/CE ratio not reported), thus it is not clear whether the test substance reached the bone marrow. Positive results were obtained for some candidate and supporting substances in the Rec, DA breaking, CH and DA synthesis assays. These results are, however, not considered valid (Table IV.4). Data on genotoxicity are summarised in Table IV.4 and IV.. Conclusion on genotoxicity For the three quinoline derivatives methylquinoline, 4methylquinoline and 4butylquinoline the data available indicate a genotoxic potential. The Panel, therefore, concluded that the Procedure could not be applied to these three substances until adequate in vivo data become available. The genotoxicity data available for the remaining candidate substances do not preclude their evaluation through the Procedure. 9. Conclusions The present Flavouring Group Evaluation deals with 3 pyridine, pyrrole, indole and quinoline derivatives. one of the 3 candidate substances can exist as geometrical or optical isomers. Twentyfive of the flavouring substances are classified into structural class II and six are classified into structural class III. Twentyeight of the flavouring substances in the present group have been reported to occur naturally in a wide range of food items. According to the default MSDI approach, the 3 flavouring substances in this group have intakes in Europe from 0.00 to 0.73 microgram/capita/day which are all below the thresholds of concern for both structural class II (40 microgram/person/day) and structural class III (90 microgram/person/day) substances. Data on genotoxicity of the candidate substances are limited and the genotoxicity could not be adequately assessed. The data available, however, give rise to some concern of a genotoxic potential of the quinoline derivatives methylquinoline [FLno: 4.38], 4methylquinoline [FLno: 4.00] and 4butylquinoline [FLno: 4.094]. The Panel, therefore, concluded that the Procedure could not be applied to these three substances until adequate in vivo data become available. The genotoxicity data available for the remaining candidate substances do not preclude their evaluation through the Procedure. Two of the 3 candidate substances [FLno: 4.0 and 4.0] in this group are expected to be metabolised to innocuous products. For the remaining 9 candidate substances it cannot be anticipated that they will be metabolised to innocuous products. It was noted that where toxicity data were available they were consistent with the conclusions in the present flavouring group evaluation using the Procedure. 6

17 Flavouring Group Evaluation 4: Pyridine, pyrrole, indole and quinoline derivatives from chemical group 8 The EFSA Journal (006) 37 It was considered that on the basis of the default MSDI approach 3 of the 8 substances to which the Procedure could be applied would not give rise to safety concerns at the estimated levels of intake arising from their use as flavouring substances. For the remaining five candidate substances acetylfurfurylpyrrole [FLno: 3.00], methylpyrrole [FLno: 4.03], acetyl methylpyrrole [FLno: 4.08],,dimethylpyrrole [FLno: 4.07] and pyrrolecarbaldehyde [FLno: 4.4], a AEL could not be derived for the substance or a structurally related substance. Accordingly, additional data are required for these five flavouring substances. When the estimated intakes were based on the mtamdi approach they ranged from 380 to 60 microgram/person/day for the flavouring substances from structural class II. The intake was above the threshold of concern of 40 microgram/person/day for one [FLno: 4.03] of these candidate substances. For the three flavouring substances [FLno: 3.00, and 4.3] taken through the Procedure and assigned to structural class III, the estimated intakes based on the mtamdi are between 0 and 380 microgram/person/day. This is above the threshold of concern for structural class III of 90 microgram/person/day. The two flavouring substances, ethyl micotinate [FLno: 4.0] and isopropyl nicotinate [FLno: 4.0] which have mtamdi intake estimates below the threshold of concern for structural class II, are also expected to be metabolised to innocuous products. Thus for the four flavouring substances acetylfurfurylpyrrole [FLno: 3.00], methylpyrrole [FLno: 4.03], acetylindole [FLno: 4.088] and methylindole [FLno: 4.3] of the 3 flavouring substances considered in this Flavouring Group Evaluation, the intakes, estimated on the basis of the mtamdi, exceed the threshold for the structural class to which the flavouring substance has been assigned. Therefore, for these four substances more reliable exposure data are required. n the basis of such additional data, these flavouring substances should be reconsidered along the steps of the Procedure. Following this procedure additional toxicological data might become necessary. In order to determine whether this conclusion could be applied to the materials of commerce, it is necessary to consider the available specifications. Adequate specifications including complete purity criteria and identity tests for the materials of commerce have been provided for all 3 flavouring substances. For three of the 3 substances: methylquinoline [FLno: 4.38], 4methylquinoline [FLno: 4.00] and 4butylquinoline [FLno: 4.094] the Procedure could not be applied until adequate in vivo genotoxicity data were available and for five substances acetylfurfurylpyrrole [FLno: 3.00], methylpyrrole [FLno: 4.03], acetylmethylpyrrole [FLno: 4.08],,dimethylpyrrole [FLno: 4.07] and pyrrolecarbaldehyde [FLno: 4.4], additional toxicity data were required. The remaining 3 substances would present no safety concern at the levels of intake estimated on the basis of the MSDI approach. 7