Results of the public consultation ... - European Commission · requested; the Commission requested the SCENIHR to provide a list of priority substances (as foreseen in Art. 6 of

1

Results of the public consultation on SCENIHR's preliminary opinion on Additives used in tobacco products

(Opinion I)

A public consultation on this Opinion was opened on the website of the Commission Scientific committees from 22 July 2015 to 23 September 2015. Information about the public consultation was broadly communicated to national authorities, international organisations and other stakeholders. 37 different contributors (providing in total 266 comments) participated in the public consultation providing input to different chapters and subchapters of the Opinion. Majority of comments came from the tobacco industry, several came from public health authorities/institutes. 35 comments were marked as confidential (these comments and the replies of the SCENIHR do not appear in the table). The SCENIHR provided an individual reply to these contributors. Each submission was carefully considered by the SCENIHR and the scientific Opinion has been revised to take account of relevant comments. The literature has been accordingly updated with relevant publications. The SCENIHR thanks all contributors for their comments and for the references provided during the public consultation. The table below shows comments received on different chapters of the Opinion and the SCENIHR's response to them. It is also indicated if the comment resulted in a change of the Opinion.

2

Comments received during the public consultation on the SCENIHR preliminary opinion on "Additives used in tobacco products (Opinion I).

No. Name of

individual/ organisation

Table of content to

which comment

refers

Submission SCENIHRs response

1. Lutz Rolf, Philip Morris International Management SA, [email protected], Other

ABSTRACT PMI welcomes a science based approach for the assessment of tobacco additives and believes the enhanced testing and reporting obligations set forth by Art. 6 of the Tobacco Products Directive 2014/40/EU (TPD) should lead to an EU harmonized evidence-based regulation of the use of additives. We kindly request to obtain as soon as possible guidance on the types of tests manufacturers will have to perform and report by November 2017 for the additives in the priority list, to be able to adequately plan the manufacturing of testing articles and testing to be done. PMI furthermore recommends a public consultation on SCENIHR’s Opinion 2 with sufficient time for a thorough analysis. We are concerned that SCENIHR’s Preliminary Opinion includes definitive statements on addictiveness, stating that “14 substances were selected because of their contribution to addictiveness to smoking due to different mechanisms”. (p.5, l.12). This statement is not supported by evidence and should be replaced with “their potential/suspected contribution”. Moreover, it surprises us that the Committee considers itself in a position to reach definitive statements on toxicity of certain additives when it recognizes that “time constraints only allowed a limited toxicological evaluation of the large number of compounds” (p.4, l.31-33).

The mandate of the Commission concerning Tobacco Additives has been divided in two parts, corresponding to 2 Opinions. Guidance on the types of tests to perform will be the main subject of the second Opinion, as clearly indicated in the Terms of Reference. Issues related to addictiveness have been clarified within the text, which has been modified accordingly. It is clear in the Terms of Reference that an exhaustive toxicological evaluation was not requested; the Commission requested the SCENIHR to provide a list of priority substances (as foreseen in Art. 6 of the TPD). The SCENIHR applied the criteria (as already described in the Preliminary Opinion) identifying the hazard(s) related to single additives, and only for transparency reasons added the information that a thorough

3

The Committee goes beyond the TPD and its Terms of References (ToR) when examining whether additives increase “attractiveness” (p.4, l.9 and l.28, p.5, l.2 and l.15). “Attractiveness” is not a relevant criteria under TPD. Neither SCENIHR nor the Commission is competent to amend the TPD. We expect that SCENIHR will respect this legal framework and will delete “attractiveness” throughout the document. We are also concerned that SCENIHR did not follow its own April 2013 Rules of Procedure’s principles of excellence, independence and transparency as well as the principles and standards for scientific advice on risk assessment. In particular, SCENIHR did not disclose how the literature search (p.4, l.26) was conducted and on which basis literature was considered relevant or dismissed. We note that SCENIHR relied extensively on secondary sources of information, while many peer reviewed publications from the industry were not considered. In particular, the Committee relies on the PITOC factsheets which were designed to inform professionals and the public and not as a thorough scientific evidence review. In addition, it is a concern that this key source of information was produced by its own experts (DKFZ and RIVM coordinated PITOC) and not, as asserted (p.4, I.35),by different groups.

toxicological/risk evaluation was not conducted. Additives ‘resulting in a characterising flavour’ (which can potentially contribute to attractiveness) are included in the SCENIHR mandate and were assessed by the SCENIHR accordingly. However, for clarification the text has been aligned with the terminology used in the mandate. In the methodology paragraph of the preliminary Opinion, the procedure followed has been described (obviously not detailed in the abstract). There it is clearly stated that the weight of evidence approach has been followed where appropriate: not being a risk assessment but a prioritization based on hazard of a large number of additives. Secondary sources of information used as evaluations carried out by other Agencies (i.e. EFSA or EPA) were considered fully reliable. The same for the additives already evaluated by the PITOC project: the information PITOC included in the fact-sheet was the result of an extensive literature search and evaluation up to 2012. More recent data was added (literature from 2012 to 2015-analysed by expert not involved in the PITOC project), increasing judgement independency. The methodology paragraph has been amended for clarification purposes. The selection of additives have been carried out regardless PITOC results. After they were selected, the factsheets were used. Therefore what has been ‘asserted’ was exactly what the

4

Had the Committee carried out a comprehensive review of all the evidence, it would have realized that most of the “gaps” it identified are in fact not gaps in the current state of science but in its literature research. Prior to requesting additional testing from manufacturers, it is essential to have completely assessed existing data and evidence. In particular, SCENIHR would have realized that,contrary to its statements, inhalation toxicity data (p.4, l.43),data on pyrolysis and exposure to combustion reactions products (p.4, l.45) and data on mixture toxicity (P.5, l.2) are not “scarce” or “negligible” but have been reported in peer reviewed publications not yet considered by SCENIHR, which we upload in the corresponding sections. Finally, SCENIHR’s grouping of substances is not done according to the ToR allowing it to group only “substances with identical or very similar properties”. The heterogeneity of the groups leaves unlimited scope to arbitrarily include additional substances which does not allow anypredictability and planning regarding the testing and use of additives by manufacturers. All 48 priority additives should be listed individually.

SCENIHR did. The current Opinion serves, as stipulated in the methodology, the compilation of a priority list. This list will assist, in line with Article 6 of Directive 2014/40/EU, the Commission to develop priority list of at least 15 additives for which enhanced reporting obligations will apply (as described in the section 1 ‘background’). The SCENIHR is therefore not in the position to request any additional testing. Provided publications have been looked at. Since no additional relevant information was found in relation to the Terms of Reference, no change in the text was deemed necessary. The grouping has been carried out considering structurally similar chemicals used as additives, which are supposed to show similar properties (with a focus on those relevant for this Opinion). This was included in the Terms of Reference.

2. Valenta Petr, Czech Chamber of Commerce, [email protected], Czech Republic

ABSTRACT In overall, we would like to highlight our conviction that the Report presents conclusive statements without required level of expertise. The Chamber believes that any regulation on EU Internal Market should be evidence-based. Regarding the SCENIHR Report, we would like to express our worries that the Report conclusions are not based on clear evidence. The Report admits this significant fact itself while stating (page 82) “there are many data gaps”. This challenges the Report in general, and puts in risk its credibility in particular when it comes to matters such as toxicological profiles, inhalation toxicology and exposure to the additive. Regulation in these areas has to be supported by serious and unchallengeable facts.

Any Regulation on EU Internal Market is out of the mandate of the SCENIHR, which was only asked to provide a priority list of additives. For these priority additives enhanced reporting obligations will apply in the form of comprehensive studies which shall examine for each additive whether it has any of the properties specified in the Opinion. Therefore the compilation of the priority list is not to be considered a risk assessment of the selected compounds. Consequently, the presence of data gaps cannot challenge the

5

Therefore we have the serious concern whether SCENIHR has carried out a thorough scientific review of the evidence. Finally, the Chamber of Commerce would call on the European Commission to issue the guidance on the manufacturers´ testing soon. In general, this mandatory step is important when it comes to stable and predictable business environment. In addition, European Commission should submit the guidance for discussion in advance. It is necessary for the manufacturers and involved third parties to allow to plan their business strategies with concern about the testing.

Report. The mandate of the Commission concerning tobacco additives has been divided in two parts, corresponding to 2 Opinions. Guidance on the types of tests to perform will be the main topic of the second Opinion, as clearly indicated in the Terms of Reference. As usual, the Preliminary Opinion will be the subject of public consultation.

3. Karin Molander Gregory, Folkhälsomyndigheten Public Health Agency of Sweden, [email protected], Sweden

ABSTRACT We welcome the preliminary Opinion on Additives used in tobacco products by the Scientific Committee on Emerging and Newly Identified Health Risks (SCENIHR). We especially welcome that the preliminary opinion highlight data gap and lack of pyrolysis studies on additives which are regarded safe when used in food products that their safety as a tobacco additive, where the additive is burnt and inhaled is not assured in the same way, as if used in food products. With regards to data on which the selection of additives are made we want to clarify that the ingredients reported by the Industry to us in our role of a Supervisory Agency are at the moment taken by face value and no further analysis of the correctness of the amount of additives in the products are made. Therefore we very much welcome the described second opinion by SCENIHR addressing guidelines for the type and criteria for studies that should be requested from manufactures to assess the relevance of individual additives. Finally regarding the priority list of at least 15 additives we suggest that additives are rather described and grouped due to function, mainly on a generic basis, to enable additives with similar functionality to be treated in the same way.

The SCENIHR thanks for the comment. Regarding a different ‘grouping’ based on their function as additives, since many of them share more than one characteristics, it was considered not feasible.

4. Rapota Maria, JT International S.A., 1, Rue de la Gabelle, 1211 Geneva 26, [email protected]

ABSTRACT JTI has reviewed SCENIHR’s “Preliminary Opinion on Additives used in tobacco products” (Opinion) and would like to make the following points with regard to the abstract: Page 4 (lines 1 – 13) As a general comment, we note that the Preliminary Opinion contains a number of

As specified in point 1 of the Terms of Reference (see chapter 2 of the Opinion) the SCENIHR has been asked by Commission services to identify, based on scientific evidence and other relevant information

6

m, Other findings to the effect that various substances do in fact exhibit CMR properties, facilitate inhalation or nicotine uptake, result in a characterizing flavor and/or contribute to toxicity or “addictiveness”. We disagree with many of these findings (as more specifically detailed elsewhere in this response) and note that it is neither SCENIHR nor the Commission’s role under TPD2 to identify whether additives in fact have these properties. Article 6(1) of TPD2 requires the Commission to identify additives of potential concern, and then for manufacturers to conduct comprehensive studies to confirm whether or not these additives exhibit any of these properties under conditions of use. It is not for the Commission or SCENIHR to pre-judge the issue, particularly in circumstances where SCENIHR’s evaluation of the evidence is (as detailed elsewhere) far from comprehensive. Page 4 (lines 24-30) The criteria that were used for the literature search are not described and the resulting report is not a systematic review as many relevant published studies were not considered in the report. Page 4 (lines 37-39, lines 40-47), Page 5 (lines 1-5) Many relevant peer-reviewed and publicly available studies on inhalation toxicity, pyrolysis and testing additives in mixtures were not considered in this report. As part of the company’s Product Stewardship program, JTI has commissioned numerous chemical and toxicological studies for the assessment of additives under conditions of use. The results of these studies have also been submitted to EU Member States under TPD1. Page 5 (lines 6-9) SCENIHR has identified too many additives (48 instead of the 20-30 which is the number mentioned in the mandate SCENHIR received from the Commission) as possible candidates for the priority list on a questionable scientific basis. Furthermore, additives should be listed as single entries,

currently available, those additives that fall/are suspected to fall within the scope of certain categories. The SCENIHR has followed the request from the Commission accordingly. However, the text has been adjusted for clarification. The SCENIHR agrees that any Regulation is out of its mandate. In this case the SC was asked to provide a priority list of additives based on their known/potential hazard (not to cary out a risk assessment, not to pre-judge). For these priority additives enhanced reporting obligations will apply (as described in section 1 ‘background’) In the methodology paragraph ot he preliminary Opinion, indication on the procedure followed has been described (not possible to give all details within the abstract). Provided publications have been looked at and since no additional relevant information was found in relation to the Terms of Reference, no change in the text was deemed necessary. The SCENIHR has prepared a list of 30 additives/ groups of additives as requested by the Commission, based on discribed criteria. It is up to the Commission to decide how to proceed.

7

not in groups. Page 5 (lines 12-13) The wording in this section suggests that SCENHIR has prejudged the issue of whether the 14 selected substances contribute to the “addictiveness” of tobacco products. As already noted, this is an issue to be determined by way of comprehensive studies that are to be conducted in due course by tobacco product manufacturers. However, the existing scientific evidence suggests that additives have no influence on the “addictiveness” of tobacco products. In its November 2010 report on ‘The addictiveness and attractiveness of tobacco additives’ SCENIHR concluded that there is no adequate method to test the effects of additives on addictiveness. Page 5 (lines 18-21) TPD2 does not allow for prioritization within the first priority list. Once the Commission adopts the list, all of the additives contained on the list have to be tested by the deadline. Therefore, it is not really clear what SCENHIR is suggesting here. Further, the fact that all additives that the Commission decides to include on the priority list will be the subject of “comprehensive studies” by industry is a key reason why it is so concerning that SCENIHR has recommended such a large number of additives for inclusion on that list. The testing burden on tobacco manufacturers would be overwhelming, if so many additives were included.

Grouping of structurally similar additives, showing similar characteristics was foreseen, and indeed in these cases the Commission asked for a prioritization among the group (done whewnever possible). Therefore the SCENIHR did not go beyond its Terms of Reference. Additives ‘resulting in a characterising flavour’ (which can potentially contribute to attractiveness) as well as additives which ‘facilitates inhalation or nicotine uptake’ potentially contributing to the addictiveness are included in the Terms of Reference. This issue has been reworded in the final Opinion in order to increase the clarity of the text.

5. AURIEMMA LUIGI, ONT ITALIA SCARL, [email protected], Italy

ABSTRACT Lo studio SCHENIR sugli additivi necessita a nostro parere di alcuni approfondimenti, in particolare si rileva (da pagina 81 del documento) che è stata adottata una metodologia fondata su dati parziali. Ciò è da imputare secondo lo stesso parere della Commissione ad un tempo limitato per le analisi (pagina 82, riga 21) ed a dati parziali (pagina 14, riga 16), ciò non ha consentito di basare le conclusioni su solide basi scientifiche. Si rileva inoltre che il documento prende in considerazione ripetutamente il concetto dell’”attrattività” dei prodotti del tabacco benché la Direttiva 2014/40/UE non avesse mai richiesto un esame circa l’”attrattività” determinata dall'utilizzo di additivi.

The SCENIHR was asked to provide a priority list of additives, which has not to be considered a ‘full’ risk assessment of the selected compounds, but identification of hazard properties. This was clearly specified in the preliminary Opinion. The list has been build up based on sound scientific criteria, described in the Opinion. As specified in point 1 of the Terms of Reference (see chapter 2 of the Opinion) The SCENIHR has been asked by Commission services to identify, based on scientific evidence and other relevant information

8

currently available, those additives that fall/are suspected to fall within the scope of certain categories. Additives ‘resulting in a characterising flavour’ (which can potentially contribute to attractiveness) are included in the SCENIHR mandate. The SCENIHR has followed the request from the Commission accordingly. This issue has been reworded within the text in order to increase the clarity of the text.

6. Wilkinson Peter, Imperial Tobacco Limited, [email protected], United Kingdom

ABSTRACT We welcome the opportunity to comment on this Preliminary Opinion. As a leading, global manufacturer of tobacco products, we have substantial expertise in tobacco product additives, including a team of highly experienced toxicologists and laboratory personnel focusing on the science behind the product. We believe that this experience and product knowledge can make a useful contribution to the wider understanding of additives used in tobacco products. The abstract refers to the mandate given to SCENIHR. The Preliminary Opinion document does not correctly cite the mandate as the following words have been added: - to point b. (line 9): 'potentially increasing attractiveness' - to point c. (line 10): '(increasing addictiveness)' We note that these additions go beyond the mandate and suggest using the original wording. In particular, the addition to point b. introduces a term not used in the provisions of Art. 6 of Directive 2014/40/EU. The abstract as well as other parts of the Preliminary Opinion refer to "attractiveness". SCENIHR (2010) addressed the substantial limitations in assessing or measuring the "attractiveness" of a tobacco additive by highlighting an absence of suitable animal models and ethical concerns should human testing be involved. It is fair to conclude from these limitations that no appropriate or validated methodology exists for measuring “attractiveness” of a tobacco additive. Consequently, from a scientific perspective, it is not

The issue has been addressed and the abstract was re-worded accordingly. As specified in point 1 of the Terms of Reference (see chapter 2 of the Opinion) The SCENIHR has been asked by Commission services to identify, based on scientific evidence and other relevant information currently available, those additives that fall/are suspected to fall within the scope of certain categories. Additives ‘resulting in a characterising flavour’ (which can potentially contribute to attractiveness) are included in the SCENIHR mandate. The SCENIHR has

9

relevant for the exercise of defining a list of priority additives. Moreover, the term is not within the mandate given to SCENIHR since "Attractiveness" per se fails established criteria for issue definition: it is lacking in any evidential foundation and is inherently uncertain and arbitrary within a scientific context. We recognise the comments of SCENIHR that time was a limiting factor to its evaluation. None-the-less, we note that a number of relevant data have not been cited in the Preliminary Opinion. We respectfully provide responses to the Preliminary Opinion as well as supplemental data to inform SCENIHR’s knowledge set in our submission. Please refer to the file attached to this section (“2015-08-20 Section ABSTRACT Final.PDF”) for more detailed comments on the text of this section.

followed the request from the Commission accordingly. This issue has been reworded within the text in order to increase the clarity of the text. The characterising flavour was only one of the criteria used. Provided publications have been looked at and since no additional relevant information was found in relation to the Terms of Reference, no change in the text was deemed necessary.

7. Chýlková Markéta, Federation of the Food and Drink Industries, [email protected], Czech Republic

ABSTRACT The Food Chamber closely followed the entire process of the TPD negotiations at the European level in 2013. When subjecting the SCENIHR report to a review, we have discovered that its focus goes well beyond the agreed format of the Directive in examination whether the additives concerned increase attractiveness of products. The concept of attractiveness is not a relevant criteria under the TPD and we are of the opinion that the Preliminary Opinion should not go beyond the scope of the agreed format under the TPD otherwise its credibility could be significantly undermined. As the principal foodstuff association we would like to emphasize that it is very important for the manufacturers that the European Commission provides them at the earliest opportunity with the specification of tests they will be obliged to submit as part of the new reporting requirements. The timely provision of these instructions is key to an accurate business planning process.

As specified in point 1 of the Terms of Reference (see chapter 2 of the Opinion) the SCENIHR has been asked by Commission services to identify, based on scientific evidence and other relevant information currently available, those additives that fall/are suspected to fall within the scope of certain categories. Additives ‘resulting in a characterising flavour’ (which can potentially contribute to attractiveness) are included in the SCENIHR mandate. The SCENIHR has followed the request from the Commission accordingly. This issue has been reworded within the text in order to increase the clarity of the text. The mandate of the Commission concerning tobacco additives has been divided in two parts, corresponding to 2 Opinions. Guidance on the types of tests to perform will be the subject of the second Opinion, as clearly indicated in the Terms of Reference.

10

8. MASIELLO GENNARINO, COLDIRETTI (PRESIDENTE REGIONE CAMPANIA), [email protected], Italy

ABSTRACT L’opinione preliminare di SCENIHR sugli additivi utilizzati nei prodotti del tabacco presenta, in alcuni punti, un approccio poco scientifico e, peraltro, giunge a fare considerazioni troppo generiche sugli additivi da includere nella lista prioritaria. In particolare, il sommario dell’opinione di SCENIHR chiarisce che, a causa del tempo limitato a disposizione, si è potuta fare solo una limitata valutazione tossicologica di un largo numero di composti, tanto che si parla di dati “scarsi” e “trascurabili” su cui si è lavorato. Ancora il documento parla di “lacune” nei dati (“Vi sono molte lacune nei dati sugli additivi”), per cui la nostra organizzazione ritiene che l’analisi di SCENIHR sia incompleta e non possa rappresentare una base per richiedere ulteriori test sugli additivi. Inoltre l’opinione fa riferimento in più punti all’ “attrattività” dei prodotti del tabacco, che aumenterebbe a causa dell’aggiunta di certi additivi: in questo specifico caso, va fatto notare che l’opinione di SCENIHR si spinge oltre quanto richiesto dalla Direttiva 2014/40/UE, il cui articolo 6 non fa riferimento al concetto di “attrattività”.

The SCENIHR was asked to provide a priority list of additives, which has not to be considered a ‘full’ risk assessment of the selected compounds. This was clearly specified in the preliminary Opinion. The list has been build up based on sound scientific criteria, although not analysed in such details to carry out a risk assessment. The presence of data gaps cannot therefore challenge the Report; if data gaps are there they will be covered once the Commission will decide about the data requirement. As specified in point 1 of the Terms of Reference (see chapter 2 of the Opinion) the SCENIHR has been asked by Commission services to identify, based on scientific evidence and other relevant information currently available, those additives that fall/are suspected to fall within the scope of certain categories. Additives ‘resulting in a characterising flavour’ (which can potentially contribute to attractiveness) are included in the SCENIHR mandate. The SCENIHR has followed the request from the Commission accordingly. This issue has been reworded within the text in order to increase the clarity of the text.

9. Ureel Ludwig, British American Tobacco, [email protected], United Kingdom

ABSTRACT We understand the purpose of the scientific opinion was to undertake a comprehensive assessment to identify additives that would assist the Commission in establishing a priority list for cigarettes and RYO. Therefore, we are surprised to note that SCENIHR acknowledge that, ‘due to time constraints’, it has only been able to conduct a limited toxicological evaluation of the additives.

The aim of the Opinion was not to undertake a comprehensive risk asessement. It was clearly specified in the Terms of Reference and in the preliminary Opinion. The SCENIHR was asked to provide a priority list of additives, which has not to be considered as a ‘full’ risk assessment of the selected compounds, but as hazard identification. The list has been build up based on sound scientific criteria, although not analysed in such details to carry out a risk

11

Given that SCENIHR’s external experts were representatives of RIVM and DKFZ - the two external bodies responsible for the PITOC review - it is concerning that SCENIHR, instead of undertaking a comprehensive evaluation of the evidence itself, appears to have placed significant reliance on PITOC. Such reliance is evidenced by SCENIHR’s failure to acknowledge the large body of peer reviewed papers that predate PITOC (and which was not considered in PITOC) as well as publications released subsequently. The fact that SCENIHR’s external experts were representatives of these entities affects their ability to objectively assess PITOC and its conclusions. Under these circumstances, we are concerned that the assessment does not comply the rules of procedure under which SCENIHR should operate(1). We also question the appropriateness of SCENIHR’s reliance on the list of single Member State as a starting point for compiling the list of priority additives. Therefore, we consider that SCENIHR’s assessment and application of ‘weight of evidence approach’ is fundamentally flawed. By way of example of SCENIHR’s failure to consider relevant evidence, the tobacco industry has generated a substantial amount of test data on additives under conditions of use, singly and in combination. Despite assertions to the contrary in the Opinion, data on the pyrolysis of additives are available (2,3). These publications demonstrate that a large proportion of additives transfer with minimal breakdown and without conversion to toxic compounds.

assessment. The selection of additives for the priority list has been carried out regardless of PITOC results; then once the compounds were selected, the factsheets for PITOC additives were used. PITOC is an EU funded project, and as such it is considered as a reliable source of information. The Commisson included PITOC results within the available information. The substances on the priority list were selected irrespectively of PITOC. The PITOC evaluation was taken on board afterwards, since the information included in the fact-sheet was the results of an extensive literature search and evaluation. The information on PITOC componds was updated by recent data (literature from 2012 to 2015). It was analysed by an expert not involved in PITOC to increase judgement independency. The methodology paragraph has been amended for clarification purposes. In the methodology paragraph, indication on the procedure followed has been described (not possible to give such a detail within the abstract). There it is clearly stated that the weight of evidence approach has been followed for some aspects. Regarding the use of a list of a single member state, it is already indicated in the preliminary Opinion: To compile the list of priority substances, the SCENIHR considered several lists. Some lists of additives were obtained from the public domain which were published by several Member States including Belgium, the Czech Republic, Germany and the Netherlands ; an additional list was received from the UK authorities; and some were taken

12

Furthermore, SCENIHR claims that “no information was available on possible mixture toxicity”. However, BAT has published papers in peer reviewed journals (4-6) in which mixtures of additives were tested, the results of which are consistent with other extensive data sets available in the public domain. We acknowledge that there is sometimes little information available on the toxicity of additives per se. However, this situation exists in other industries and, for example, the Joint FAO/WHO Expert Committee on Food Additives (JECFA), acknowledges that "it is neither possible nor necessary to conduct toxicological studies on all individual flavouring substances used in food. The vast majority of flavouring substances are members of groups of substances with common metabolic pathways, and typically, individual members of such a group display a similar toxicity profile" (26). Hence, JECFA introduced a “Procedure for the Safety Evaluation of Flavouring Agents” which does not mandate a full risk assessment of individual additives if no toxicological data exists, but instead recommends the assessment of groups of structurally

from other jurisdictions (e.g., USA, Canada, and Brazil). However, except for the Netherlands, the comprehensive list of ingredients that was submitted to regulators in all Member States was not available. Therefore, due to time constraints, the list from the Netherlands. containing 1260 compounds, was used as a typical example. The text has been partially amended to clarify this point. The actual ‘use and frequency’ on a EU-wide basis (not available to the SC) is an issue to be considered by the Commission during preparation of the legislation. The only available information is related to specific ‘blend’ and therefore not useful for this Opinion. Being unfeasible to carry out studies on the exaggerated number of possible mixture, the only possibility is to use the component-based approach to evaluate independent action (absence of interaction), additiveness or interactions (see SC opinion on mixtures). The SCENIHR is fully aware of the JECFA approach (used also by CEF Panel in EFSA), as well as of the grouping approach in REACH. However, this is otside the scope of this first

13

related flavour compounds (26). Similarly, under REACH there is no requirement to undertake studies to identify specific toxicological end-points on individual additives. Instead, REACH allows for the extrapolation of toxicological data. For example, a short term feeding study is used to determine a Derived No Effect Level and deemed to be safe in chronic inhalation exposure (27). This is a common practice, known as ‘read across’, and as such is aligned with the requirements for replacement, refinement and reduction of the use of animals for toxicological testing. This suggests that, when only limited information on toxicity is available, unnecessary testing can be avoided as described in the above approaches.

Opinion and it will be considered in the second Opinion which will address testing strategy definition.

10. SOSSI FRANCESCO, COLDIRETTI (FEDERAZIONE PROVINCIALE BENEVENTO), [email protected], Italy

ABSTRACT L’opinione preliminare di SCENIHR sugli additivi utilizzati nei prodotti del tabacco presenta, in alcuni punti, un approccio poco scientifico e, peraltro, giunge a fare considerazioni troppo generiche sugli additivi da includere nella lista prioritaria. In particolare, il sommario dell’opinione di SCENIHR chiarisce che, a causa del tempo limitato a disposizione, si è potuta fare solo una limitata valutazione tossicologica di un largo numero di composti, tanto che si parla di dati “scarsi” e “trascurabili” su cui si è lavorato. Ancora il documento parla di “lacune” nei dati (“Vi sono molte lacune nei dati sugli additivi”), per cui la nostra organizzazione ritiene che l’analisi di SCENIHR sia incompleta e non possa rappresentare una base per richiedere ulteriori test sugli additivi. Inoltre l’opinione fa riferimento in più punti all’ “attrattività” dei prodotti del tabacco, che aumenterebbe a causa dell’aggiunta di certi additivi: in questo specifico caso, va fatto notare che l’opinione di SCENIHR si spinge oltre quanto richiesto dalla Direttiva 2014/40/UE, il cui articolo 6 non fa riferimento al concetto di “attrattività”.

The SCENIHR was asked to provide a priority list of additives, which has not to be considered as ‘full’ risk assessment of the selected compounds but hazard identification. This was clearly specified in the preliminary Opinion. The list has been build up based on sound scientific criteria, although not analysed to carry out a risk assessment. The presence of data gaps cannot therefore challenge the Report; if data gaps are there, they will be covered once the Commission will decide about the data requirement. Additives ‘resulting in a characterising flavour’ (which can potentially contribute to attractiveness) as well as additives which ‘facilitates inhalation or nicotine uptake’ are included in the SCENIHR Terms of Reference. Therefore the SCENIHR followed exactly what is requested by the Commission. This issue has been reworded to increase clarity of the text.

14


ABSTRACT We understand the purpose of the scientific opinion was to undertake a comprehensive assessment to identify additives that would assist the Commission in establishing a priority list for cigarettes and RYO. Therefore, we are surprised to note that SCENIHR acknowledge that, ‘due to time constraints’, it has only been able to conduct a limited toxicological evaluation of the additives. Given that SCENIHR’s external experts were representatives of RIVM and DKFZ - the two external bodies responsible for the PITOC review - it is concerning that SCENIHR, instead of undertaking a comprehensive evaluation of the evidence itself, appears to have placed significant reliance on PITOC. Such reliance is evidenced by SCENIHR’s failure to acknowledge the large body of peer reviewed papers that predate PITOC (and which was not considered in PITOC) as well as publications released subsequently. The fact that SCENIHR’s external experts were representatives of these entities affects their ability to objectively assess PITOC and its conclusions. Under these circumstances, we are concerned that the assessment does not comply the rules of procedure under which SCENIHR should operate(1).

The aim of the Opinion was not undertake a comprehensive risk asessement. It was clearly specified in the Terms of Reference and within the text of the preliminary Opinion. The SCENIHR was asked to provide a priority list of additives, which has not to be considered as risk assessment of the selected compounds but as hazard identification. The list has been build up based on sound scientific criteria described in the Opinion. The selection of additives for the priority list has been carried out regardless of PITOC results; then once the compounds were selected, the factsheets for PITOC additives were used. PITOC is an EU funded project, and as such it is considered as a reliable source of information. The Commisson included PITOC results as ‘supporting’ information. The additives were selected irrespectively of PITOC, then the evaluation was taken on board, since the information included in the fact-sheet was the results of an extensive literature search and evaluation. In addition the results of the literature search carried out to update information from 2012 to 2015 were analysed by the SCENIHR members not previously involved in PITOC to increase judgement independency. The methodology paragraph has been amended for clarification purposes. In the methodology paragraph ot he preliminary Opinion, indication on the procedure followed has been described (not possible to give such a detail within the abstract). There it is clearly stated that the

15

We also question the appropriateness of SCENIHR’s reliance on the list of single Member State as a starting point for compiling the list of priority additives. In light of the above, we consider that SCENIHR’s assessment and application of ‘weight of evidence approach’ is fundamentally flawed. By way of example of SCENIHR’s failure to consider relevant evidence, the tobacco industry has generated a substantial amount of test data on additives under conditions of use, singly and in combination. Despite assertions to the contrary in the Opinion, data on the pyrolysis of additives are available (2,3). These publications demonstrate that a large proportion of additives transfer with minimal breakdown and without conversion to toxic compounds. Furthermore, SCENIHR claims that “no information was available on possible mixture toxicity”. However, BAT has published papers in peer reviewed journals (4-6) in which mixtures of additives were tested, the results of which are consistent with other extensive data sets available in the public domain. We acknowledge that there is sometimes little information available on the toxicity of additives per se.

weight of evidence approach has been followed for some aspects. Regarding the use of a list of a single member states, it is already indicated in the preliminary opinion (page 15) To compile the list of priority substances, SCENIHR considered several lists. Some lists of additives were obtained from the public domain which were published by several Member States including Belgium, the Czech Republic, Germany and the Netherlands ; an additional list was received from the UK authorities; and some were taken from other jurisdictions (e.g., USA, Canada, and Brazil). However, except for the Netherlands, the comprehensive list of ingredients that was submitted to regulators in all Member States was not available. Therefore, due to time constraints, the list from the Netherlands containing 1260 compounds, was used as a typical example. The text has been partially amended to clarify this point. The actual ‘use and frequency’ on an EU-wide basis (not available to the SCENIHR) is an issue to be considered by the Commission during preparation of the legislation. The available information on ‘mixture toxicity’ is related to specific ‘blend’ and therefore not useful for this opinion. Being unfeasible to carry out studies on the almost ‘infinite’ number of possible mixture, the only possibility is to use the component-based approach to evaluate independent action (absence of interaction), additiveness or interactions (see SC opinion on mixtures).

16

However, this situation exists in other industries and, for example, the Joint FAO/WHO Expert Committee on Food Additives (JECFA), acknowledges that "it is neither possible nor necessary to conduct toxicological studies on all individual flavouring substances used in food. The vast majority of flavouring substances are members of groups of substances with common metabolic pathways, and typically, individual members of such a group display a similar toxicity profile" (26). Hence, JECFA introduced a “Procedure for the Safety Evaluation of Flavouring Agents” which does not mandate a full risk assessment of individual additives if no toxicological data exists, but instead recommends the assessment of groups of structurally related flavour compounds (26). Similarly, under REACH there is no requirement to undertake studies to identify specific toxicological end-points on individual additives. Instead, REACH allows for the extrapolation of toxicological data. For example, a short term feeding study is used to determine a Derived No Effect Level and deemed to be safe in chronic inhalation exposure (27). This is a common practice, known as ‘read across’, and as such is aligned with the requirements for replacement, refinement and reduction of the use of animals for toxicological testing. This suggests that, when only limited information on toxicity is available, unnecessary testing can be avoided as described in the above approaches.

The SCENIHR is fully aware of the JECFA approach (used also by CEF Panel in EFSA), as well as of the grouping approach in REACH. However, this is outside the scope of this first Opinion. It will be considered in the second Opinion which will address testing strategy definition.

12. Federici Giulio, Coldiretti Caserta, [email protected], Italy

ABSTRACT L’opinione preliminare di SCENIHR sugli additivi utilizzati nei prodotti del tabacco presenta, in alcuni punti, un approccio poco scientifico e, peraltro, giunge a fare considerazioni troppo generiche sugli additivi da includere nella lista prioritaria. In particolare, il sommario dell’opinione di SCENIHR chiarisce che, a causa del tempo limitato a disposizione, si è potuta fare solo una limitata valutazione tossicologica di un largo numero di composti, tanto che si parla di dati “scarsi” e “trascurabili” su cui si è lavorato. Ancora il documento parla di “lacune” nei dati (“Vi sono molte lacune nei dati sugli additivi”), per cui la nostra organizzazione ritiene che l’analisi di SCENIHR sia

The SCENIHR was asked to provide a priority list of additives, which has not to be considered as ‘full’ risk assessment of the selected compounds, but as hazard identification. This was already specified in the preliminary Opinion. The list has been build up based on sound scientific criteria. In addition data on exposure were not available on an EU wide basis, therefore a risk assessment was not possible at all. The presence of data gaps cannot therefore

17

incompleta e non possa rappresentare una base per richiedere ulteriori test sugli additivi. Inoltre l’opinione fa riferimento in più punti all’ “attrattività” dei prodotti del tabacco, che aumenterebbe a causa dell’aggiunta di certi additivi: in questo specifico caso, va fatto notare che l’opinione di SCENIHR si spinge oltre quanto richiesto dalla Direttiva 2014/40/UE, il cui articolo 6 non fa riferimento al concetto di “attrattività”.

challenge the Report; if data gaps are there they will be covered once the Commission will decide about the data requirement.

13. MELENEKLIS VASSILIOS, HELLENIC INTERPROFESSIONAL ORGANIZATION OF RAW TOBACCO, [email protected], Greece

ABSTRACT We are concerned about SCENIHR’s approach and its far-reaching conclusions regarding certain additives included in the priority list of additives. Also, the paper does not comply with the Tobacco Products Directive. The most indicative example is its examination of whether additives increase “attractiveness”. “Attractiveness” is a notion that is irrelevant to the Directive’s scope in terms of criteria setting. The paper’s approach practically amends the Directive, something that neither SCENIHR nor the Commission is competent to do. Finally, the paper has not been subject to a thorough scientific review of the evidence that presents.

The SCENIHR was asked to provide a priority list of additives, which is not to be considered as ‘full’ risk assessment of the selected compounds but as hazard identification. This was already specified in the preliminary Opinion. The list has been build up based on sound scientific criteria. The presence of data gaps cannot therefore challenge the Report; if data gaps are there they will be covered once the Commission will decide about the data requirement. As specified in point 1 of the Terms of Reference (see chapter 2 of the Opinion) the SCENIHR has been asked by Commission services to identify, based on scientific evidence and other relevant information currently available, those additives that fall/are suspected to fall within the scope of certain categories. Additives ‘resulting in a characterising flavour’ (which can potentially contribute to attractiveness) are included in the SCENIHR mandate. The SCENIHR has followed the request from the Commission accordingly. This issue has been reworded within the text in order to increase the clarity of the text.

14. Lutz Rolf, Philip Morris International Management SA, [email protected],

3 SCIENTIFIC RATIONALE

The section at p.12, l.3-13, should mention the distinction between blended cigarettes, which use additives, and Virginia cigarettes, which use no or very few tobacco additives (see e.g. Lee et al, 2012; Sanders et al, 2012;

The distinction between different tobacco products is outside the Terms of Reference for the SCENIHR, which was asked to provide a priority list of additives used in tobacco

18

Other SCENIHR 2010). Otherwise it creates the erroneous impression that all cigarettes contain tobacco additives. Whether additives increase “attractiveness” is not relevant under TPD Art. 6. Contrary to SCENIHR’s statement (p.14, l.45 sq), attractiveness and addictiveness are completely different concepts and should not be equated. The Committee itself in 2010 made a very clear distinction between both concepts in its review of the “Addictiveness and Attractiveness of Tobacco Additives” (SCENIHR 2010). TPD Art. 2(26) defines addictiveness as “the pharmacological potential of a substance to cause addiction, a state which affects an individual's ability to control his or her behaviour, typically by instilling a reward or a relief from withdrawal symptoms, or both”. This definition does in no way refer to attractiveness as a relevant criteria to determine addictiveness. We expect SCENIHR to respect the TPD’s legal framework and delete all references to attractiveness (p.12, l.21-22; p.13, l.9 and l.17-19). The sentence on p.12, l.41 is misleading as it implies that the formation of aldehydes is only due to the pyrolysis of added sugars while, as SCENIHR recognizes before (l.37), the “combustion of tobacco generates substances that may be toxic”. The sentence should be amended to clarify that aldehydes are also formed by the pyrolysis of tobacco. Cigarettes made of Virginia tobacco without any added sugars have generally higher aldehyde yields than blended cigarettes with added sugars. Contrary to SCENIHR’s statement on p.14, l.1-2 it is unfounded to conclude that Menthol and Thymol facilitate inhalation. SCENIHR itself stated in SCENIHR 2010, p.55: “It has been proposed, that the cooling and local anaesthetic effects [of menthol] could lead to deeper inhalation of the smoke and higher exposure to other smoke constituents, but current data are inconclusive”. Similarly, FDA PSE 2013, p.43 states: “…the weight of

products (as a general category). Therefore no distinction has to be made between blended and Virginia cigarettes: this clarification will be made in the related legislation (risk management issue). Additives ‘resulting in a characterising flavour’ (which can potentially contribute to attractiveness) as well as additives which ‘facilitates inhalation or nicotine uptake’ (mechanisms potentially contributing to addictiveness) are included in the SCENIHR Terms of Reference (see also previous answer on the same issue). There is in full consistency with the previous SCENIHR Opinion, from which definition and main concepts have been mutuated. The text has been amended to increase the clarity of these two concepts. The text has been amended to improve its clarity.

19

evidence supports the conclusion that menthol in cigarettes is likely not associated with increased or decreased levels of biomarkers of exposure”. SCENIHR should therefore amend its statement as follows: “Their action (delete: “is to”) could decrease the perception of the smoke irritating effects, which may induces the smoker to inhale the smoke deeper and more frequently but evidence so far is insufficient”. SCENIHR should also include in its review studies finding no such association, which we upload in the corresponding sections. Contrary to SCENIHR’s statement on p.14, l.2-4 it is unfounded to conclude that additives with bronchodilating effects in other contexts of use facilitate inhalation. SCENIHR itself stated in SCENIHR 2010, p.42: “[…] the content of theobromine per cigarette will be too low to have a bronchodilating effect on the lungs and thereby increase the absorption of nicotine”. Similarly, RIVM 2002 found “The general conclusion is that the level of these compounds [including theobromine] in added cocoa in cigarettes is not sufficient to increase the addiction to cigarette smoking”. SCENIHR should therefore amend its statement as follows: “A similar result (delete: “can") might be obtained also using bronchodilators, such as theobromine, generated from cocoa, caffeine and glycyrrhizine but evidence so far is insufficient”.

‘Is to’ deleted and changed in ‘could’ Regarding menthol, please also look at the new data described in the specific section (already included in the preliminary Opinion). The entire paragraph actually use ‘may’ and ‘might’ to express the concept of potentiality. The verb ‘can’ has been replaced with 'might' as suggested.

15. AURIEMMA LUIGI, ONT ITALIA SCARL, [email protected], Italy


È necessaria una netta distinzione, pagina 12 (righe da 3 a 13), tra prodotti del tabacco “American blend” e prodotti “Virginia”. I prodotti “American blend” sono costituiti da una miscela che comprende l’aggiunta di diversi additivi ai tabacchi utilizzati, i prodotti “Virginia” invece normalmente non richiedono additivi. Di conseguenza è necessario chiarire che non tutte le sigarette contengono additivi aggiunti al tabacco. Del resto la formazione di aldeidi non è dovuta alla presenza di additivi, dal momento che si formano sia nel caso di prodotti “American blend” che di prodotti “Virginia”. Come riportato nel testo del documento, le sostanze potenzialmente tossiche sono generate dalla combustione del tabacco delle sigarette (pagina 12, riga 37) e non dalla sola aggiunta di particolari additivi.

The distinction between different tobacco products is outside the Terms of Reference for the SCENIHR, which was asked to provide a priority list of additives used in tobacco products (as a general category). Therefore no distinction has to be made between blended and Virginia cigarettes: this clarification will be made in the related legislation (risk management issue). The text has been amended to increase the clarity of these two concepts.

20

16. Chýlková Markéta, Federation of the Food and Drink Industries of the Czech Republic, [email protected], Czech Republic


We are in particular concerned by the fact that the opinion uncompromisingly concludes that menthol helps inhalation without being able to back up this claim by concrete scientific research. Furthermore, SCENIHR itself stated in one of its reports published in 2010 that although it has been proposed that the cooling effects of menthol could lead to deeper inhalation, current data are not conclusive on the matter. We believe that this claim should be restated in the Preliminary opinion to make it more accurate.

Additional information has been made available regarding menthol, since the 2010 SCENIHR Opinion, and are reported in the preliminary Opinion within the Menthol specific information. The German Cancer Research Centre (DKFZ) position paper in 2013 supported the proposal for a revision of the EU Tobacco Product Directive, which includeda ban on menthol. Support for this proposal is that 1) menthol increases the attractiveness of tobacco products, 2) menthol targets youth and 3) menthol may promote smoking initiation (DKFZ, 2013)1. This information was included in the section on menthol inthe PITOC-report. In 2013, the FDA published a review entitled ‘Preliminary scientific evaluation of the possible public health effects of menthol versus non-menthol cigarettes’2, which concluded that although there is little evidence that menthol cigarettes are more toxic than menthol-free cigarettes, adequate data indicate that menthol is associated with increased smoking initiation and greater addiction, especially among young people. This was confirmed by the study of Nonnemaker et al. (2013). Indeed, smokers smoking menthol cigarettes have agreaterdependence on nicotine, which is likely due to the anaesthetic properties which reduce the harshness of the smoke. In addition, menthol cigarette smokers are less successful in quitting from smoking (Smith et

1http://www.dkfz.de/de/tabakkontrolle/download/DKFZ-Position-Paper-Menthol-July2013.pdf

2http://www.fda.gov/downloads/ScienceResearch/.../UCM361598.pdf

21

al., 2015). 17. MASIELLO

GENNARINO, COLDIRETTI (PRESIDENTE REGIONE CAMPANIA), [email protected], Italy


In questa sezione sarebbe opportuno inserire la distinzione tra prodotti miscelati o “American blend”, che normalmente utilizzano additivi, e prodotti “Virginia” che ne utilizzano pochi o nessuno. Questa distinzione è importante per chiarire che ci sono prodotti con additivi e anche prodotti privi di additivi. Infatti, sempre nella stessa sezione, si fa riferimento alle aldeidi che si formano per la pirolisi di vari zuccheri aggiunti ai prodotti del tabacco. Tale affermazione va chiarita, specificando che le aldeidi sono prodotte anche dalla pirolisi del tabacco e quindi presenti, ad esempio, anche nei prodotti di tabacco “Virginia” (con pochi o senza additivi).




The Committee used data on additive concentrations from one source only, the Netherlands; however, as this is not in the public domain, it is impossible to verifying the relevance, or accuracy of the information it contains. It is important to note that not all tobacco companies use the same additives, at the same levels, or with the same functions.

Regarding the use of a list of a single member state, it is already indicated in the preliminary Opinion To compile the list of priority substances, the SCENIHR considered several lists. Some lists of additives were obtained from the public domain which were published by several Member States including Belgium, the Czech Republic, Germany and the Netherlands ; an additional list was received from the UK authorities; and some were taken from other jurisdictions (e.g., USA, Canada, and Brazil). However, except for the Netherlands, the comprehensive list of ingredients that was submitted to regulators in all Member States was not available. Therefore, due to time constraints, the list from the Netherlands. containing 1260 compounds, was used as a typical example. The text has been partially amended to stress this point. The actual ‘use and frequency’ on an EU-wide basis (not available to the SCENIHR) is an issue to be considered by the Commission during preparation of the

22

Similarly, use levels can change over time, so the information on which the Opinion was based may no longer be applicable. SCENIHR appears to have not taken into consideration, the large body of peer reviewed literature published by the tobacco industry and other regarded institutes on additives, and the results of test data generated on those additives under conditions of use, both singly and in combination. Complex “single” ingredients from natural sources (e.g. caramel colours, carob bean extract, cocoa, natural/botanical extracts etc.) should be assessed as they occur in tobacco products, as opposed to studying their individual composition suggested in this Opinion. The rationale to investigate the individual compounds, or components of composition is scientifically flawed, because such approach would not identify how additives actually behave when smoked as tobacco products. Similarly, just because a constituent is high, or low, in percentage does not mean it has greater or lesser degree of chemical reactivity and toxicity. Hence, complex single additives should be assessed as intended to be used in cigarettes, and tested by their effect on the product via smoke chemistry and toxicological testing.

legislation.

The possibility that exposure can change over time is valid for almost all the opinions. Once the Commission will have available a full data set on the priority list as well as and EU-wide information on frequency and use it will be possible to carry out a risk assessment on the additive based on the real condition of use. Content and frequency were among the selection criteria but were not the only ones. However the hazard properties are not going to change. Being unfeasible (as proposed by the commenter) to carry out studies on the almost ‘infinite’ number of possible mixture, the only possibility is to use the component-based approach to evaluate independent action (absence of interaction), additiveness or interactions (see SC opinion on mixtures).

19. SOSSI FRANCESCO, COLDIRETTI (FEDERAZIONE PROVINCIALE BENEVENTO), [email protected], Italy


In questa sezione sarebbe opportuno inserire la distinzione tra prodotti miscelati o “American blend”, che normalmente utilizzano additivi, e prodotti “Virginia” che ne utilizzano pochi o nessuno. Questa distinzione è importante per chiarire che ci sono prodotti con additivi e anche prodotti privi di additivi. Infatti, sempre nella stessa sezione, si fa riferimento alle aldeidi che si formano per la pirolisi di vari zuccheri aggiunti ai prodotti del tabacco. Tale affermazione va chiarita, specificando che le aldeidi sono prodotte anche dalla pirolisi del tabacco e quindi presenti, ad esempio, anche nei prodotti di tabacco “Virginia” (con pochi o senza

The distinction between different tobacco products is outside the Terms of Reference for the SCENIHR, which was asked to provide a priority list of additives used in tobacco products (as a general category). Therefore no distinction has to be made between blended and Virginia cigarettes: this clarification will be made in the related legislation (risk management issue). The text has been amended to increase the

23

additivi). clarity of these two concepts. 20. Ureel Ludwig, British

American Tobacco, [email protected], United Kingdom


The Committee used data on additive concentrations from one source only, the Netherlands; however, as this is not in the public domain, it is impossible to verifying the relevance, or accuracy of the information it contains. It is important to note that not all tobacco companies use the same additives, at the same levels, or with the same functions. Similarly, use levels can change over time, so the information on which the Opinion was based may no longer be applicable. SCENIHR appears to have not taken into consideration, the large body of peer reviewed literature published by the tobacco industry and other regarded institutes on additives, and the results of test data generated on those additives under conditions of use, both singly and in combination. Complex “single” ingredients from natural sources (e.g. caramel colours, carob bean extract, cocoa,

Regarding the use of a list of a single MS, it is already indicated in the preliminary opinion (page 15) To compile the list of priority substances, the SCENIHR considered several lists. Some lists of additives were obtained from the public domain which were published by several Member States including Belgium, the Czech Republic, Germany and the Netherlands ; an additional list was received from the UK authorities; and some were taken from other jurisdictions (e.g., USA, Canada, and Brazil). However, except for the Netherlands, the comprehensive list of ingredients that was submitted to regulators in all Member States was not available. Therefore, due to time constraints, the list from the Netherlands. containing 1260 compounds, was used as a typical example. The text has been partially amended to stress this point. The actual ‘use and frequency’ on an EU-wide basis (not available to the SCENIHR) is an issue to be considered by the Commission during preparation of the legislation. The possibility that exposure can change over time is valid for almost all the opinions. Once the Commission will have available a full data set on the priority list as well as and EU-wide information on frequency and use it will be possible to carry out a risk assessment on the additive based on the real condition of use. Content and frequency were among the selection criteria but were not the only ones. However the hazard properties are not going to change. The available information on ‘mixture toxicity’ is related to specific ‘blend’ and therefore not

24

natural/botanical extracts etc.) should be assessed as they occur in tobacco products, as opposed to studying their individual composition suggested in this Opinion. The rationale to investigate the individual compounds, or components of composition is scientifically flawed, because such approach would not identify how additives actually behave when smoked as tobacco products. Similarly, just because a constituent is high, or low, in percentage does not mean it has greater or lesser degree of chemical reactivity and toxicity. Hence, complex single additives should be assessed as intended to be used in cigarettes, and tested by their effect on the product via smoke chemistry and toxicological testing.

useful for this Opinion. Being unfeasible (as proposed by the commenter) to carry out studies on the almost ‘infinite’ number of possible mixture, the only possibility is to use the component-based approach to evaluate independent action (absence of interaction), additiveness or interactions (see SC opinion on mixtures).

21. Federici Giulio, Coldiretti Caserta, [email protected], Italy


In questa sezione sarebbe opportuno inserire la distinzione tra prodotti miscelati o “American blend”, che normalmente utilizzano additivi, e prodotti “Virginia” che ne utilizzano pochi o nessuno. Questa distinzione è importante per chiarire che ci sono prodotti con additivi e anche prodotti privi di additivi. Infatti, sempre nella stessa sezione, si fa riferimento alle aldeidi che si formano per la pirolisi di vari zuccheri aggiunti ai prodotti del tabacco. Tale affermazione va chiarita, specificando che le aldeidi sono prodotte anche dalla pirolisi del tabacco e quindi presenti, ad esempio, anche nei prodotti di tabacco “Virginia” (con pochi o senza additivi).


22. diop maodo malick, MANE, [email protected], France


in this priority list, substance of concern contribute to the toxicity of the products concern and/or increases the toxicity of any of the products concerned to a significant or measurable degree, Why it define a limit of 15 substances for May 2016? If the rational of introduction is based on toxicity we think that number of substance will be only based on safety concern for public health and can be more or less than 15. In conclusion as producers of those flavouring substances we are very concern by any “affirmation” regarding the impact on health of our substances, and we need robust scientific data to understand why we have to consider those ingredients

This was foreseen by the TPD and what is included in the mandate received by the SCENIHR. It is clearly stated in the preliminary Opinion.

25

«Toxic, addictive, and attractive” 23. MELENEKLIS

VASSILIOS, HELLENIC INTERPROFESSIONAL ORGANIZATION OF RAW TOBACCO, [email protected], Greece


The section at page12 (l.3-13) should mention the distinction between blended cigarettes, which use additives, and Virginia cigarettes, which use no or very few tobacco additives. If this is not done, then an erroneous impression is created that all cigarettes contain tobacco additives. Also, the paper makes statements on toxicity and addictiveness of additives that are not supported by clear definitive evidence. It is indicative of this lack of evidence the fact that in the same document it is recognized that “time constraints only allowed a limited toxicological evaluation of the large number of compounds” (p. 82) and that “there are many data gaps” (p. 14) including on toxicological profiles, inhalation toxicology and exposure to the additive. The paper reaches definitive conclusions based on data it considers “scant” (p. 4) or “negligible” (p. 4).

The distinction between different tobacco products is outside the Terms of Reference for the SCENIHR, which was asked to provide a priority list of additives used in tobacco products (as a general category). Therefore no distinction has to be made between blended and Virginia cigarettes: this clarification will be made in the related legislation (risk management issue). Additives ‘resulting in a characterising flavour’ (which can potentially contribute to attractiveness) as well as additives which ‘facilitate inhalation or nicotine uptake’ (mechanisms potentially contributing to addictiveness) are included in the SCENIHR Terms of Reference. The SCENIHR was asked to provide a priority list of additives, which is not to be considered a ‘full’ risk assessment of the selected compounds but as hazard identification. This was clearly specified in the preliminary Opinion. The list has been build up based on sound scientific criteria. The presence of data gaps cannot therefore challenge the Report; if data gaps are there they will be covered once the Commission will decide about the data requirement.

24. Lutz Rolf, Philip Morris International Management SA, [email protected], Other

3.1 Introduction

P.14, l.16-30 seems to insinuate that, in its risk assessment, the tobacco industry solely relies on the fact that additives have been approved for use in food or cosmetics. This is wrong. Unburnt toxicology is only the first step in the industry’s assessment of additives as described, e.g., in Hahn and Schaub (2010) or Dempsey et al (2012), two peer reviewed publications submitted by PMI

It was not the SCENIHR intention to ‘insinuate’ anything. However, the sentence has been re-phrased. Provided publications have been looked at and since no additional relevant information was found in relation to the Terms of Reference, no change in the text was deemed necessary.

26

as part of its December 2014 offering of information to SCENIHR and not yet considered by SCENIHR. The key in vitro and in vivo assays are designed to evaluate the additive in the particular context of smoking. The ingredient is mixed with tobacco and subsequently burnt and inhaled. It should be clarified that the tobacco industry does far from only rely on substances approved for use by the FDA, JECFA, EFSA or FEMA in its assessment. It is not correct that data on inhalation toxicity (p.14, l.31) and mixture toxicity (p.14,l.42) do not exist or are scant. The results of in vivo studies (rat inhalation studies) and mix studies have been reported in peer reviewed publications which we upload in the corresponding sections. SCENIHR should consider these studies or at least explain why they did not consider them. While SCENIHR’s own Rules of Procedure require scientific advice to represent “the best information and guidance” (principle of excellence) and should be based on “the best reasonably obtainable date and information” (B. Methods, 9. Use best data), SCENIHR relies for almost one third of the additives (14 out of 48) solely on the PITOC factsheets, a secondary source, and leaves aside many peer reviewed publications. The PITOC factsheets were designed to inform professionals and the public and do not constitute a thorough scientific review of the evidence. Taking sugars as an example, PITOC solely refers to two peer reviewed studies (Baker 2004 and Talhout 2006) out of 9 relevant peer reviewed publications at the time. Other secondary sources have been used by SCENIHR, such as a DKFZ position paper (l.14) and a FDA review (l.19). In order to meet its own standards, SCENIHR should revise its opinion to include all relevant primary sources instead of relying on secondary sources. SCENIHR states it carried out an “update of the literature for the period 2012-2014” (P.16, l.10). However, contrary to its own Rules of Procedure SCENIHR does not clearly identify the “strategies and

Data on specific mixtures cannot be used to generalise, and in addition the toxicity of tested ingredients (in terms of hazard identification) cannot be regarded on overall toxicity of cigarette smoke, but on the basis of their own toxicity. In the methodology paragraph of the preliminary Opinion, indication on the procedure followed has been described. There it is clearly stated that the weight of evidence approach has been followed for some aspects, not being a risk assessment but a prioritization based on hazard on a high number of additives. Secondary sources of information used as evaluations carried out by other Agencies (i.e. EFSA or EPA) were considered fully reliable. PITOC is an EU funded project, considered as reliable. The Commisson included PITOC results within the ‘supporting’ documents. The additives were selected regardless of PITOC results; then once the compounds were selected, the factsheets for PITOC additives were used, since the information included was the results of an extensive literature search and evaluation. In addition the results of the literature search carried out to update information from 2012 to 2015 were analysed by the SCENIHR members not previously involved in the PITOC project to increase judgement independency.

27

procedures for identifying and acquiring data and information” (B. Methods, 9. Use of best data). In particular, SCENIHR’s Preliminary Opinion is not transparent regarding inclusion and exclusion criteria, number of studies identified and retained. SCENIHR should add this information to be transparent and to meet its own standards. P.16, l.11-12: Absent relevant information on methodology, SCENIHR’s statement that “no new relevant information was available that would require adaptation or changes of the report's conclusion” cannot be verified. This does not meet the Rules of Procedures: “The opinions should include all the information necessary for the understanding and, as far as possible, reproducibility of processes and results. All the important steps, assumptions, calculations made should be documented” (24. Completeness of Opinions). SCENIHR’s grouping of substances (P.17, l.5-11) is not done according to the oR allowing it to group only “substances with identical or very similar properties”. E.g., propylene glycol, glycerol and sorbitol, grouped under humectants, do not fulfill these criteria. The heterogeneity of the groups leaves unlimited scope to arbitrarily include additional substances which does not allow any predictability and planning regarding the testing and use of additives by manufacturers. All 48 priority additives should be listed individually.

Therefore what has been ‘asserted’ was exactly what the SCENIHR did. The methodology paragraph has been amended for clarification purposes. The current Opinion serves, as stipulated in the Terms of Reference, the compilation of a priority list. This list will assist, in line with Article 6 of Directive 2014/40/EU, the Commission to develop priority list of at least 15 additives for which enhanced reporting obligations will apply (as described in the section 1 ‘background’). If all the information is already available to Tobacco Industry, it will be easier to answer to the Commission data requirements.The SCENIHR is not in the position to request any additional testing. The grouping has been carried out considering structurally similar chemicals used as additives, which are supposed to show similar properties (with a focus on those relevant for this Opinion). This was included in the Terms of Reference.

28

25. Rapota Maria, JT International S.A., 1, Rue de la Gabelle, 1211 Geneva 26, [email protected], Other

3.1 Introduction

Page 12, lines 14-16 SCENIHR’s 2010 report was inconclusive as to how either “addictiveness” or “attractiveness” could be reliably measured for tobacco additives. Page 13, lines 9-16 We agree with the SCENIHR 2010 report that there is no widely agreed universal standard as to what tests (if any) could be employed to measure the “addictiveness” of specific tobacco product ingredients or combinations of the same under intended conditions of use. We note that, whatever definition of “addictiveness” is used, this does not affect the fact that people can stop smoking if they are determined to do so. Based upon the available scientific evidence our view is that tobacco products with added ingredients are no more difficult to quit than those which do not contain added ingredients (Lee et al., 2009, Werley et al., 2007, Sanders et al., 2012). The WHO (2007) definition of “attractiveness” fails established criteria for issue definition and no scientific criteria have been developed to assess the “attractiveness” of tobacco products, let alone to regulate on that basis. Page 13, lines 17-33 The concept of “attractiveness” described here is both inherently subjective and so broad as to be meaningless. JTI does not accept the suggestion that a policy objective of ingredient regulation should be to make smoking less pleasurable. Page 13, line 18 JTI has never used “light” or similar branding to reduce smokers’ concerns about the health risks of smoking. Page 13, lines 34-45 and Page 14, lines 1-14 The hypothetical physiological mechanisms described here have yet to be conclusively demonstrated and the presumed role of added ingredients has yet to be established. Page 13, lines 34-43 JTI does not use additives with the intention of altering the pH of cigarette smoke and does not believe doing so would in any case affect the bioavailability of nicotine to the smoker. Page 13 (lines 43-) JTI does not use additives with the intent of increasing bronchodilation and/or creating anaesthetising effects and does not believe that the additives it uses have such effects. Page 14 (lines 7-10) “Additives, which interfere with nicotine kinetics…” –

We agree with the SCENIHR 2010 report that there is no widely agreed universal standard to measure the 'addictiveness' or ‘actractiveness’ of specific tobacco products. However, additives ‘resulting in a characterising flavour’ (which can potentially contribute to attractiveness) as well as additives which ‘facilitates inhalation or nicotine uptake’ (which can potentially contribute to addictiveness) are included in the SCENIHR Terms of Teference. There is full consistency with the previous SCENIHR Opinion, from which definition and main concepts have been mutuated. This issue has been reworded within the text. It was not the intention of the SCENHIR to accuse anybody, but this is the perception of the general public. The entire paragraph actually use ‘may’ and ‘might’ to express the concept of potentiality. As stated in the Data gap paragraph, although information about possible effects on addictiveness and attractiveness does not exist (due to lack of specific test), information can be derived based on the mode of action of the single chemical used as additive. The text of the Opinion has been amended to make concepts related to addictveness and attractiveness more understandable and in

29

SCENIHR’s 2010 report did not conclusively identify any additives that in fact had these properties. Page 14, lines 16-37, lines 38-43, Page 15, lines 22-29 Many published pyrolysis studies and toxicological studies testing additive mixtures are not mentioned in the report. Page 14, lines 44-45 and Page 15, line 1-2 “In addition, it is difficult, in many cases, to distinguish between addictiveness and attractiveness, because addiction to a tobacco product is, to a large extent, also related to the attractiveness of the product.” The statement does not correspond to the pharmacological model of nicotine addiction that was central to the SCENIHR 2010 report, to references in this report to additives which supposedly enhance the pharmacological effects of nicotine or to the TPD2 definition of “addictiveness” as “the pharmacological potential of a substance to cause addiction, a state which affects an individual’s ability to control his or her behaviour…”. What it does demonstrate is that the concept of “attractiveness” is so vague as to be meaningless. Page 16, lines 13-27 The conclusions on menthol are not supported by the majority of published literature on the topic, please refer to the menthol section. Page 17, lines 5-11 SCENIHR identified 48 instead of 20-30 additives, which is not in line with the mandate from the Commission. Additives should also be listed as single entries, not in groups. Page 17, lines 30-33 The search terms used by SCENHIR should be transparent. Many relevant published studies are not referred to in the report.

line with TPD. Provided publications have been looked at and since no relevant information was found in relation to the Terms of Reference, no change in the text was deemed necessary. Data on specific mixtures cannot be used to generalise, and in addition the toxicity of tested ingredients (in terms of hazard identification) cannot be regarded on overall toxicity of cigarette smoke, but on the basis of their own toxicity. Additional information has been made available regarding menthol, since the 2010 SCENIHR Opinion, and are reported in the preliminary Opinion. See Menthol specific section.

26. Wilkinson Peter, Imperial Tobacco Limited, [email protected], United Kingdom

3.1 Introduction

This section lacks scientific detail, and would benefit from appropriate referencing. There is a statement that additives can be toxic per se; this is an oversimplification. In the absence of further qualification, the same statement can be made for all chemical substances We firmly believe that there is more public domain information available on additives used in tobacco products than implied by this section; with additional time available to SCENHIR it would have been possible to review other relevant data.

The section is only an Introduction, to explain the criteria used in order to compile the priority list. The SCENIHR agrees that all chemical substances are potentially toxic per se, therefore there is no reason to consider the statement not valid for chemicals used as additives in tobacco product. The use of the phrase was introduced to distinguish between the additives as such and its combustion

30

SCENHIR acknowledges that only a limited toxicological evaluation was possible due to time constraints and a large number of compounds involved. This is compounded with the primary focus on additive information from a single EU Member State, rather than reflecting the whole EU. Indeed, in providing Terms of reference the EU Commission specifically invited the committee to consult the data reported by the industry under Directive 2001/37/EC. As this evaluation is, however, the most relevant part of the task, it gives rise to doubts on the completeness of the Preliminary Opinion.

products. However, since it can be confused with toxicity due to parent compound rather than biotransformation products within the organisme, the expression per sè has been replaced by ‘in their unburnt form’. The SCENIHR was asked to provide a priority list of additives, which is not to be considered as ‘full’ risk assessment of the selected compounds, but as hazard identification. This was clearly specified in the preliminary Opinion. The list has been build up based on sound scientific criteria. Regarding the use of a list of a single member state, it is already indicated in the preliminary Opinion To compile the list of priority substances, the SCENIHR considered several lists. Some lists of additives, obtained from the public domain, were published by several Member States including Belgium, the Czech Republic, Germany and the Netherlands; an additional list was received from the UK authorities; and some were taken from other jurisdictions (e.g., USA, Canada, and Brazil). However, except for the Netherlands, the comprehensive list of ingredients that was submitted to regulators in all Member States was not available. Therefore, due to time constraints, the list from the Netherlands containing 1260 compounds, was used as a typical example. The text has been partially amended to stress this point. However, the actual ‘use and frequency’ on an EU-wide basis (not available to the SCENIHR) is an issue to be considered by the Commission during preparation of the legislation

31

We respectfully include some relevant materials within our submission. This section contains contradictory statements and is not qualified to become part of the final Opinion. For example, it argues that tobacco is alkalised to alter the pH and supposedly increase “free” nicotine. While perhaps compelling from a simplistic chemistry perspective, it is in direct opposition to the statement on page 13, Lines 1-4, where the Preliminary Opinion also states that alkaline smoke is harsh and irritating and so sugars and acids are added to restore the pH to lower levels. Moreover, SCENIHR omits its own conclusion in 2010 on “bronchodilators”; in the current document it is suggested that cocoa, caffeine and glycyrrhizine act a bronchodilators to facilitate smoke inhalation. In 2010 SCHENIHR “Based on the information available on the PMI and BAT websites the percentage of cocoa used in cigarettes ranges from 0.2% to 0.66%. Taking this information into account, the content of theobromine per cigarette will be too low to have a bronchodilating effect on the lungs and thereby increase the absorption of nicotine”. If SCENIHR has new evidence to the contrary this should be cited. Also, multiple modes of action are suggested by which nicotine kinetics are affected by additives; significantly, no references are provided to substantiate this claim. This section refers to "attractiveness". SCENIHR (2010) addressed the substantial limitations in assessing or measuring the "attractiveness" of a tobacco additive by highlighting an absence of suitable animal models and ethical concerns should human testing be involved. It is fair to conclude from these limitations that no appropriate or validated methodology exists for measuring “attractiveness” of a tobacco additive. Consequently, from a scientific perspective, it is not relevant for the exercise of defining a list of priority additives. Moreover, the term is not within the mandate

Provided publications have been looked at and since no additional relevant information were found in relation to the Terms of Reference, no change in the text was deemed necessary. The SCENIHR does not see any contractidction in the text of the Opinion. There is full consistency with the previous SCENIHR Opinion, from which definition and main concepts have been mutuated. But it has been

Results of the public consultation ... - European Commission · requested; the Commission requested the SCENIHR to provide a list of priority substances (as foreseen in Art. 6 of

Documents