[Federal Register Volume 83, Number 195 (Tuesday, October 9, 2018)]
[Rules and Regulations]
[Pages 50490-50503]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2018-21807]


-----------------------------------------------------------------------

DEPARTMENT OF HEALTH AND HUMAN SERVICES

Food and Drug Administration

21 CFR Parts 172 and 177

[Docket No. FDA-2015-F-4317]


Food Additive Regulations; Synthetic Flavoring Agents and 
Adjuvants

AGENCY: Food and Drug Administration, HHS.

ACTION: Final rule; notification of partial denial of petition.

-----------------------------------------------------------------------

SUMMARY: The Food and Drug Administration (FDA, the Agency, or we) is 
partially granting a petition submitted by the Breast Cancer Fund (now 
known as the Breast Cancer Prevention Partners), Center for 
Environmental Health, Center for Food Safety, Center for Science in the 
Public Interest, Consumers Union, Environmental Defense Fund, 
Environmental Working Group, Improving Kids' Environment, Natural 
Resources Defense Council, WE ACT for Environmental Justice, and Mr. 
James Huff, by amending the food additive regulations to no longer 
authorize the use of benzophenone, ethyl acrylate, eugenyl methyl 
ether, myrcene, pulegone, and pyridine as synthetic flavoring 
substances for use in food. We are taking this action because, despite 
FDA's scientific analysis and determination that these substances do 
not pose a risk to public health under the conditions of their intended 
use, the petitioners provided data demonstrating that these additives 
induce cancer in laboratory animals, and, as a result of this finding 
in animals, FDA cannot as a matter of law maintain the listing of these 
synthetic flavoring substances in the food additive regulations. 
Because of evidence that benzophenone causes cancer in animals, FDA 
also is amending the food additive regulations to no longer provide for 
the use of benzophenone as a plasticizer in rubber articles intended 
for repeated use in contact with food. FDA is denying as moot the 
portions of the petition proposing that the food additive regulations 
be amended to no longer authorize the use of styrene as a synthetic 
flavoring substance because this use has been permanently and 
completely abandoned. In addition, FDA is declining to act on the 
petitioners' request to issue a regulation to prohibit the use of these 
synthetic flavoring substances in food because that issue is not the 
proper subject of a food additive petition.

DATES: This rule is effective October 9, 2018. See section IX for 
further information on the filing of objections. Submit either 
electronic or written objections and requests for a hearing on the 
final rule by November 8, 2018.

ADDRESSES: You may submit objections and requests for a hearing as 
follows. Please note that late, untimely filed objections will not be 
considered. Electronic objections must be submitted on or before 
November 8, 2018. Objections received by mail/hand delivery/courier 
(for written/paper submissions) will be considered timely if they are 
postmarked or the delivery service acceptance receipt is on or before 
that date.

Electronic Submissions

    Submit electronic objections in the following way:
     Federal eRulemaking Portal: https://www.regulations.gov. 
Follow the instructions for submitting comments. Objections submitted 
electronically, including attachments, to https://www.regulations.gov 
will be posted to the docket unchanged. Because your objection will be 
made public, you are solely responsible for ensuring that your 
objection does not include any confidential information that you or a 
third party may not wish to be posted, such as medical information, 
your or anyone else's Social Security number, or confidential business 
information, such as a manufacturing process. Please note that if you 
include your name, contact information, or other information that 
identifies you in the body of your objection, that information will be 
posted on https://www.regulations.gov.
     If you want to submit an objection with confidential 
information that you do not wish to be made available to the public, 
submit the objection as a written/paper submission and in the manner 
detailed (see ``Written/Paper Submissions'' and ``Instructions'').

Written/Paper Submissions

    Submit written/paper submissions as follows:
     Mail/Hand delivery/Courier (for written/paper 
submissions): Dockets Management Staff (HFA-305), Food and Drug 
Administration, 5630 Fishers Lane, Rm. 1061, Rockville, MD 20852.
     For written/paper objections submitted to the Dockets 
Management Staff, FDA will post your objection, as

[[Page 50491]]

well as any attachments, except for information submitted, marked and 
identified, as confidential, if submitted as detailed in 
``Instructions.''
    Instructions: All submissions received must include the Docket No. 
FDA-2015-F-4317 for ``Food Additives Permitted for Direct Addition to 
Food for Human Consumption; Synthetic Flavoring Agents and Adjuvants.'' 
Received objections, those filed in a timely manner (see ADDRESSES), 
will be placed in the docket and, except for those submitted as 
``Confidential Submissions,'' publicly viewable at https://www.regulations.gov or at the Dockets Management Staff between 9 a.m. 
and 4 p.m., Monday through Friday.
     Confidential Submissions--To submit an objection with 
confidential information that you do not wish to be made publicly 
available, submit your objections only as a written/paper submission. 
You should submit two copies total. One copy will include the 
information you claim to be confidential with a heading or cover note 
that states ``THIS DOCUMENT CONTAINS CONFIDENTIAL INFORMATION.'' We 
will review this copy, including the claimed confidential information, 
in our consideration of comments. The second copy, which will have the 
claimed confidential information redacted/blacked out, will be 
available for public viewing and posted on https://www.regulations.gov. 
Submit both copies to the Dockets Management Staff. If you do not wish 
your name and contact information to be made publicly available, you 
can provide this information on the cover sheet and not in the body of 
your comments and you must identify this information as 
``confidential.'' Any information marked as ``confidential'' will not 
be disclosed except in accordance with 21 CFR 10.20 and other 
applicable disclosure law. For more information about FDA's posting of 
comments to public dockets, see 80 FR 56469, September 18, 2015, or 
access the information at: https://www.gpo.gov/fdsys/pkg/FR-2015-09-18/pdf/2015-23389.pdf.
    Docket: For access to the docket to read background documents or 
the electronic and written/paper comments received, go to https://www.regulations.gov and insert the docket number, found in brackets in 
the heading of this document, into the ``Search'' box and follow the 
prompts and/or go to the Dockets Management Staff, 5630 Fishers Lane, 
Rm. 1061, Rockville, MD 20852.

FOR FURTHER INFORMATION CONTACT: Judith Kidwell, Center for Food Safety 
and Applied Nutrition (HFS-265), Food and Drug Administration, 5001 
Campus Dr., College Park, MD 20740-3835, 240-402-1071.

SUPPLEMENTARY INFORMATION: 

Table of Contents:

I. Introduction
II. Background
    A. Statutory and Regulatory Background
    B. Abandonment of Use of Styrene Authorized Under 21 CFR 172.515
    C. History of the Regulation of the Synthetic Flavoring 
Substances and Adjuvants
    D. Summary and Context of Determination
III. Evaluation of Carcinogenicity
    A. Benzophenone
    B. Ethyl Acrylate
    C. Methyl Eugenol
    D. Myrcene
    E. Pulegone
    F. Pyridine
IV. Comments on the Notice of Petition
    A. Legal and Policy Issues
    B. Scientific Issues
V. Conclusion
VI. Public Disclosure
VII. Analysis of Environmental Impacts
VIII. Paperwork Reduction Act
IX. Objections
X. References

I. Introduction

    In the Federal Register of January 4, 2016 (81 FR 42), we announced 
that the Center for Science in the Public Interest, Natural Resources 
Defense Council, Center for Food Safety, Consumers Union, Improving 
Kids' Environment, Center for Environmental Health, Environmental 
Working Group, Environmental Defense Fund, and James Huff (the 
petitioners), c/o Mr. Tom Neltner, 1875 Connecticut Ave. NW, 
Washington, DC 2009, had jointly filed a food additive petition (FAP 
5A4810). Subsequently, the Breast Cancer Fund (now known as Breast 
Cancer Prevention Partners) and WE ACT for Environmental Justice joined 
as co-petitioners.
    The petition proposed that we take two separate regulatory actions: 
(1) Amend the food additive regulations in Sec.  172.515 Synthetic 
flavoring substances and adjuvants (21 CFR 172.515) to no longer 
authorize the use of seven listed synthetic flavoring food additives 
and (2) to establish zero tolerances in Sec.  172.515 for these 
additives. However, the food additive regulation is not the appropriate 
section for a ``zero tolerance,'' and this request is not the proper 
subject of a food additive petition. A food additive petition must 
either propose the issuance of a regulation prescribing the conditions 
under which a food additive may be safely used (see section 409(b)(1) 
of the Federal Food, Drug, & Cosmetic Act (FD&C Act) (21 U.S.C. 
348(b)(1)), or propose the amendment or repeal of an existing food 
additive regulation (see section 409(i) of the FD&C Act. Only the 
petitioners' request to amend Sec.  [thinsp]172.515 to remove the seven 
synthetic flavorings and adjuvants from FDA's regulations permitting 
their use as additives in food falls within the statutory scope of a 
food additive petition. Therefore, the petitioners' request that we 
establish zero tolerances for these seven flavoring additives falls 
outside the scope of a food additive petition. As a result, we are not 
addressing that request further in this rule. (An interested person may 
use the citizen petition process to request the issuance of a 
regulation, including a request to establish a ``zero tolerance,'' 
which we interpret as a request to issue a regulation prohibiting a 
substance from human food under part 189 (see 21 CFR 189.1(c) 
(referring to 21 CFR part 10, which sets forth FDA's citizen petition 
process)). (In addition, we understand the petitioners are no longer 
pursuing this request based on a public filing with a U.S. court of 
appeals (stating ``[t]he Petition also requested that FDA `establish a 
zero tolerance [standard]. . . for the use of these seven flavors.' . . 
. Petitioners are no longer pursuing this aspect of the Petition''). 
(See In Re Breast Cancer Prevention Partners, No. 18-71260 (9th Cir.)). 
Thus, in this rule we focus solely on the request to amend the food 
additive regulations.
    The seven food additives that are the subject of this petition are:
    1. Benzophenone (also known as diphenylketone) (CAS No. 119-61-9);
    2. Ethyl acrylate (CAS No. 140-88-5);
    3. Eugenyl methyl ether (also known as 4-allylveratrole or methyl 
eugenol) (CAS No. 93-15-2);
    4. Myrcene (also known as 7-methyl-3-methylene-1,6-octadiene) (CAS 
No. 123-35-3);
    5. Pulegone (also known as p-menth-4(8)-en-3-one) (CAS No. 89-82-
7);
    6. Pyridine (CAS No. 110-86-1); and
    7. Styrene (CAS No. 100-42-5).
    We stated in the notice of petition that, although the petition 
only proposes to amend Sec.  172.515 to no longer provide for the use 
of these seven synthetic flavoring substances, FDA's action in response 
to the petition could affect other regulations that provide for the use 
of the additives. Specifically, in the notice we identified the use of 
benzophenone, which is approved as an indirect food additive, i.e., a 
plasticizer (diphenylketone in Sec.  177.2600 (21 CFR 
177.2600(c)(4)(iv))), as potentially being impacted by our regulatory 
decision. The notice of petition gave interested parties until March 4, 
2016, to submit

[[Page 50492]]

comments on the filed food additive petition. In response to a written 
request submitted to the docket, we extended the comment period to May 
3, 2016 (81 FR 8867, February 23, 2016).
    This final rule partially granting the request to revise the 
regulations to no longer provide for the use of these synthetic 
flavorings in food, and the partial denial given the petitioners' 
request falls outside the scope of the food additive petition process, 
completely responds to the petition.

II. Background

A. Statutory and Regulatory Background of Food Additive Regulation

    The FD&C Act authorizes us to regulate ``food additives'' (see 
section 409(a) of the FD&C Act). The FD&C Act defines ``food 
additive,'' in relevant part, as any substance the intended use of 
which results or may reasonably be expected to result, directly or 
indirectly, in its becoming a component of food (see section 201(s) of 
the FD&C Act (21 U.S.C. 321(s))). Food additives can include both 
substances added directly to food and ``food contact substance[s]'' 
(i.e., substances intended for use in materials that come into contact 
with food, for instance in food packaging or manufacturing, but which 
are not intended to have any technical effect in the food (see Sec.  
170.3(e)(3) (21 CFR 170.3(e)(3))). Food additives are deemed unsafe and 
prohibited except to the extent that we approve their use (see, e.g., 
section 301(a) and (k) (21 U.S.C. 331(a) and (k)) and 409(a) of the 
FD&C Act).
    The FD&C Act provides a process through which persons who wish to 
use a food additive may submit a petition proposing the issuance of a 
regulation prescribing the conditions under which the additive may be 
safely used (see section 409(b)(1) of the FD&C Act). Such a petition is 
referred to as a ``food additive petition.'' A food additive petition 
must either propose the issuance of a regulation prescribing the 
conditions under which a food additive may be safely used (see section 
409(b)(1) of the FD&C Act), or propose the amendment or repeal of an 
existing food additive regulation (see section 409(i) of the FD&C Act). 
When we conclude that a proposed use of a food additive is safe, we 
issue a regulation called a ``food additive regulation'' authorizing a 
specific use of the substance.
    A food additive cannot be approved for use unless the data 
presented to FDA establish that the food additive is safe for that use 
(section 409(c)(3)(A) of the FD&C Act). To determine whether a food 
additive is safe, the FD&C Act requires FDA to consider, among other 
relevant factors: (1) Probable consumption of the additive; (2) 
cumulative effect of such additive ``in the diet of man or animals''; 
and (3) safety factors recognized by experts ``as appropriate for the 
use of animal experimentation data'' (section 409(c)(5) of the FD&C 
Act). FDA's determination that a food additive use is safe means that 
there is a ``reasonable certainty in the minds of competent scientists 
that the substance is not harmful under the intended conditions of 
use'' (Sec.  170.3(i)). However, FDA cannot approve a food additive if 
it is found ``to induce cancer when ingested by man or animal, or if it 
is found, after tests which are appropriate for the evaluation of the 
safety of food additives, to induce cancer in man or animal'' (section 
409(c)(3)(A) of the FD&C Act). This provision, which is often referred 
to as the ``Delaney Clause,'' was added to the FD&C Act by the Food 
Additives Amendment of 1958 (Pub. L. 85-929). The Delaney Clause limits 
FDA's discretion to determine the safety of food additives, in that it 
prevents FDA from finding a food additive to be safe if it has been 
found to induce cancer when ingested by humans or animals, regardless 
of the probability, or risk, of cancer associated with exposure to the 
additive or of the extent to which the experimental conditions of the 
animal study or the carcinogenic mode of action provide insight into 
the health effects of human consumption and use of the additive in 
question. In Public Citizen v. Young, the DC Circuit Court of Appeals 
held that Congress intended for the Delaney Clause to be 
``extraordinarily rigid,'' to protect the public from cancer-causing 
substances without exception, rejecting FDA's argument that a 
particular color additive, which was subject to a similarly worded 
Delaney Clause for color additives, should be approved because it did 
not pose more than a de minimis cancer risk (831 F.2d 1108, 1122 (DC 
Cir. 1987); see also Les v. Reilly, 968 F.2d 985, 986 (9th Cir. 1992) 
(holding that the Environmental Protection Agency's refusal to revoke 
regulations permitting the use of certain pesticides (which were 
regulated as food additives at the time of the court decision), on the 
grounds that they pose a de minimis cancer risk, is contrary to the 
provisions of the Delaney Clause).
    The FD&C Act provides that FDA must by regulation prescribe the 
procedure by which a food additive regulation may be amended or 
repealed (see section 409(i) of the FD&C Act). Our regulation specific 
to the administrative actions for food additives provides that the 
Commissioner of Food and Drugs (the Commissioner), on his or her own 
initiative or on the petition of any interested person, may propose the 
issuance of a regulation amending or repealing a regulation pertaining 
to a food additive (see Sec.  [thinsp]171.130(a) (21 CFR 171.130(a))). 
Our regulation, at Sec.  [thinsp]171.130(b), further provides that any 
such petition must include an assertion of facts, supported by data, 
showing that new information exists with respect to the food additive 
or that new uses have been developed or old uses abandoned, that new 
data are available as to toxicity of the chemical, or that experience 
with the existing regulation or exemption may justify its amendment or 
repeal.
    The specific food additive regulation at issue in the petition, 
Sec.  [thinsp]172.515, lists synthetic flavoring substances and 
adjuvants that may be safely used in food in accordance with the 
conditions in the regulation. At issue in the petition are seven 
synthetic flavorings and adjuvants listed in this regulation: 
Benzophenone (also known as diphenylketone), ethyl acrylate, eugenyl 
methyl ether (also known as 4-allylveratrole or methyl eugenol), 
myrcene (also known as 7-methyl-3-methylene-1,6-octadiene), pulegone 
(also known as p-menth-4(8)-en-3-one, pyridine, and styrene. The 
petitioners assert that new data establish that these synthetic 
flavoring additives are carcinogenic and therefore not safe for use in 
food pursuant to the Delaney Clause.

B. Abandonment of Use of Styrene Authorized Under 21 CFR 172.515

    Related to FAP 5A4810, in a document published in the Federal 
Register on June 15, 2016 (81 FR 38984), we announced that we filed a 
food additive petition (FAP 6A4817) proposing that we amend Sec.  
172.515 to no longer provide for the use of styrene as a synthetic 
flavoring substance and adjuvant in food because the use has been 
abandoned. Elsewhere in this issue of the Federal Register, we have 
published a final rule in response to FAP 6A4817 granting that petition 
and amending Sec.  172.515 to no longer authorize the use of styrene as 
a synthetic flavoring substance and adjuvant in food because its use 
under Sec.  172.515 has been permanently and completely abandoned. 
Because the final rule issued in response to FAP 6A4817 removes styrene 
from Sec.  172.515--thereby taking one of the actions requested in this 
petition--the petitioners' request is moot, and it is neither necessary 
nor an efficient use of our resources to address the petitioners'

[[Page 50493]]

assertions regarding the safety of the food additive use of styrene 
that is no longer authorized. Therefore, we are denying as moot the 
request in FAP 5A4810 to remove styrene from Sec.  172.515.

C. History of the Regulation of the Synthetic Flavoring Substances and 
Adjuvants

    In the Federal Register of May 27, 1964 (29 FR 6957), FDA published 
a proposed rule to establish a regulation for synthetic flavoring 
substances and adjuvants used in food. The purpose of the proposed 
regulation was to identify those synthetic substances that may be 
safely used as flavoring substances or flavor adjuvants in food. The 
proposed regulation listed many synthetic flavoring substances and 
adjuvants in use at the time, including benzophenone, ethyl acrylate, 
eugenyl methyl ether, myrcene, pulegone, and pyridine. The proposed 
rule stated that, in reaching a conclusion about the safety of the 
substances listed in the proposed order, FDA relied upon experience 
based on the common use of these substances in food prior to 1958; the 
fact that many of the synthetic flavoring substances have a natural 
counterpart in food or in natural substances used to flavor foods; that 
metabolic and toxicity data representing studies made on selected 
flavoring substances were reviewed and safety established; and that 
relatively low and essentially self-limiting quantities are involved 
when these substances are used in food, consistent with good 
manufacturing practice. (29 FR 6957). In the Federal Register of 
October 27, 1964 (29 FR 14625), FDA issued a final rule based on this 
proposal with a few changes based on comments that were received and 
established this regulation in 21 CFR 121.1164. This regulation also 
limited the amount of the synthetic flavoring substance that could be 
added to food to the smallest amount necessary to achieve the desired 
flavoring effect. In the Federal Register of March 15, 1977 (42 FR 
14302 at 14492), 21 CFR 121.1164 was redesignated Sec.  172.515.

D. Summary and Context of Determination

    We have evaluated the data and information submitted by the 
petitioners, as well as other relevant carcinogenicity data and 
information, and have determined the remaining six synthetic flavoring 
substances (i.e., other than styrene) that are the subject of FAP 
5A4810 are unlikely to pose a potential or significant carcinogenic 
risk for humans at the levels that these synthetic flavoring substances 
are used in foods, and that the use of these food additives is safe for 
human consumption. In other words, FDA has a reasonable certainty that 
the substances do no harm under the intended conditions of use (the 
standard for approving food additives). However, because data submitted 
by the petitioners demonstrate that these synthetic flavoring 
substances have been shown to induce cancer in animal studies, FDA 
cannot consider these synthetic flavoring substances to be safe as a 
matter of law because of the Delaney Clause, and must revoke the 
listings providing for the use of these synthetic flavoring substances 
and adjuvants, as described further in section III.
    In making this determination, we reiterate the point, first made in 
our 1964 proposed rulemaking, that all of the synthetic flavoring 
substances that are the subject of the petition have a natural 
counterpart in food or in natural substances used to flavor foods. For 
example, benzophenone is present in grapes, ethyl acrylate is present 
in pineapple, eugenyl methyl ether (methyl eugenol) is present in 
basil, myrcene is present in citrus fruit, pulegone is present in 
peppermint, and pyridine is present in coffee. FDA's revocation of the 
listings providing for the use of these synthetic flavoring substances 
and adjuvants does not affect the legal status of foods containing 
natural counterparts or non-synthetic flavoring substances extracted 
from food, and there is nothing in the data FDA has reviewed in 
responding to the pending food additive petition that causes FDA 
concern about the safety of foods that contain natural counterparts or 
extracts from such foods.

III. Evaluation of Carcinogenicity

    The petitioners assert that each of the synthetic flavoring 
substances (i.e., benzophenone, ethyl acrylate, methyl eugenol, 
myrcene, pulegone, and pyridine) has been shown to induce cancer in 
animals by studies sponsored by the Department of Health and Human 
Services' National Toxicology Program (NTP). The petitioners also cite 
conclusions of the International Agency for Research on Cancer (IARC) 
and the California Environmental Protection Agency's Office of 
Environmental Health Hazard Assessment (OEHHA), and assert that 
information that became available after these food additives were 
listed in Sec.  172.515 demonstrates that ``they are not safe for use 
in food pursuant to the Delaney Clause''; however, we note that the 
conclusions from IARC and OEHHA are based primarily on results from the 
NTP studies. Thus, our review of whether the synthetic flavoring 
substances that are the subject of the petition induce cancer in humans 
or animals focused on results of the NTP studies, as well as other 
available relevant information discussed in this rule.
    As part of our scientific review, we also evaluated the 
genotoxicity of the synthetic flavoring substances. Based on their 
biological activities, chemical carcinogens can be classified as 
genotoxic (directly DNA reactive) and non-genotoxic (not directly DNA 
reactive but operating through a secondary mechanism) (Ref. 1). In 
cancer risk assessments, the traditional assumption for chemicals that 
are genotoxic is that there is no threshold exposure level below which 
there is no risk of cancer and that there is a risk of cancer at any 
level of exposure. In contrast, non-genotoxic carcinogens are assumed 
to have a threshold of exposure level below which tumor development is 
not anticipated and the risk of cancer is negligible (Ref. 2).
    Additionally, as part of our review, we calculated Margins of 
Exposure (MOE) for each of the six synthetic flavoring substances. The 
MOE is the ratio between a point of departure (e.g., no-observed-
adverse-effect-dose or benchmark dose) and estimates of human dietary 
exposure. As a risk characterization tool, the MOE can be used to 
provide information on the level of public health concern. The MOE is 
invaluable in risk management for chemicals present in food, when a 
health-based guidance level is impossible to derive, such as with 
genotoxic and carcinogenic contaminants and veterinary drug residues 
(Refs. 2 and 3). If the MOE is very large (such as greater than 
10,000), it can be an indication of a low level of human health risk 
(Ref. 3).
    We also estimated dietary exposure for the six synthetic flavoring 
substances using information from the 2015 Poundage and Technical 
Effects Survey that the Flavor and Extract Manufacturers Association 
(FEMA) collected from its member companies that formulate flavoring 
substances (Ref. 4). (The acronym FEMA, as used throughout this rule, 
refers to the Flavor and Extract Manufacturers Association. It should 
not be confused with the Federal Emergency Management Agency that 
commonly is referred to by this same acronym.) Every 5 years FEMA 
surveys its members to estimate the total volume of flavoring 
substances added to food, or ``poundage data.'' (The 2015 poundage data 
were the most recent available.) FEMA's members include flavor 
manufacturers, flavor users, flavor

[[Page 50494]]

ingredient suppliers, and others with an interest in the U.S. flavor 
industry. According to FEMA, their flavor manufacturing members produce 
more than 95 percent of flavors consumed in the United States.
    To estimate dietary exposure to the synthetic flavoring substances, 
we used a ``per-capita times ten'' approach that conservatively assumes 
10 percent of the population consumes 100 percent of the available 
flavoring substance. Because the FEMA poundage data include the total 
poundage for both synthetic and naturally-sourced flavoring substances, 
our estimates of dietary exposure assumed that all of the flavoring 
substances added annually to food are synthetic; thus, for most of 
these substances, actual exposure to these synthetic flavoring 
substances is less than our conservative exposure estimates (Refs. 5 
and 6).
    As explained in more detail later in this section, although there 
were findings of carcinogenicity in animal studies, none of the data in 
our evaluations of the six synthetic flavoring substances supports a 
finding that they are human carcinogens when consumed at the levels of 
intended use. Additionally, with the exception of the data concerning 
methyl eugenol, the data from the animal studies demonstrated that the 
modes of action (MOA) of carcinogenicity are not acting through 
mechanisms of genotoxic alterations and are not relevant to humans.
    For methyl eugenol, the data showed evidence for a potential 
concern for carcinogenic risk to humans based on the findings that: (1) 
A metabolite of methyl eugenol was found to be genotoxic and able to 
covalently bind with DNA to form DNA adducts (a DNA adduct is a segment 
of DNA bound to a cancer causing chemical); (2) methyl eugenol-DNA 
adducts have been detected in human lung and liver tissues; and (3) 
there is a potential metabolic pathway by which methyl eugenol could 
metabolize to a reactive metabolite, under specific reaction conditions 
that then may proceed to tumor formation and carcinogenesis. However, 
there are no available clinical or epidemiological data reporting tumor 
formation and carcinogenicity from methyl eugenol exposure in humans.
    Additionally, we concluded that the risk of carcinogenicity in 
humans from consumption of methyl eugenol added to food as a synthetic 
flavoring substance is further reduced by the following mitigating 
factors: (1) The metabolic pathway, in which methyl eugenol converts to 
a genotoxic metabolite subsequently leading to tumor formation, does 
not serve as the primary metabolic/detoxification pathway for methyl 
eugenol in humans and the amount of the genotoxic metabolite generated 
is dose-dependent, occurring at higher doses and (2) compared to the 
low levels of added synthetic methyl eugenol as a flavoring substance, 
the levels of methyl eugenol tested in the NTP animal studies were very 
high test doses that likely overwhelmed physiological conditions of 
normalcy and overloaded systemic repair systems.
    In assessing the potential human carcinogenicity of methyl eugenol 
associated specifically with the use of synthetic methyl eugenol as a 
flavoring substance, we also considered data indicating that exposure 
to methyl eugenol from foods that naturally contain methyl eugenol 
(e.g., basil and other spices/herbs) is significantly higher 
(approximately 488 times higher) than exposure expected from the 
addition of synthetic methyl eugenol as a flavoring substance, and that 
these foods have been ingested by humans for millennia without apparent 
harm (Ref. 7). Based on our review of published literature up to May 
2018, there is no clinical or epidemiological evidence suggesting an 
association between the typical dietary consumption of food items that 
naturally contain methyl eugenol and carcinogenic effects.
    In sum, although the data do not indicate that these synthetic 
flavoring substances pose a public health risk as a human carcinogen, 
because these six synthetic flavoring substances have been found to 
induce cancer in animal studies, the Delaney Clause requires that FDA 
consider these synthetic flavoring substances to be unsafe as a matter 
of law, and FDA must revoke the listings providing for the use of these 
synthetic flavoring substances.
    Below is a summary of FDA's analysis of each of the six synthetic 
flavoring substances and adjuvants.

A. Benzophenone

1. Exposure
    Under Sec.  172.515, benzophenone is permitted for use as a 
synthetic flavoring substance and adjuvant in foods in accordance with 
current good manufacturing practices (CGMP). FEMA estimated an annual 
production volume of 5 kilograms (kg) for benzophenone used as a 
flavoring substance and adjuvant in food based on information from the 
2015 FEMA Poundage and Technical Effects Survey (Ref. 4). FEMA also 
estimated that 133 kg of benzophenone are available for consumption 
annually in the United States from its natural presence in foods (Ref. 
8). Thus, benzophenone is present from natural sources in the food 
supply (e.g., grapes) at a level 27 times greater than that from its 
use as a flavoring substance and adjuvant. Using the FEMA poundage data 
(assuming all reported poundage is for the synthetically-prepared 
flavoring substance) and a ``per-capita times ten'' approach, we 
estimated dietary exposure from benzophenone added to food as a 
synthetic flavoring and adjuvant to be 0.43 micrograms per person per 
day ([micro]g/p/d), or 7.2 x 10\-3\ [micro]g/kilogram body weight/d 
([micro]g/kg bw/d) for a 60 kg person (Refs. 6 and 9).
    Benzophenone also is permitted for use as a plasticizer in rubber 
articles intended for repeated use under Sec.  177.2600. The upper-
bound limit to the dietary exposure for benzophenone from this use is 
estimated to be 45 [micro]g/p/d. This estimate assumes that 100 percent 
of an individual's diet is processed using rubber articles containing 
benzophenone as a plasticizer. While the exposure estimate for the use 
of benzophenone as a plasticizer in repeat use rubber articles is an 
overestimate of the actual exposure from this use, the estimated 
exposure is greater than that from the use of benzophenone as a 
flavoring substance by a factor of approximately 500. Thus, the 
combined exposure to benzophenone from its uses as a flavoring 
substance and as a plasticizer in food contact applications was 
estimated to be no more than 45 [micro]g/p/d, or 0.75 [micro]g/kg bw/d 
(Refs. 5 and 9).
2. Toxicology Studies
    FDA reviewed data from 2 NTP-sponsored 105-week carcinogenic 
bioassays on benzophenone in F344/N rats and B6C3F1 mice. In these 
studies, the rats and mice were administered feed containing 
benzophenone at 0, 312, 625, or 1,250 parts per million per day (ppm/d) 
or milligrams per kilogram of feed/day (mg/kg/d). This dosing is 
equivalent to average daily doses of approximately 15, 30, and 60 mg 
benzophenone/kg bw to male rats and 15, 30, and 65 mg/kg bw to female 
rats; equivalent to average daily doses of approximately 40, 80, and 
160 mg/kg bw to male mice and 35, 70, and 150 mg/kg bw to female mice 
(Ref. 9).
    The NTP reported several carcinogenicity findings from these 
studies. They noted that there was some evidence of carcinogenicity due 
to increased incidence of renal (kidney) tubular tumors in treated male 
rats and increased incidence of mononuclear cell leukemia (MNCL) in all 
treated female rats. The mean incidence of MNCL in

[[Page 50495]]

the 625 ppm female dose group was significantly greater than that in 
the control female rats. The NTP also reported some evidence of 
carcinogenic activity in male mice based on increased incidence of 
hepatocellular (liver) neoplasms and some evidence of carcinogenicity 
in female mice based on increased incidence of histiocytic (originating 
from blood cells) sarcomas. Results showed that benzophenone produced 
tumors at the two highest doses in the studies. Occurrence of the key 
tumor types (i.e., those tumor types the NTP considered to constitute 
``some evidence'' of carcinogenicity) in animals at the lowest dose was 
not significantly different from that of the control groups. The NTP 
classified the occurrence of the key tumor types as constituting some 
evidence of carcinogenic activity rather than being clear evidence of 
carcinogenic activity (NTP's highest level of evidence of 
carcinogenicity). Benzophenone also was tested in several genotoxicity 
assays and found to be non-genotoxic.
    Based on results from the NTP studies, FDA concluded that, under 
the conditions of the 2-year NTP bioassays, benzophenone induced renal 
tubular tumors in male rats and hepatocellular tumors in male mice 
(Ref. 9).
3. Risk Characterization
    Based on the results of the NTP 2-year carcinogenicity studies we 
concluded that benzophenone induced cancer in animals under the test 
conditions of the studies. However, benzophenone is not genotoxic and 
unlikely to produce cancer through a direct DNA-reactive mechanism. 
Chronic progressive nephropathy (CPN, a spontaneous age-related disease 
that occurs commonly in rats) may be involved in benzophenone inducing 
renal tumors in rats; however, CPN as a MOA, a biologically plausible 
sequence of key events leading to an observed endpoint supported by 
robust experimental observations and mechanistic data (Ref. 10), for 
renal tumors in humans has not been established. Regarding the 
incidence of MNCL in female F344/N rats, we determined that it was not 
dose-dependent and that the incidence of this tumor in the control rats 
was outside the historical range. Therefore, we concluded that the 
occurrence of renal tumors in this study is not related to treatment 
with benzophenone. Additionally, MNCL is species- and strain-specific 
to the F344/N rat, and of little or no relevance to humans (Ref. 9).
    Regarding the results from the mouse study, several authors have 
observed that hepatocellular neoplasms seen in 2-year bioassays in 
B6C3F1 mice typically are secondary responses to chronic hepatic 
toxicity and regenerative cellular proliferation or hypertrophy as a 
function of dose (Ref. 9). Evidence of hepatotoxicity in short duration 
studies also has been shown to be a good predictor of hepatic neoplasia 
in chronic studies and the higher susceptibility of the male mouse 
(Ref. 9). Although there is no definitive MOA for the development of 
benzophenone-associated liver tumors in the NTP study, the B6C3F1 male 
mouse has been shown to have a high incidence of spontaneously-
occurring hepatocellular tumors, which is elevated after chemical 
exposure. Introduction of high doses of benzophenone may produce 
hepatotoxicity that exacerbates this propensity toward tumor 
development and results in their increased occurrence by a non-
genotoxic mechanism. Although rarely reported in NTP studies, 
histiocytic sarcomas observed in the B6C3F1 mice have been reported to 
occur at a mean incidence of 5.5 percent in female B6C3F1 mice used as 
controls in 2-year carcinogenicity studies conducted at Bayer AG, 
Institute of Toxicology. This result was based on historical data 
accumulated over a 10-year period (1986-1996) and is in line with the 6 
percent occurrence observed in the high dose (1,250 ppm) group in the 
benzophenone NTP study. Other authors also reported similar findings in 
B6C3F1 mice, with incidences of 3.5 percent and 5.5 percent in control 
males and females, respectively. Histiocytic sarcomas are rarely 
reported in humans, accounting for less than 1 percent of all the 
neoplasms reported in the lymph nodes or soft tissues. The histiocytic 
sarcomas identified in the female mice in the NTP study were not dose 
related (i.e., 5/50 at 625 ppm and 3/50 at 1,250 ppm) and were found 
only at dose levels that induced overt toxicity (Ref. 9).
    The lowest test dose (312 ppm) in the NTP 2-year studies was a dose 
at which no statistically significant treatment-related increase in 
tumor incidence was reported in rats or mice. This finding suggests 
that there may be a threshold level below which benzophenone does not 
induce tumors in rodents. Additionally, there is a large margin of 
exposure (MOE; 2.1 x 10\6\ for rats, 4.7 x 10\6\ for male mice, and 5.6 
x 10\6\ for female mice) between the lowest test dose and the estimated 
dietary exposure of 0.43 [micro]g benzophenone/p/day (equivalent to 7.2 
x 10 -\3\ [micro]g/kg bw/day) from its use as a flavoring 
substance. When benzophenone is used as a plasticizer in repeat use 
rubber articles exposed to food, the MOE for male and female rats is 
calculated to be 2 x 10\4\ and for male and female mice, 5.3 x 10\4\ 
and 4.7 x 10\4\, respectively. Although these MOE values are lower than 
those for benzophenone's use as a synthetic flavoring substance, they 
are still sufficient to ensure an acceptable margin of safety (Ref. 9). 
It should also be noted that these results are based on estimated 
worst-case dietary exposure of 45 [micro]g/person/d (0.75 [micro]g/kg 
bw/d) from its use as a plasticizer (Ref. 5) and actual MOEs for this 
use probably would be higher. Considering these findings in a weight-
of-evidence analysis, we concluded that benzophenone is unlikely to 
induce tumors in humans at current use levels as a synthetic flavoring 
substance and adjuvant in food (Ref. 9).

B. Ethyl Acrylate

1. Exposure
    Under Sec.  172.515, ethyl acrylate is permitted for use as a 
synthetic flavoring substance and adjuvant in foods in accordance with 
CGMP. FEMA estimated an annual production volume of 18 kg for ethyl 
acrylate used as a flavoring substance and adjuvant in food based on 
information from the 2015 FEMA Poundage and Technical Effects Survey 
(Ref. 4). FEMA also estimated that 9.2 kg of ethyl acrylate are 
available for consumption annually in the United States from its 
natural presence in foods (e.g., pineapple) (Ref. 8). Thus, ethyl 
acrylate is present in foods from natural sources at 50 percent of the 
level from its use as a flavoring substance. Using the FEMA poundage 
data (assuming all reported poundage is for the synthetically-prepared 
flavoring substance) and a ``per-capita times ten'' approach, we 
estimated dietary exposure from ethyl acrylate's use as a synthetic 
flavoring substance and adjuvant in food to be 1.5 [micro]g/person/d, 
or 0.025 [micro]g/kg bw/d for a 60 kg person (Refs. 6 and 11).
2. Toxicology Studies
    FDA reviewed data from 2 NTP-sponsored 103-week carcinogenic 
bioassays on ethyl acrylate in F344/N rats and B6C3F1 mice. In these 
studies, rats and mice were administered ethyl acrylate at 0, 100, or 
200 mg/kg bw by gavage 5 days per week. The NTP reported 
carcinogenicity findings were confined to the forestomach of rats and 
mice. They also reported that the occurrence of these forestomach 
tumors had a statistically positive trend compared to the control 
animals. Ethyl acrylate also was tested in several genotoxicity 
studies. Based on the available data from these studies, we

[[Page 50496]]

concluded that ethyl acrylate is not genotoxic (Ref. 11).
    We also concluded that under the test conditions of NTP's 2-year 
hazard assessment studies ethyl acrylate is a rodent carcinogen. 
Evidence, however, supports the findings that these tumors were 
produced by a non-genotoxic mechanism (Ref. 11).
3. Risk Characterization
    The tumors observed in the NTP study were initiated by 
administering bolus doses of ethyl acrylate by gavage onto the 
forestomach of the treated rats and mice, which resulted in irritation, 
inflammation, and hyperplasia of the forestomach mucosa. Repeated 
dosing over a 2-year period exacerbated this irritation and resulted in 
the development of papillomas and carcinomas, which were confined to 
the forestomach. No other treatment-related tumors were observed in the 
animals. Forestomach tumors were observed at both doses tested (100 mg/
kg bw and 200 mg/kg bw) in both male and female mice and rats. Humans 
do not have a forestomach and a human counterpart for the forestomach 
does not exist. The function of the rodent forestomach is to store and 
concentrate feed; therefore, high concentrations of ethyl acrylate were 
present in the forestomach over the duration of the 2-year study. This 
concentration effect precluded our determining a no-significant-effect-
level for the occurrence of the forestomach tumors. Therefore, we 
cannot make an MOE comparison between a no-effect-dose level for 
significant incidences of tumors and the estimated dietary exposure of 
ethyl acrylate as a synthetic flavoring substance and adjuvant in food 
(1.5 [mu]g ethyl acrylate/p/d, or 0.025 [micro]g/kg bw/d) (Ref. 11).
    The 2-year NTP studies were conducted at doses higher than the 
expected exposures for flavoring substances. In general, flavoring 
substances have significantly lower dietary exposures than the doses 
used in 2-year carcinogenicity studies. For example, the lowest dose of 
ethyl acrylate tested in the NTP studies was 100 mg/kg bw, or 
approximately 1.8 x 10 \6\ times greater than the estimated dietary 
exposure from its use as a synthetic flavoring substance and adjuvant 
in food (Ref. 11).
    Importantly, the NTP Board of Scientific Counselors Report on 
Carcinogens (RoC) Subcommittee concluded, based on the totality of the 
evidence, that ethyl acrylate should not be considered a human 
carcinogen (Ref. 12). We concur with the RoC and concluded that ethyl 
acrylate is a non-genotoxic rodent carcinogen with a carcinogenic 
effect limited to the rodent forestomach (a rodent-specific organ) due 
to chronic irritation. This MOA is not relevant to humans and, at the 
current intake level, there is no concern of carcinogenicity from the 
intake of ethyl acrylate intentionally added to food as a flavoring 
substance and adjuvant (Ref. 11).

C. Eugenyl Methyl Ether (Methyl Eugenol)

1. Exposure
    Under Sec.  172.515, methyl eugenol is permitted for use as a 
synthetic flavoring substance and adjuvant in foods in accordance with 
CGMP. FEMA estimated an annual production volume of 86 kg for methyl 
eugenol used as a flavoring substance and adjuvant in food based on 
information from the 2015 FEMA Poundage and Technical Effects Survey 
(Ref. 4). FEMA also estimated that 447,450 kg of methyl eugenol are 
available for consumption annually in the United States from its 
natural presence in foods (e.g., basil) (Ref. 8). The 69th Joint Food 
and Agriculture Organization/World Health Organization (WHO) Expert 
Committee on Food Additives (JECFA) estimated an upper bound annual 
volume for methyl eugenol of 41,992 kg from its natural presence in 
herbs and spices. The most significant difference between the two 
estimates is that FEMA presumed a maximum content of methyl eugenol in 
basil of 4.1 percent, whereas JECFA presumed a maximum content of 0.118 
percent (Refs. 5 and 8). Natural sources of basil have varying levels 
of methyl eugenol. It is unlikely, however, that most basil used in the 
United States would consistently have levels as high as 4.1 percent 
and, as such, JECFA's estimate of the amount of methyl eugenol from 
natural sources is suitably conservative and representative of probable 
consumption. Using the JECFA estimate, methyl eugenol is estimated to 
be present in the food supply from natural sources at a level 488 times 
greater than that from its use as a synthetic flavoring substance or 
adjuvant in food. Using the FEMA poundage data (assuming all reported 
poundage is for the synthetically prepared flavor) and a ``per-capita 
times ten'' approach, we estimated dietary exposure from methyl 
eugenol's use as a synthetic flavoring substance and adjuvant in food 
to be 7.4 [micro]g/person/d, or 0.12 [micro]g/kg bw/d for a 60 kg 
person (Refs. 6 and 13).
2. Toxicology Studies
    FDA reviewed data from 2 NTP-sponsored 2-year carcinogenicity 
bioassays on methyl eugenol in F344/N rats and B6C3F1 mice. In these 
studies, methyl eugenol was administered to the animals at 0, 37, 75, 
or 150 mg/kg bw by gavage, 5 days per week, for 105 weeks. These test 
doses are 220,000 to 890,000 times higher than the estimated human 
dietary exposure from its use as a flavoring substance.
    The NTP reported significantly increased incidence of liver tumors 
(combined adenomas or carcinomas), compared to the concurrent control 
groups, occurring in a dose-dependent manner across the treatment 
groups in both genders of rats and mice. Although the mortality in some 
treated groups was higher than 50 percent, tumors were the main cause 
of death in these groups. Further, most deaths occurred late in the 
studies. Another type of tumor, glandular stomach neuroendocrine 
neoplasms, were found in both genders of rats, but in only two male 
mice. The NTP, JECFA, and FDA do not consider these glandular stomach 
neuroendocrine neoplasms relevant to tumor formation in humans due to 
considerations of the mechanism of development of these neoplasms. 
Based on the overall data, we concluded that methyl eugenol, under the 
test conditions of the NTP 2-year carcinogenicity bioassays, induced 
cancer in rodents (Ref. 13).
    Regarding the genotoxicity potential of methyl eugenol, results 
from several genotoxicity assays were negative; however, in testing 
systems that provided adequate metabolic activation, specifically 1'-
hydroxylation and sulfonation, or those systems directly testing the 
1'-hydroxyl metabolite of methyl eugenol, positive genotoxic effects 
were observed.
    There is evidence showing that methyl eugenol treatment leads to 
the formation of covalent DNA adducts in vitro and in vivo. In cancer 
risk assessment, the formation of DNA adducts is a biomarker of 
exposure and suggestive of potential cancer risk. However, the 
observation of adducts itself should not be used to predict cancer. The 
relevance of DNA adducts for cancer assessment should be investigated 
in the context of other information, such as the quantity and 
persistency of the adducts. The level of methyl eugenol-specific 
adducts was shown to be dose-dependent in experimental animals. 
Therefore, since human dietary consumption of methyl eugenol from use 
as a synthetic flavoring substance in food is much lower than the dose 
received by the animals in the NTP studies, much lower levels of DNA 
adducts would be formed in humans compared to that in the test

[[Page 50497]]

animals. Additionally, there is evidence that the formation of these 
adducts requires specific metabolic activation of methyl eugenol (i.e., 
hydroxylation followed by sulfonation, leading to the formation of 1'-
sulfooexymethyleugenol, the ultimate metabolite that binds to DNA). 
Based on the physiology-based pharmacokinetic model of methyl eugenol, 
this pathway is not a major metabolic pathway in humans. Even after 
hydroxylation occurs, the hydroxylated intermediates can be eliminated 
by glucuronization and oxidation, so that only a trace amount of 
ingested methyl eugenol is metabolized to 1'-sulfooexymethyleugenol. In 
regard to the persistence of the adducts, there is evidence showing 
that in rats given methyl eugenol, the levels of methyl eugenol-
specific adducts reduced after the treatment was stopped, suggesting 
that these adducts are repairable with considerable low persistency 
(Ref. 13).
    There are only few studies measuring methyl eugenol-specific DNA 
adducts in humans. The adducts have been detected in 150 of 151 human 
liver biopsy samples and 10 of 10 tested human lung biopsy samples, 
indicating that the bioactive metabolites form in these subjects with 
typical dietary exposure, and are capable of binding with human DNA. 
However, these human data have limitations. We note that all but one 
the human tissue donors in these studies were patients with cancer or 
chronic liver diseases, who may have DNA-repair deficiencies, 
compromised detoxification pathways, or weakened control mechanisms 
that prevent the promotion of carcinogenesis from DNA adducts, whereas 
such control mechanisms would be expected to be operable in healthy 
humans. Therefore, it is difficult to extrapolate DNA-adduct results 
found in these unhealthy subpopulations to the general healthy 
population (Ref. 13).
3. Risk Characterization
    In our evaluation of the carcinogenic potential of methyl eugenol 
in humans using a weight-of-evidence approach, we concluded that a 
genotoxic MOA is likely involved in the carcinogenicity observed in the 
NTP animal studies. This MOA involves formation of a bioactivated 
metabolite that forms DNA-adducts that leads to subsequent cancer 
initiation and development. Current scientific data on methyl eugenol 
suggest that bioactivation to the DNA-reactive metabolite, DNA adduct 
formation, and subsequent tumor formations are dose-dependent. Although 
methyl eugenol-specific DNA adducts have been identified in 
hospitalized subpopulations, there are no clinical or epidemiological 
data that provide concrete evidence that methyl eugenol is a human 
carcinogen. In the general healthy population, DNA-repair mechanisms 
and damage-response pathways may effectively prevent cancer development 
from an initiation event such as a DNA adduct. Therefore, the extremely 
low level of DNA adducts formed in humans from dietary exposure to 
methyl eugenol as an added food flavoring substance likely is below a 
threshold level necessary for subsequent cancer development. However, 
the current science is inadequate to quantitate the carcinogenic 
potential risk (if any) of methyl eugenol in humans (Ref. 13).
    Carcinogenicity data on methyl eugenol also demonstrated that non-
genotoxic MOAs for the observed tumors in animals, especially in mice, 
may be operating in conjunction with the genotoxic MOA. However, data 
for the non-genotoxic MOA are insufficient (Ref. 13).
    The MOE for synthetic methyl eugenol as a flavoring substance and 
adjuvant in food is very large. Two dose-response assessments have been 
conducted to derive a point of departure for the liver carcinogenicity 
of methyl eugenol; both derived a lower bound benchmark dose 
(BMDL10) based on data from the NTP bioassays. Using the 
more conservative BMDL10 (7.7 mg/kg/d), and the estimated 
dietary exposure of methyl eugenol as a flavoring substance (0.12 
[micro]g/kg bw), the MOE is approximately 6.4 x 10 \4\. This MOE is 
based on an estimated dietary exposure that assumed 100 percent of the 
reported poundage data are exclusively synthetic methyl eugenol. Thus, 
the actual MOE for synthetically prepared methyl eugenol added to foods 
likely is larger. Although the carcinogenic potential cannot be 
definitively ruled out, this large MOE translates into a very small 
risk for carcinogenicity in humans and a low public health concern 
(Ref. 13).
    As for methyl eugenol from natural sources, other components in 
such sources may modulate bioactivation and/or detoxification, so the 
toxicity data related to the use as a synthetic flavoring substance may 
not be relevant to its presence from natural sources. For example, a 
flavonoid derived from basil extracts, nevadensin, was found to be a 
sulfotransferase inhibitor, and it significantly reduced methyl 
eugenol-induced DNA adduct formation in F344/N rats (Ref. 13).
    In conclusion, although there is evidence of genotoxicity for a 
bioactive metabolite of methyl eugenol, we concluded based on currently 
available scientific evidence that, despite the potential carcinogenic 
concern and lack of definitive quantitative cancer risk measurement, 
such risk in humans is mitigated by factors such as low exposure from 
its use as a flavoring substance, pharmacokinetics/metabolism, DNA-
repair mechanisms, and the lack of clinical and epidemiological 
evidence of the carcinogenic effect in humans from oral exposure to 
methyl eugenol. Therefore, it is unlikely that consumption of methyl 
eugenol presents a risk to public health from use as a flavoring 
substance.

D. Myrcene

1. Exposure
    Under Sec.  172.515, myrcene is permitted for use as a synthetic 
flavoring substance and adjuvant in foods in accordance with CGMP. FEMA 
estimated an annual production volume of 860 kg for myrcene used as a 
flavoring substance and adjuvant in food based on information from the 
2015 FEMA Poundage and Technical Effects Survey (Ref. 4). FEMA also 
estimated that 14,177,215 kg of myrcene are available for consumption 
annually in the United States from its natural presence in foods (e.g., 
citrus juices) (Ref. 8). Thus, myrcene is present naturally in foods at 
a level 16,500 times greater than that from use as a flavoring 
substance and adjuvant. We estimated dietary exposure to myrcene as a 
synthetic flavoring substance using the FEMA poundage data (assuming 
all reported poundage is for the synthetically prepared flavoring 
substance) and a ``per-capita times ten'' approach to be 74 [micro]g/
person/d, or 1.23 [micro]g/kg bw/d for a 60 kg person (Refs. 6 and 14).
2. Toxicology Studies
    FDA reviewed data from 2 NTP-sponsored carcinogenicity bioassays on 
myrcene ([beta]-myrcene) in F344/N rats and B6C3F1 mice. In the rat 
study, male and female rats were administered 0, 0.25, 0.50 or 1.0 g 
myrcene/kg bw by gavage, 5 days per week for up to 105 weeks. Results 
from the study showed increased incidence of renal tubule tumors in 
both sexes. All high dose (1 g/kg bw) male rats died prior to the end 
of the study due to renal toxicity. Incidence of nephrosis were 
significantly increased in all dosed male and female rats when compared 
to controls. Incidence of CPN were significantly increased in all 
myrcene-treated female rats but not male rats. There also was 
significantly increased incidence of nephrosis in all myrcene-

[[Page 50498]]

treated male and female rats compared to controls. However, incidence 
of mineralization of renal papilla also was significantly increased in 
all dosed male rats but not in female rats. Based on increased 
incidence of renal tubule neoplasms, NTP concluded that there was clear 
evidence of carcinogenic activity of myrcene in male F344/N rats and 
equivocal evidence of carcinogenic activity of myrcene in female rats 
(Ref. 14).
    In the NTP mouse study, male and female mice were administered 0, 
0.25, 0.50 or 1.0 g myrcene/kg bw by gavage, 5 days per week for up to 
104 (females) and 105 weeks (males). Based on increased incidence of 
liver neoplasms, NTP concluded that there was clear evidence of 
carcinogenic activity of myrcene in male mice and equivocal evidence of 
carcinogenic activity of myrcene in female mice (Ref. 14).
    Myrcene also was tested in several in vivo and in vitro 
genotoxicity assays sponsored by the NTP. The NTP concluded that 
myrcene was not genotoxic based on the negative Ames assays (Salmonella 
typhimurium (S. typhimurium) and Escherichia coli (E. coli)) and in 
vivo micronucleus assays in male and female B6C3F1 mice (Ref. 14).
    Based on our evaluation of the data in the NTP 2-year myrcene 
studies, we concluded that, under the test conditions of the studies, 
myrcene induced renal tubular tumors in F344/N rats and hepatocellular 
tumors in B6C3F1 mice. We also concluded that myrcene is non-genotoxic 
(Ref. 14).
3. Risk Characterization
    Our review of relevant scientific data and information suggests 
that myrcene may be operating through multiple MOAs to induce kidney 
and liver tumors in rodents. While, a definitive MOA for the induction 
of tumors by myrcene in rodents has not been established, because 
myrcene is not genotoxic, the induction of rodent tumors likely is 
occurring through an indirect non-DNA mediated MOA. One potential MOA 
in male and female rats is an unusual nephrosis. Another potential MOA, 
[alpha]2u-globulin (a low molecular-weight protein synthesized in the 
male rat liver) hyaline nephropathy, and renal tubular hyperplasia may 
collectively contribute to the development of renal tubule neoplasia in 
male rats following myrcene treatment (the [alpha]-2u-globulin 
nephropathy occurs only in male rats and is not operative in humans) 
(Ref. 14).
    The B6C3F1 mouse strain used in the NTP-sponsored study with 
myrcene is known to have a high spontaneous background incidence of 
liver neoplasms and is a sensitive strain for the induction of liver 
tumors. The observed hepatocellular tumors in myrcene-dosed mice 
exceeded concurrent and historical controls. The MOA for the induction 
of hepatocellular tumors in myrcene dosed mice is not well understood. 
We are not aware of any robust mechanistic studies conducted to 
determine the MOA(s) responsible for the induction of hepatocellular 
neoplasia reported in myrcene-treated mice (Ref. 14).
    In the NTP 2-year rat study, increased incidence of renal tubular 
tumors was observed in all doses of myrcene treated male rats. Because 
a no significant effect dose level was not observed in this study, we 
derived a BMDL10 of 64,000 [micro]g/kg bw/d based on the 
most sensitive endpoint, the combined renal tubular adenomas and 
carcinomas in male rats. Based on this BMDL10 and the 
estimated dietary exposure to myrcene, we calculated an MOE of 5.2 x 10 
\4\ (Ref. 14).
    Using a weight-of-evidence analysis, we concluded that myrcene is 
unlikely to induce tumors in humans at its current exposure level when 
used as a synthetic flavoring substance and adjuvant in food based on 
the following: (1) Myrcene is non-genotoxic; (2) the MOA for kidney 
tubule tumors likely involves multiple MOAs that may include renal 
toxicity (nephrosis), [alpha]2u-globulin nephropathy (a mechanism not 
operative in humans), and hyperplasia in male rats. In female rats, 
nephrosis and hyperplasia are likely MOAs; (3) B6C3F1 mice are prone to 
spontaneous hepatocellular adenomas, carcinomas, and hepatoblastomas 
with high background tumor incidence, and (4) a MOE of 5.2 x10 \4\ 
indicates a low risk concern from a public health point of view (Ref. 
14).

E. Pulegone

1. Exposure
    Under Sec.  172.515, pulegone is permitted for use as a synthetic 
flavoring substance and adjuvant in foods in accordance with CGMP. FEMA 
estimated an annual production volume of 6 kg for pulegone used as a 
flavoring substance and adjuvant in food based on information from the 
2015 FEMA Poundage and Technical Effects Survey (Ref. 4). FEMA 
estimated that 866 kg of pulegone are available for consumption 
annually in the U.S. from its natural presence in foods (e.g., mint) 
(Ref. 8). Thus, pulegone is present from natural sources in the food 
supply at a level 144 times greater than that from use as a flavoring 
substance and adjuvant. Using FEMA poundage data (assuming all reported 
poundage is for the synthetically prepared flavor) and a ``per-capita 
times ten'' approach, we estimated dietary exposure from pulegone's use 
as a synthetic flavoring substance and adjuvant in food to be 0.5 
[micro]g/person/d, equivalent to 0.008 [micro]g/kg bw/d for a 60 kg 
person (Refs. 6 and 15).
2. Toxicology studies
    FDA reviewed data from 2 NTP-sponsored 2-year carcinogenicity 
bioassays on pulegone in F344/N rats and B6C3F1 mice. In the rat study, 
pulegone was administered by gavage at 0, 18.75, 37.5, or 75 mg 
pulegone/kg bw to male rats and 0, 37.5, 75, or 150 mg pulegone/kg bw 
to female rats 5 days a week for up to 104 weeks. The NTP reported 
that, in female rats, the primary tumors observed were urinary bladder 
papillomas and carcinomas. In male rats, no urinary bladder neoplasms 
were reported. Only transitional epithelial hyperplasia was observed in 
the pulegone-treated male rats at the lowest dose tested; no epithelial 
hyperplasia was observed in male rats at the mid or high doses. 
Pulegone administration also was associated with the occurrence of non-
neoplastic lesions in the liver and nose of male and female rats, and 
in the forestomach of male rats. The NTP concluded that under the 
conditions of the experiment, there was no evidence of carcinogenic 
activity of pulegone in male F344/N rats and clear evidence of 
carcinogenic activity of pulegone in female F344/N rats based on 
increased incidence of urinary bladder neoplasms.
    In the mouse study, pulegone was administered by gavage at 0, 37.5, 
75 or 150 mg/kg bw 5 days a week for 105 weeks. The NTP reported that 
the primary tumors observed in the study were liver neoplasms in male 
and female mice. The NTP concluded that under the conditions of the 
experiment, there was clear evidence of carcinogenic activity of 
pulegone in male and female B6C3F1 mice.
    Pulegone also was tested in several in vitro and in vivo 
genotoxicity assays. Overall, results were mostly negative. However, 
NTP concluded that pulegone is genotoxic based on a single positive 
result in the Ames Assay in S. typhimurium strain TA 98 and E. coli 
strain WP2 uvrA/PKM101 in the presence of metabolic activation.
    Based on the findings of statistically significant increased 
incidence of urinary bladder papilloma and carcinoma in female F344/N 
rats and liver neoplasms in B6C3F1 male and female mice in the 2-year 
NTP

[[Page 50499]]

bioassays, we concluded that under the conditions of the 2 NTP studies, 
pulegone is a rodent carcinogen. Based on the totality of evidence from 
available genotoxicity studies, we also concluded that pulegone is 
likely non-genotoxic (Ref. 15).
3. Risk Characterization
    According to NTP, the dose-related increase in the incidence of 
urinary bladder neoplasms in female rats was most likely related to the 
genotoxic activity of pulegone. However, we concluded that pulegone 
likely is non-genotoxic based on negative results in the majority of 
genotoxicity studies, along with a lack of available evidence reporting 
that DNA adducts related to pulegone treatments are formed. This 
suggests that the urinary bladder neoplasms observed in female F344/N 
rats treated with pulegone were caused by a non-genotoxic MOA.
    Urinary bladder carcinogenesis likely is occurring in the rat 
through cytotoxicity as a result of chronic exposure to high 
concentrations of pulegone and its metabolites, followed by 
regenerative urothelial cell (a cell type that lines much of the 
urinary tract) proliferation, that further led to urothelial tumors 
(Ref. 15). Da Rocha et al. (2012) (Ref. 16) concluded that the 
carcinogenic MOA for urinary bladder tumors was not relevant to humans, 
based on the assertion that humans would never be exposed to pulegone 
long enough to develop hyperplasia because pulegone is highly volatile, 
noxious, and a nasal irritant, and that genotoxicity of pulegone has 
not been demonstrated.
    The metabolic fate of pulegone has been studied extensively in 
rodents and is well understood. Pulegone is metabolized by multiple 
pathways in the rodent. One important intoxication (bioactivation) 
pathway involves the formation of menthofuran, the proximate toxic 
metabolite of pulegone, which is further oxidized in the liver to yield 
[gamma]-ketoenal, 8-pulegone aldehyde. [gamma]-ketoenal, 8-pulegone 
aldehyde is the ultimate toxic metabolite of pulegone in rodents. In 
general, at dose levels at or below 80 mg/kg bw, cellular 
concentrations of pulegone and its metabolites are effectively 
detoxified by conjugation with glutathione and glucuronic acid in 
rodents (Ref. 15).
    In a human metabolism study in which pulegone was administered 
orally at doses of 0.5 to 1 mg/kg bw, 10-hydroxypulegone, not 
menthofuran, was the major metabolite. In this study, 10-
hydroxypulegone was conjugated with glucuronic acid or sulfuric acid 
and detoxified. Based on the limited, available human metabolism data, 
the toxic metabolite of pulegone, menthofuran, is not formed at 
toxicologically significant levels in humans at the dietary exposure 
levels expected from the use of pulegone as a flavoring substance (Ref. 
15).
    Protein adduct formation and glutathione depletion have been 
postulated as potential MOAs of pulegone via menthofuran formation, 
which could cause cytotoxicity and chronic cell proliferation, and 
ultimately lead to liver neoplasms. In vivo and in vitro studies showed 
an association between hepatocellular damage caused by menthofuran and 
its metabolic activation to [gamma]-ketoenal, 8-pulegone aldehyde and 
covalent binding to target organ proteins. Further, p-cresol, another 
pulegone metabolite produced in rodents given high doses of pulegone, 
depletes glutathione levels. This may lead to chronic regenerative cell 
proliferation, which may be related to the liver carcinogenicity 
observed in experimental B6C3F1 mice (Ref. 15)
    Considering genotoxicity data, metabolism, MOA, and the sensitivity 
of the B6C3F1 strain to develop hepatocellular tumors, the mouse liver 
tumors likely are not relevant to humans at the current use level of 
pulegone as a synthetic flavoring substance and adjuvant in food (Ref. 
15).
    An MOE was calculated using the no-significant effect level at 
which no treatment-related tumors were reported in the 2-year NTP mouse 
study of pulegone in male rats (i.e., no significant effect level 
(18.75 mg/kg bw, equivalent to 13.39 m g/kg bw/day)). This dose was 
selected because in female rats, combined incidence of urinary bladder 
papilloma or carcinoma (a rare tumor) was significantly increased at 
the high dose (150 mg/kg bw), exceeding historical control ranges for 
2-year corn oil gavage studies and concurrent controls. In male mice, 
the incidence of hepatocellular adenomas in the 37.5 mg/kg bw dose 
group exceeded that in the concurrent and historical control ranges for 
2-year corn oil gavage studies. In addition, in female mice, the 
incidence of hepatocellular adenomas in the 37.5 mg/kg bw dose group 
exceeded that in the concurrent and historical control ranges for 2-
year corn oil gavage studies. Although not statistically significant, 
these occurrences may be biologically relevant, given that they 
exceeded those of the historical and concurrent controls, and there 
were statistically significant increases in some proliferative non-
neoplastic lesions in the liver of male mice at this dose. The MOE 
based on the estimated dietary exposure of 0.5 [micro]g/p/d (equivalent 
to 0.008 [micro]g/kg bw/d) for pulegone as a flavoring substance in 
humans is 1.7 x 10 \6\, which indicates a very low potential 
carcinogenic risk for humans (Ref. 15).
    Using a weight-of-evidence analysis considering that: (1) Pulegone 
is non-genotoxic; (2) pulegone has a potential cytotoxicity MOA; (3) 
available data suggest a dose-dependent, metabolic activation of 
pulegone in humans and rodents, an indication of a threshold effect; 
(4) there is a no-significant effect level below which no tumors were 
formed in the 2 NTP year studies; (5) dietary exposure from use as a 
synthetic flavoring substance added to food is low with a MOE of 1.7 x 
10 \6\, we concluded that pulegone at its current use level as a 
synthetic flavoring substance and adjuvant in food, is unlikely to 
induce urinary bladder cancer and liver neoplasms in humans and does 
not pose a public health concern (Ref. 15).

F. Pyridine

1. Exposure
    Under Sec.  172.515, pyridine is permitted for use as a synthetic 
flavoring substance and adjuvant in foods in accordance with CGMP. FEMA 
estimated an annual production volume of 27 kg for pyridine used as a 
flavoring substance and adjuvant in food based on information from the 
2015 FEMA Poundage and Technical Effects Survey (Ref. 4). FEMA also 
estimated that 73,861 kg of pyridine are available for consumption 
annually in the U.S. from its natural presence in foods (e.g., coffee) 
(Ref. 8). Thus, pyridine is present from natural sources in the food 
supply at a level 2,736 times greater than that from use as a flavoring 
substance. Using the FEMA poundage data (assuming all reported poundage 
is for the synthetically prepared flavoring substance) and a ``per-
capita times ten'' approach, we estimated dietary exposure from 
pyridine's use as a synthetic flavoring substance and adjuvant in food 
to be 2.3 [micro]g/person/day, or 0.038 [micro]g/kg bw/d for a 60 kg 
person (Refs. 6 and 17).
2. Toxicology studies
    FDA reviewed data from 3 NTP-sponsored 2-year carcinogenicity 
bioassays on pyridine in F344/N rats, Wistar rats, and B6C3F1 mice. In 
the F344/N rat study, pyridine was administered in drinking water at 0, 
100, 200, or 400 ppm (mg pyridine/kg drinking water) for 104 (males) 
and 105 (females) weeks. These dose levels were equivalent to doses of 
7, 14, or 33 mg pyridine/kg bw/d, respectively. The

[[Page 50500]]

NTP reported a statistically significant increased incidence of renal 
tubule adenomas and renal tubule hyperplasia only in the high dose 
F344/N male rats. In addition, NTP reported significantly elevated 
incidences of MNCL in F344/N female rats at the 200 ppm and 400 ppm 
dose levels. MNCL is a commonly occurring spontaneous neoplasm in 
untreated, older F344/N rats. One study found that MNCL occurs in 
untreated, aged F344/N rats at a high and variable rate; that MNCL as a 
lesion is uncommon in most other rat strains; and the background 
incidence of MNCL in F344/N rats has increased significantly over the 
years (Ref. 17).
    Recognizing the species specificity and high background levels of 
MNCL in F344/N rats, the NTP conducted a 2-year carcinogenicity study 
in male Wistar rats (a rat strain that does not have a high background 
of MNCL neoplasms). In this study, pyridine was administered in 
drinking water at 0, 100, 200, or 400 ppm for 104 weeks to male Wistar 
rats. These dose levels were equivalent to doses of 8, 17, or 36 mg 
pyridine/kg bw/d. The study showed no increased incidences of MNCL in 
any of the treatment groups. The NTP reported a statistically 
significant increased incidence of interstitial cell adenomas in the 
400 ppm dose group. Observed increased incidence of interstitial cell 
adenomas of the testes in Wistar rats exposed to 400 ppm pyridine were 
marginally above the historical control range. A statistically 
significant increased incidence of kidney hyperplasia was observed at 
the 100 ppm dose group, along with increased incidence of kidney 
adenomas that were not statistically significant. There also was 
increased incidence of nephropathy in all pyridine-treated Wistar rats 
as well as in the controls (Ref. 17).
    The NTP concluded that under the conditions of the 2-year F344/N 
rat oral drinking water study there was some evidence of carcinogenic 
activity of pyridine in male F344/N rats based on increased incidence 
of renal tubule neoplasms and equivocal evidence of carcinogenic 
activity of pyridine in female F344/N rats based on increased incidence 
of MNCL. The NTP considered the increased incidence of interstitial 
cell adenomas of the testes in the Wistar rat study to be equivocal 
evidence for carcinogenicity.
    In the mouse study, pyridine was administered in drinking water to 
male B6C3F1 mice at concentrations of 0, 250, 500 or 1,000 ppm (doses 
equivalent to 35, 65, or 110 mg pyridine/kg bw/d, respectively) for 104 
weeks. Groups of female B6C3F1 mice were administered pyridine at doses 
of 0, 125, 250 or 500 ppm (doses equivalent to 15, 35, or 70 mg 
pyridine/kg bw/d, respectively) in drinking water for 105 weeks. The 
NTP reported statistically significant increased incidence of 
hepatocellular carcinomas at all dose levels in the male and female 
mice and concluded that there was clear evidence of carcinogenic 
activity of pyridine in male and female B6C3F1 mice.
    Pyridine also was tested in several in vitro and in vivo 
genotoxicity assays. The NTP concluded that pyridine was non-genotoxic. 
Based on evidence from available studies, we also concluded that 
pyridine is non-genotoxic (Ref. 17).
    Under the test conditions of the 2-year NTP studies, we concluded 
that pyridine is a rodent carcinogen based on the observed pyridine-
induced renal tumors in male F344/N rats and pyridine-induced liver 
tumors in B6C3F1 mice (Ref. 17).
3. Risk Characterization
    Our review of relevant scientific data and information suggests 
that pyridine may be operating through multiple MOAs in its capability 
to induce kidney and liver tumors in rodents. A definitive MOA for the 
induction of tumors in rodents has not been established. However, 
because pyridine is not genotoxic, the induction of rodent tumors 
likely is occurring through an indirect non-DNA mediated MOA.
    While NTP discounted the kidney neoplasms observed in the F344/N 
rats as being associated with an [alpha]2[micro]-globulin MOA, we 
concluded that pyridine may be a weak inducer of [alpha]2[micro]-
globulin in F344/N male rats, based on the observation of statistically 
significant increased incidence in granular casts and hyaline 
degeneration in the 1000 ppm pyridine-treated rats along with higher 
staining intensity for [alpha]2[micro]-globulin in the kidney tissues 
from F344/N male rats exposed to 1000 ppm pyridine (Ref. 17).
    Using a weight-of-evidence analysis, we concluded that pyridine is 
unlikely to induce tumors in humans at its current exposure level as a 
synthetic flavoring substance and adjuvant in foods based on the 
following: (1) Pyridine is non-genotoxic; (2) renal tubule neoplasms 
likely involve multiple MOAs that may include [alpha]2[micro]-globulin 
nephropathy and CPN, which are not relevant to humans. These postulated 
mechanisms, specifically [alpha]2[micro]-globulin nephropathy, are 
species- and sex-specific; (3) B6C3F1 mice are prone to spontaneous 
hepatocellular adenomas, carcinomas, and hepatoblastomas with high 
background incidence; and (4) active metabolites of pyridine differ 
across species and appear to be dose-dependent.
    Further, there is a large MOE (3.7 x 10\5\) between the estimated 
dietary exposure of pyridine as a synthetic flavoring substance 
intentionally added to food (0.038 [micro]g/kg bw/d) compared to the 
highest dose of pyridine at which no treatment-related, statistically 
significant tumors were observed in the NTP studies (14,000 [micro]g/kg 
bw/d (rats)) (Ref. 17). This large MOE further supports our conclusion 
that pyridine, when used as a flavoring substance, is unlikely to 
induce cancer in humans.

IV. Comments on the Notice of Petition

    FDA received a number of comments in response to the notice of the 
petition. Most comments expressed general support for revocation of the 
regulations for the seven synthetic flavoring substances, without 
providing any additional information. Several comments expressed 
concern about the safety of these synthetic flavoring substances and 
asked that FDA ban them from foods; however, these comments did not 
provide any information to support their claim that the use of these 
additives is unsafe.
    We summarize and respond to relevant portions of comments in this 
final rule. To make it easier to identify comments and FDA's responses 
to the comments, the word ``Comment'' will appear in parentheses before 
the description of the comment, and the word, ``Response'' will appear 
in parentheses before FDA's response. We have also numbered each 
comment to make it easier to identify a particular comment. The number 
assigned to each comment is for organizational purposes only and does 
not signify the comment's value, importance, or the order in which it 
was submitted.

A. Legal and Policy Issues

    (Comment 1) One comment stated that these synthetic flavoring 
substances should not be revoked based on the Delaney Clause because 
``. . . the Delaney Clause does not mandate that FDA flatly prohibit 
the use of the substance under any circumstances.'' The comment goes on 
to say that ``[t]he determination that a substance triggers the Delaney 
Clause is not the same as a determination that the substance is 
necessarily unsafe in food and that ``. . . an outright ban of any of 
the flavorings identified by the petitioner would require FDA to 
explain--in a rulemaking procedure--why the substance not only triggers 
the Delaney Clause but also why there are no circumstances under which 
the substance could otherwise be

[[Page 50501]]

considered safe for food use under specified conditions of use.'' 
Several comments stated that FDA should interpret the Delaney Clause in 
a manner similar to the approach used by FDA in its Constituents Policy 
(i.e., FDA may determine that a food or color additive is ``safe'' if 
it contains a carcinogenic constituent but is not itself carcinogenic, 
see 47 FR 14464, April 2, 1982) for carcinogenic contaminants present 
in certain food additives.
    (Response 1) We disagree. The language of the Delaney Clause is 
straightforward. For most food additives, FDA has discretion to review 
a number of factors to determine whether a food additive is safe 
(section 409(c)(5) of the FD&C Act). However, for food additives that 
are shown ``to induce cancer in man or animal,'' the Delaney Clause 
limits FDA's discretion and requires that FDA conclude that the food 
additive is not safe. Furthermore, as described above, courts have 
rejected the interpretations of the Delaney Clause suggested in the 
comments and have concluded that the Delaney Clause completely bans 
additives found to induce cancer in humans or animals. Thus, as a 
matter of law, FDA cannot find these synthetic flavoring substances to 
be safe.
    (Comment 2) One comment said that the Delaney Clause applies only 
to food additives that induce cancer in test animals through a direct, 
genotoxic mechanism of carcinogenicity. The comment further stated that 
there are numerous examples of food ingredients that produce increased 
incidence of tumors in high dose rodent studies through a threshold 
secondary mechanism.
    (Response 2) We disagree. The Delaney Clause does not differentiate 
between non-genotoxic and genotoxic carcinogens. Nor does it permit FDA 
to find a food additive safe for human consumption if the food additive 
has induced cancer in animal. The Delaney Clause is a strict legal 
standard that precludes FDA from using its expertise to evaluate a 
substance under its intended condition of use and its risk to public 
health.
    (Comment 3) One comment stated that the petitioners call for a 
radical departure from long-established regulatory framework of FDA 
conducting its own comprehensive review of the scientific data that 
bear on the safety assessment. Further, the comment stated that the 
petitioners' approach is contrary to the statute and cannot be 
implemented without amendment of the law. The comment stated that if, 
contrary to the statute and long precedent, FDA believes it should 
delegate its authority to external organizations, it must consider such 
policy changes through notice-and-comment rulemaking. The comment also 
stated that while an FAP is the correct vehicle to appeal/amend a food 
additive regulation, it is not appropriate for FDA to consider, much 
less implement, ``radical new interpretations'' of the statute through 
a food additive petition.
    (Response 3) FDA disagrees with this comment. FDA's regulations 
permit petitioning the agency to revoke a food additive regulation. In 
response to such a petition, FDA conducts its own review of scientific 
data that bear on the petition. FDA then takes action based on its own 
evaluation of the data in accordance with the FD&C Act and its 
implementing regulations. The Delaney Clause is in the FD&C Act and 
this rulemaking is in accordance with the language of the law and case 
law interpreting it.

B. Scientific Issues

    (Comment 4) One comment included a lengthy discussion of relevant 
carcinogenicity and genotoxicity studies for each of the additives that 
are the subject of the petition and argued that none of the synthetic 
flavoring substances are direct carcinogens. Instead, the comment 
contended that tumors observed in the NTP studies were the result of 
secondary mechanisms and not direct, genotoxic effects.
    (Response 4) Our review included an evaluation of all relevant 
carcinogenicity studies for each of the additives. The toxicology 
memoranda for each of the six synthetic flavoring substances and 
section III include a full discussion of the relevant studies and 
address each scientific point outlined in the comment.
    (Comment 5) Several comments believed that FDA should not base its 
safety decision solely on classifications by NTP or IARC and that any 
decision should be based on an independent FDA assessment. Another 
comment stated that FDA must consider new studies since the NTP and 
IARC reviews were completed.
    (Response 5) FDA agrees with the comments and has conducted its own 
evaluation of available relevant data to reach its conclusions on each 
synthetic flavoring substance, and did not solely rely on NTP and IARC 
classifications as the basis for our decision.
    (Comment 6) One comment noted that IARC is not subject to U.S. law 
and relying on its conclusions is inappropriate and legally vulnerable 
for FDA. Another comment noted that IARC warns that its monographs are 
not the basis for governmental action, pointing out that the preamble 
to IARC monographs is clear that they are a starting place for 
government agencies, not a basis for regulation.
    (Response 6) We agree that relying solely on IARC conclusions would 
not be appropriate in making a decision on the petition, and, as such, 
FDA has conducted its own comprehensive carcinogenicity evaluation of 
the flavoring substances using all available relevant information.
    (Comment 7) One comment stated that the international health and 
safety community has moved away from rote reliance on IARC and NTP. The 
comment further said that the NTP and IARC classifications do not make 
those substances carcinogens under the Occupational Safety and Health 
Administration (OSHA) Hazard Communication Standard and that these 
reviews are not viewed as weight-of-evidence conclusions by 
international authorities; therefore, it would be incongruent for FDA 
to view them in this manner. The comment cited an action in 2012, where 
OSHA reversed three decades of automatically requiring employers to 
classify a substance as a carcinogen based on an NTP or IARC 
classification.
    (Response 7) FDA acknowledges that the NTP studies are designed for 
hazard identification and not for assessing the human carcinogenicity 
risk of chemicals under specific conditions of use; however, FDA must 
evaluate the results from the NTP studies and other available 
information within the context of the FD&C Act, including the Delaney 
Clause.
    (Comment 8) Some comments expressed concern that compliance and 
enforcement of a zero tolerance policy is not possible and that a zero 
tolerance policy is not feasible for naturally occurring substances.
    (Response 8) FDA has not addressed the request for FDA to establish 
zero tolerances for the food additives that are the subject of this 
petition because such a request is not the proper subject of a food 
additive petition, and because the petitioners have indicated that they 
are abandoning that claim.
    (Comment 9) Several comments expressed concern over the use of 
these substances in food packaging applications.
    (Response 9) Benzophenone is the only synthetic flavoring substance 
that is the subject of this petition that also is approved as a food 
additive for use in food packaging (Sec.  177.2600(c)(4)(iv) 
diphenylketone). As explained earlier, we are repealing the regulation 
for the

[[Page 50502]]

use of this substance as a plasticizer in food packaging based on 
results of the NTP studies.
    (Comment 10) One comment said that the use of ethyl acrylate should 
not be revoked, because the studies used to assess carcinogenicity were 
not appropriate and noted that NTP has removed it from its list of 
human carcinogens.
    (Response 10) FDA acknowledges that NTP has removed ethyl acrylate 
from its list of human carcinogens; however, the flavoring substance 
induced cancer in animals under the conditions of the 2-year NTP 
carcinogenicity studies. As such, we are required under the Delaney 
Clause to deem the additive to be unsafe as a matter of law. (See 
Section III.B, Ethyl Acrylate.)
    (Comment 11) One comment submitted on behalf of several industry 
interests supported removal of styrene from Sec.  172.515 based solely 
on abandonment and subsequently submitted a petition (FAP 6A4817 (81 FR 
38984)) providing data to support their claim.
    (Response 11) FDA is responding to this comment as part of our 
response to FAP 6A4817, which is published elsewhere in this edition of 
the Federal Register.
    (Comment 12) One comment stated that the petitioner should follow 
the National Environmental Policy Act and submit an environmental 
assessment but did not provide any supporting data.
    (Response 12) FDA disagrees. As discussed in section VII, we have 
determined that the action we are taking on the petition does not have 
a significant effect on the human environment and neither an 
environmental assessment nor an environmental impact statement is 
required.

V. Conclusion

    Upon review of the available information, we have determined that 
the information provided in the petition and other publicly available 
relevant data demonstrate that synthetic benzophenone, ethyl acrylate, 
methyl eugenol, myrcene, pulegone, and pyridine have been shown to 
cause cancer in animals. Despite FDA's scientific analysis and 
determination that these substances do not pose a risk to public health 
under the conditions of their intended use, under the Delaney Clause 
this finding of carcinogenicity renders the additives ``unsafe'' as a 
matter of law and FDA is compelled to amend the authorizations for 
these substances as food additives to no longer provide for the use of 
these synthetic flavoring substances. Additionally, because of evidence 
that benzophenone causes cancer in animals, FDA also is amending the 
food additive regulations to no longer provide for the use of 
benzophenone as a plasticizer in rubber articles intended for repeated 
use in contact with food. Therefore, we are amending parts 172 and 177 
as set forth in this document. Upon the publication, these food 
additive uses are no longer authorized.
    FDA realizes that the food industry needs sufficient time to 
identify suitable replacement ingredients for these synthetic flavoring 
substances and reformulate products and for these products to work 
their way through distribution. Therefore, FDA intends to not enforce 
applicable requirements of the final rule with regard to food products 
manufactured (domestically and internationally) prior to October 9, 
2020 that contain one or more of these six synthetic flavoring 
substances, to provide an opportunity for companies to reformulate 
products prior to enforcing the requirements of this final rule.

VI. Public Disclosure

    In accordance with Sec.  171.1(h) (21 CFR 171.1(h)), the petition 
and the documents that we considered and relied upon in reaching our 
decision to approve the petition will be made available for public 
disclosure (see FOR FURTHER INFORMATION CONTACT). As provided in Sec.  
171.1(h), we will delete from the documents any materials that are not 
available for public disclosure.

VII. Analysis of Environmental Impacts

    As stated in the January 4, 2016, Federal Register notice of 
petition for FAP 5A4810 (81 FR 42), the petitioners claimed a 
categorical exclusion from preparing an environmental assessment or 
environmental impact statement under 21 CFR 25.32(m). We have 
determined that the categorical exclusion under Sec.  25.32(m) for 
actions to prohibit or otherwise restrict or reduce the use of a 
substance in food, food packaging, or cosmetics is warranted. We have 
determined under Sec.  25.32(m) that this action is of a type that does 
not individually or cumulatively have a significant effect on the human 
environment. Therefore, neither an environmental assessment nor an 
environmental impact statement is required.

VIII. Paperwork Reduction Act of 1995

    This final rule contains no collection of information. Therefore, 
clearance by the Office of Management and Budget under the Paperwork 
Reduction Act of 1995 is not required.

IX. Objections

    If you will be adversely affected by one or more provisions of this 
regulation, you may file with the Dockets Management Staff (see 
ADDRESSES) either electronic or written objections. You must separately 
number each objection, and within each numbered objection you must 
specify with particularity the provision(s) to which you object, and 
the grounds for your objection. Within each numbered objection, you 
must specifically state whether you are requesting a hearing on the 
particular provision that you specify in that numbered objection. If 
you do not request a hearing for any particular objection, you waive 
the right to a hearing on that objection. If you request a hearing, 
your objection must include a detailed description and analysis of the 
specific factual information you intend to present in support of the 
objection in the event that a hearing is held. If you do not include 
such a description and analysis for any particular objection, you waive 
the right to a hearing on the objection.
    Any objections received in response to the regulation may be seen 
in the Dockets Management Staff between 9 a.m. and 4 p.m., Monday 
through Friday, and will be posted to the docket at https://www.regulations.gov.

X. References

    The following references marked with an asterisk (*) are on display 
at the Dockets Management Staff (see ADDRESSES), under Docket No. FDA-
2015-F-4317, and are available for viewing by interested persons 
between 9 a.m. and 4 p.m., Monday through Friday, they also are 
available electronically at https://www.regulations.gov. References 
without asterisks are not on display; they are available as published 
articles and books.

1. Bevan, R.J. (2017). ``Threshold and Non-Threshold Chemical 
Carcinogens: A survey of the Present Regulatory Landscape.'' 
Regulatory Toxicology and Pharmacology, 88, 291-302.
2. JECFA (2006). ``The Formulation of Advice on Compounds That are 
Both Genotoxic and Carcinogenic.'' WHO Food Additives Series No. 55, 
Annex 4.
3. Barlow, S. et al. (2006). ``Risk Assessment of Substances That 
are Both Genotoxic and Carcinogenic: Report of an International 
Conference organized by EFSA and WHO with Support of ILSI Europe.'' 
Food and Chemical Toxicology, 44, 1636-1650.
4. Flavor and Extract Manufacturers Association Transmittal Letter 
to Szabina Stice (FDA, CFSAN), April 27, 2018.*
5. FDA Memorandum from D. Folmer, CFSAN Chemistry Review Group,

[[Page 50503]]

Division of Petition Review, to J. Kidwell, Regulatory Group I, 
Division of Petition Review, June 24, 2016.*
6. FDA Memorandum from D. Folmer, CFSAN Chemistry Review Group, 
Division of Petition Review, to J. Kidwell, Regulatory Group I, 
Division of Petition Review, June 20, 2018.*
7. Food and Agriculture Organization of the United Nations and the 
World Health Organization. Principles and Methods for the Risk 
Assessment of Chemicals in Food. 2009. Available at http://www.inchem.org/documents/ehc/ehc/ehc240_index.htm. (Last accessed 
September 12, 2017.)
8. Flavor and Extract Manufacturers Association Letter to Judith 
Kidwell (FDA, CFSAN), April 11, 2016.*
9. FDA Memorandum from S. Thurmond, CFSAN Toxicology Team, Division 
of Petition Review, to J. Kidwell, Regulatory Group I, Division of 
Petition Review, June 21, 2018. *
10. Boobis, A.R. et al. (2006). ``IPCS Framework for Analyzing the 
Relevance of a Cancer Mode of Action for Humans.'' Critical Reviews 
in Toxicology, 36:10, 781-792.
11. FDA Memorandum from S. Thurmond, CFSAN Toxicology Team, Division 
of Petition Review, to J. Kidwell, Regulatory Group I, Division of 
Petition Review, June 21, 2018.*
12. National Toxicology Program. Report on Carcinogens Background 
Document for Ethyl Acrylate. December 2-3, 1998.
13. FDA Memorandum from J. Zang, CFSAN Toxicology Team, Division of 
Petition Review, to J. Kidwell, Regulatory Group I, Division of 
Petition Review, June 21, 2018.*
14. FDA Memorandum from A. Khan, CFSAN Toxicology Team, Division of 
Petition Review, to J. Kidwell, Regulatory Group I, Division of 
Petition Review, June 21, 2018.*
15. FDA Memorandum from N. Anyangwe, CFSAN Toxicology Team, Division 
of Petition Review, to J. Kidwell, Regulatory Group I, Division of 
Petition Review, June 21, 2018.*
16. Da Rocha, M.S., Dodmane, P.R., Arnold, L.L., et al. (2012). 
``Mode of Action of Pulegone on the Urinary Bladder of F344 Rats.'' 
Toxicological Sciences, kfs035.
17. FDA Memorandum from T. Tyler, CFSAN Toxicology Team, Division of 
Petition Review, to J. Kidwell, Regulatory Group I, Division of 
Petition Review, June 27, 2018.*

List of Subjects

21 CFR Part 172

    Food additives, Reporting and recordkeeping requirements.

21 CFR Part 177

    Food additives, Food packaging.

    Therefore, under the Federal Food, Drug, and Cosmetic Act and under 
authority delegated to the Commissioner of Food and Drugs, 21 CFR parts 
172 and 177 are amended as follows:

PART 172--FOOD ADDITIVES PERMITTED FOR DIRECT ADDITION TO FOOD FOR 
HUMAN CONSUMPTION

0
1. The authority citation for part 172 continues to read as follows:

    Authority:  21 U.S.C. 321, 341, 342, 348, 371, 379e.


Sec.  172.515   [Amended]

0
2. Amend Sec.  172.515(b) by removing the entries for ``benzophenone; 
diphenylketone,'' ``ethyl acrylate,'' ``eugenyl methyl ether; 4-
allylveratrole; methyl eugenol,'' ``myrcene; 7-methyl-3-methylene-1,6-
octadiene,'' ``pulegone; p-menth-4(8)-en-3-one,'' and ``pyridine.''

PART 177--INDIRECT FOOD ADDITIVES: POLYMERS

0
3. The authority citation for part 177 continues to read as follows:

    Authority:  21 U.S.C. 321, 342, 348, 379e.


Sec.  177.2600   [Amended]

0
4. In Sec.  177.2600(c)(4)(iv), remove the entry for ``diphenyl 
ketone.''

    Dated: October 2, 2018.
Leslie Kux,
Associate Commissioner for Policy.
[FR Doc. 2018-21807 Filed 10-5-18; 8:45 am]
 BILLING CODE 4164-01-P


