oversight

Food Additive, Acrylonitrile, Banned in Beverage Containers

Published by the Government Accountability Office on 1977-11-02.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                        DLCUNEIIT RESUME      1/-     9-'77
04045 -   B3074327]

Food Additive, Acrylonitrile, Banned in Beveraae Containers.
4RD-78-9; B-164031(2).  ovember 2, 1977. Released ovember 9,
1977. 17 pp.
Report to Rep. Toby offett; by Paul G. Dembling, Acting
Comptroller General.

Issue Area: Food (1700); Consumer and orker Protection: Safety
    of Food Additives (905).
Contact: Human Resources Div.
Budget Function: Health: PreventJ.oa and Control of Health
    Problems (553).
Organization Concerned: Food and Drug Administration.
Congressional Relevance: Rep. Toby olof'ett.
Authority: Food, Drug, and Cosmetic Act, as amended; Food
    Additives Amendment of 1958 (21 O.S.C. 348). 21 C.F.R.
    170.3(e). 21 C.F.R. 170.22.
          Acrylonitrile is a volatile, clear liquid, which cn be
formed  into more complex compounds known as polymers that are
used to make various plastic articles. Beverage bottles made of
acrylonitrile copolymers are lightweight, do not cause injury if
broker,, and are, therefore, desirable to consumers. A
characteristic of such bottles, however, is that afte,
pclymerization, a small amoant of "residual" acrylcnitrile that
has not combined with other monomers remains in the plastic and
may become part of the substance in te   bottle. Some
acrylonitrile copolymers also depolymerize to some ext 4 ent,
allowing additional acrylonitrile to migrate to the rood.
Findings/Conclusions: In January 1977, Food ad Drug
Administration (FDA) officials decided that the use of
acrylonitrile in making plastic bottles for carbonated beveraes
and beer should be banned and that all other acrylcnitrile uses
should be restricted to a maximum permissible migration level of
0.05 ppm. n arch 7,. 1977, Monsanto Coapany filed a moticn in
the U.S. court of appeals requesting a review of FDA's
suspension of its regulati-n authorizing the use of
acrylonitrile in bottles intended to hold soft drinks. In its
motion, Monsanto Company maintained that PA had not followed
the procedures required by section 4C9 of the Federal Food,
Druq, and Cosmetic Act in suspending the regulation. The primary
issue of this motion was whether FDA could remove a previously
lawful product from the market without notice and opportaniry
for affected parties to contest the action. The ccurt crdered
that FDA's suspension of the regulation be lifted until ay 18,
1977, and that FDA hold the required public hearing roaptly,.
The outcome was that acrylonitrile copolymers used to fabricate
beverage containers were judged to be food additives and were
not safe for use in food. FDA tereinated all regulations that
permit acrylonitrile in beverage containers. (SW)
        -4


..cJ'         'ai        REPORT OF THE
o            ;X}~BP      COMPTROLLER GENERAL
                ~N'~'    OF THE UNITED STATES

                         R£2T CT4b -
                                        e.            R E                     DE
                                             '9       H rE'cta
                                                     w.F-*ide   ied                  e   noral
                         Accauwt *.
                                 .- ' :i: e·          '-;   :. _ .r - '   ~    plctic appoval
                         b te A..d In Aid         ,.J.)iV        ;^^:>n

                         Food Additive, Acrylonitrile,
                         Banned in Beverage Containers

                         Acrylonitrile has been used for about 30
                         years to make various plastic food contain-
                         ers. Recent studies show that, when used in
                         beverage containers, this substance migrates
                         to the packaged beverage, contaminating it.
                        Bated oc these studies and on the results of a
                        public hearing concerning these issues, on
                        Sep.amt3r 23, 1977', the Food and Drug Ad-
                        ministration published a final order erminat-
                        ing all regulations that permit acr!;',litile in
                        beverage containers.




                        HRD-78-9                                          NOVEMBER 2, 1977
                 COMPTROLLFR GENERAL OF THE UNITED STATES
                            WASHINGTON. D.C. 20!48




B-164031(2)



The Honorable Toby MoffetL
House of Representatives

Dear Mr. Mor.cett:

     In your letter of March 15, 1977, you requested our
observations on the Food and Drug Administration's (FDA's)
March 11, 1977, decision to stay its food additive regula-
tions that had permitted acrylonitrile to be used in making
beverage containers. You indicated that FDA's only scien-
tific data on which to base such   decision was the prelimi-
nary report of a 2-year study which was only half completed.

     We interviewed FDA officials; reviewed legislation,
regulations, and practices relating to FDA's regulation of
food additives; examined FDA's records relating to tne
regulatory status of acrylonitrile; and reviewed testimony
presented at a public hearing on acrylonitrile used to make
beverage containers.

     As agreed with your office, we are sending copies of
this report to the Secretary of Health, Education, end Welfare
and the FDA Commissioner.  Unless you publicly announce its
contents earlier, we will not distribute this report further
until 7 days from the date of the report.

REGULATION OF FOOD ADDITIVES

     The Federal Food, Drug, and Cosmetic Act, as amended by
the Food Additives Amendment of 1958, requires FDA to estab-
lish regulations prescribing the conditions under which a food
additive may be safely used (21 U.S.C. 348). The act defines
"food additive" as any substance that becomes or may reason-
ably be expected to become a component of food, either
directl" or indirectly, or which may otherwise affect the
characteristics of the food. The effective date of a regula-
tion may be stayed if a hearing is sought by any person ad-
versely affected by such a regulation.
B-164031(2)



     Substances used in accordance with approvals given by the
Department of Agriculture and FDA before the enactment of the
1958 Food Additives Amendment are not subject to such regula-
tions.  Tnese approvals, generally known as prior-sanctioned
approvals, were based on safety requirements less formal and
restrictive than those provided by the 1958 amendment.

     FDA's food additive regulations (21 C.F.R. 170.3(e))
state that a material used to produce containers and packages
is considered a food additive if it may reasonably be expected
to become a component, or to directly or indirectly affect
the characteristics of food packed in the container.  Packag-
ing material that migrates to the food is an indirect food
additive.

     The Food Additives Amendment of 1958 and FDA regulations
require that a petitioner requesting the approval of an in-
direct food additive demonstrate the safety of the additive
with appropriate data showing the amount of the substance
that is expected to migrate to food and the results of
toxicological studies. The amendment also requires FDA to
consider the probable consumption of the additive when deter-
mining the safety of the proposed use of a food additive.

      According to an FDA Division of Toxicology official, the
number and type of toxicological tests the agency requires
for a safety evaluation of an indirect additive depend on the
amount of the chemical that is expected to migrate to the
food.   He said that FDA requires only acute toxicity data to
insure that a substance is not a highly toxic chemical when
anticipated migration is "virtually nil" (generally less than
0.05 parts per millicn (ppm)), provided that the substance
is not a heavy metal or a pesticide and there is no reason
to suspect carcinogenicity or teratogenicity.   More extensive
toxicological test data is required when migration of a sub-
stance is anticipated to be "negligible" (more than 0.05 ppm
but less than about 1.0 ppm).

WHAT IS ACRYLONITRILE?

     Acrylonitrile is a volatile, clear liquid. Because of
its relatively low molecular weight, it is classified as a
monomer, which, when joined with certain other monomers
through a process known as polymerization, can form more
complex compounds known as polymers that are used to make
various plastic articles.



                              2
 B-164031(2)



     Acrylonitrile copolymers, a class of polymers, have
been used for about 30 years in various food containers,
such as margarine tubs, vegetable oil bottles, and pipes
                                                          for
handling food products.  Copolymers with high concentrations
of acrylonitrile have excellent barrier (pressure-holdina)
characteristics. Since such copolymers have low permeability
to gases, they are suitable for carbonated beverage bottles
that must retain gas for extended periods.

      Beverage bottles made of acrylonitrile copolymers are
lightweight, do not cause injury if broken, and are therefore
desirable to consumers.   A characteristic of such bottles,
however, is that after polymerization, a small amount of
"residual" acrylonitrile that has not combined with other
monomers remains in the plastic and may become prt of the
substance in the bottle. Some acrylonitrile copolymers also
depolymerize to some extent, allowing additional acrylonitrile
to migrat'e to the food.

BASIS FOR INITIAL SAFETY DETERMINATION

     FDA's initial prior-sanctioned approvals of acryloni-
trile's use in food-contact articles were based on analytical
data indicating that there was no significant migration of
acrylonitrile to food under the conditions of its intended
use.  These approvals included uses for oleomargarine wraps
and equipment and containers used in fod processing.
     Pursuant to the 1958 Food Additives Akendment the ap-
proval process for food additives requires the submission
                                                           of
a petition proposing the issuance of a regulation that would
prescribe the conditions under which an additive may be safely
used. Since 1958 FDA has issued 20 food additive regulations
approving various additional acrylonitrile uses in food-
contact articles.  These approvals were based on data sub-
mitted with petitions which showed no significant. migration
of acrylonitrile to the packaged food.  The first regulation
allowing the use of acrylonitrile in plastic beverage bottles
was issued to Vistron Corporation on June 11, 1970.

SAFETY LATER QUESTIONED

     Extraction studies submitted by the E. I. duPont de
Nemours Company with its June 14, 1973, petition to FDA
requesting approval for use of its acrylonitrile bottle to
package soft drinks showed that acrylonitrile migrated from
bottles to food when stored at elevated temperatures for


                              3
B-164031(2)



extended periods. Extraction stu-.es can measure, throughl
analytical methods, the aunt of a substance that has
migrated to a food.  Based on the new evidence and the fact
that improved procedures for detecting and measuring acrylo-
nitrile had been developed, FDA began a complete review of
all safety data available on this substance.

     FDA found that available toxicity data on acrylonitrile
consisted primarily of two series of long-term feeding
studies. One series, entitled "The Pharmacology and Toxic-
ity of Acrylonitrile and Acrylon" (by P. E. Tullar, George
Washington University, Nov. 1, 194 ), does not, according
to FDA, meet current standards for the toxicological testing
of substances such as food additives. An FDA Division of
Toxicology official said that, at best, a tentative "no-
adverse-effect" level for acrylonitrile could be established
from this study at 38 ppm.

     The other series of studies available at that time was
made by Svirbely and Floyd of the U.S. Public Health ScLvice
at the Taft Sanitary Engineering Center in Cincinnati in the
late 1950s. The series consisted of 2-year fding studies
in rats and dogs and a 3-generation rat r:production study.
A final report on this series was nZver pepared, and accord-
ing t FDA the available aLa provided only limited informa-
tion that was inadequate to establish a no-adverse-effect
level for acrylonitrile.

     Based on the results of its review, FDA concluded that
the safety of acryloni-ile was questionable, that the evel
of acceptable migration should be limited, and that addi-
tional information and studies were needed on (1) analytical
methods concerning migration and (2) toxicity.

    FDA regulations (21 C.F.R. 180.?) state that substances
with a history of use in food fo£ human consumption or in
food-contact surfaces may at any time have their safety or
functionality bought into question by new information that
in itself is inconclusive. An interim food additive regula-
tion for the use of any such substance may be issued

     "* * * when new information raises a substantial
     question about the safety or functionality of the
     substance but there is a reasonable certainty that
     the substance is not harmful and that no harm to
     the public health will result from the continued
     use of the substance for a limited period of time


                               4
B-164031(2)



     while the question raised is being resolved by
     further study."

     On November 4, 1974, FDA published in the Federal
Register a proposed interim ooC additive regulation for
acrylonitrile intended for use in food-contact articles.
This regulation proposed, for an interim period; a limit
of 0.3 ppm as the maximum amount of acrylonitrile allowed
to iarate to food and required that toxicological studies
be conducted to determine the safety of lcw levels of acry-
Ionitrile ingestion.

     The 0.3-ppm migration limitation was established by
applying a 100 to 1 safety factor to a tentative no-adverse-
effect evel of 38 ppm, which was based on the 1947 P. E.
Tullar study.  FDA regulations (21 C.F.R. 170.22) provide
that

     "* * * a food additive for use by man will not
     be granted a tolerance that will exceed 1/100th
     of the maximum amount demonstrated to be without
     harm to experimental animals."

     On June 19, 1975, the Manufacturing Chemists Association
submitted, for FDA review, protocols for a series of studies
designed to provide the toxicity data called for by the pro-
posed interim regulation.. On July 18, 1975, FDA informed the
association that the protocols, with some technical modifica-
tions, were acceptable. The association contracted with the
Do Chemical Company to make the studies. The studies were
sponsored by the following manufacturers or major users of
acrylonitrile: American Cyanamid Company, Borg-Warner
Chemicals, Dow Chemical Company, E. I. duPont de Nemours and
Company, Monsanto Company, Standard Oil Company (Ohio),
Tennessee Eastman Company, and Uniroyal, Incorporated.

     After publishing its proposed interim regulation on
acrylonitrile, FDA issued regulations permitting duPont,
Borg-Warner, Monsanto, and Dow to use acrylonitrile to make
plastic bottles.

ISSUANCE OF FINAL INTERIM REGULATION

     On June 14, 1976, FDA issued in final form the interim
food additive regulation on the use of acrylonitrile in food-
contact articles. The regulation established the maximum
permissible level of acrylonitrile migration at 0.3 ppm


                              5
B-164031(2)



pending the results of the ongoing toxicity studies.
Thirty days were allowed for comments on and objections
to the regulation.

     Four manufacturers and one industry association sub-
mitted comments to FDA on the interim regulation. These
comments included questions about

     -- the appropriateness of a 120 degree FaLrenheit tem-
        perature requirement for migration studies,

     -- the technical inconsistency of the volume-to-surfacc
        ratio requirements for migration measurement,

     -- the Fossitility of extending the 30-day comment period
        to allow time to obtain FDA's migration test method-
        clogy and evaluate the regulation:   iTm-ct on the
        product being manufactured, and

     -- the lack of prior knowledge about the existence of
        the proposed regulation.

     Only the Natural Resources Defense Council filed an
objection questicning the safety of acrylonitrile in food-
contact applications.  It cited several reasons for ques-
tioning the safety of acrylonitrile.

     -- The 1947 P. E. Tullar study, which concluded that
        acrylonitrile causes cancer.

     -- The fact that acrylonitrile, also known as vinyl
        cyanide, is chemically related to vinyl chloride,
        a known carcinogen.

     -- An epidemiology study of Japanese acrylonitrile
        workers which found a correlation between exposure
        to the substance and impairment of liver function.

Based on the lack of sufficient data supporting the safety
of acrylonitrile, the council on July 13, 1976, requested
that a stay be granted of any food additive regulation
under which acrylonitrile might reasonably be expected
to migrate co a food product. The council requested that
the stay remain in effect pending a hearing and a decision
on its objection.




                              6
[-164031(2)



STUDY RESULTS

     Studies done in response to the need for additional
information on acrylonitrile include toxicity studies iden-
tified in the interim food additive regulation and migration
studies designed to measure very small amounts of the sub-
stance in fcod.

Toxicity studies

     On Decembe:: 9, 1975, the Manufacturing Chemists Associa-
tion submitted to FDA results of two toxicity studies by Dow
Chemical.   In the first study, rats were fed acrylonltrile
at levels of 0, 35, 85, 210, and 500 ppm in their drinking
water for 90 days.

     The study report stated that:

     "Results of this study indicate that male and
     female rats may ingest water containing at
     least 85 ppm of acrylonitrile for a period f
     90 days without a deleterious effect on any of
     the arameters evaluated."

The repozt noted that decreased body weight gains were ob-
served in male rats in the top dose level and in female rats
in the two top dose levels.  It further reported that small
bu. significant increases in liver to body weight ratios were
observed ii male and female rats on water containing 500 and
210 ppm acrylonitrile.

     Dur'.na the second study, purebred beagle dogs were fed
acrylonitrile at levels of 0, 100, 200, and 300 ppm (eight
dogs per level) ii their drinking water for 6 months.

     The study report concluded that:

     "Ingestion of AN [acrylonitrile] in the drinking
     water by male and female dogs at 200 and 300 ppm
     for a period of 6 months was associated with
     toxicological manifestations primarily and
     secondarily related to AN.

     "Ingestion of AN in the drinking water at 100 ppm
     for 6 months resulted in no toxicological mani-
     festations in male or female dogs."



                              7
 B-164031(2)



     The FDA toxicologist who evaluated the study, however,
stated in a February 1977 memorandum to the Director of
FDA's Division of Food and Color Additives that:

     "With the normal criteria of toxicity data
     evaluations, the no-effect level appears to be
     lower than 100 ppm (8-10 mg/kgm [milligrams per
     kilogram]).   The effect at this level was a
     slight lag in    owth, and microscopic changes
     in the esophagus, with increased thickening of
     the epithelium of the tongue."
He said that at the two higher dose levels, effects such
as growth lags, increased mortality, and reduced liver
function were noted.

     On Aril 19, 1976 the Manufacturing Chemists Associa-
tion submitted to FDA the results of a study conducted
Younger Laboratories .. te joint toxic action between by
acrylonitrile and potassium cyanide. This study
male and female rats were orally given mixtures ofin acrylo-
                                                      which
nitrile and potassium cyanide showed no evidence of syner-
gistic effect on acute toxicity.
     On October 13, 1976, the association submitted results
from nutagenici`y studies made in three different labora-
tories. Only one laboratory reported evidence of mutagenic
effects from one of several tests that it conducted.

      In November 1976 the association submitted to FDA a
report entitled "Teratologic Evaluation of Acrylonitrile
Monomer Given to Rats by Gavage." The study, conducted by
Dow Chemical, was designed to evaluate the effects of acry-
ionitrile on embryonal and fetal development. In the study
pregnant rats were given, oii days 6 through 15 of gestation,
either 0, .0, 25, or 65 mg of acrylonitrile per kilogram
body weight by gavage (oral administration of a substance of
directly to the stomach through a feeding tube). At the
65-mg feeding level, effects such as maternal toxicity, in-
creased numbers of fetal malformations, decreased  fetal
wei,ht, and minor skeletal variants occurred. At the 25-mg
feeding level, maternal toxicity and malformations also
occurred, but with less frequency. The report stated that
at the 10-mg (approximately 100-ppm) feeding level, there
                                                           was
no evidence of toxicity to either the mother or her develop-
ing embryo or fetus.



                              8
B-164031 2)


     On January 14, 1977, the sscciation submitted to FDA
a 13-month interim report on a 2-year rat feeding study
being made by Dow Chemical. The study, scheduled for com-
pletion about November 1977, involves feeding rats acryloni-
trile in their drinking water at concentrations of 0, 35,
100, and 300 ppm. With respect to eedings at the 100- and
300-ppm levels, the interim report stated that:
     "It is concluded hat administration of AN
     under the condition of this study has signifi-
     cantly lowered body weight, produced pathologic
     changes in the gastric epithelium, increased the
     incidence of masses of the ear duct and produced
     proliferative lesions in the centLal nervous
     system of rats. The significance of the higher
     incidence of subcutaneous masses in the mammary
     region of the rats is less clear, and its resolu-
     tion will have to await further progress in the
     study."
     On April 11, 1977, the Manufacturing Chemists Associa-
tion provided FDA with updated information on the interim
findings of Dow's 2-year rat feeding study. According to
the association, the study investigator had found tumors of
the central nervous system in two rats on the 35-ppm feeding
level of the same type found earlier in rats on the 100.- and
300-ppm feeding levels.
     The association also notified FDA that another 2-year
study in which rats were being exposed to 0, 20, or 80 ppm
acrvlonitrile by inhalation showed the following interim
results.
     -- Three rats in the highest dose group (80 ppm) devel-
        oped tumors of the central nervous system comparable
        to those reported in the 2-year ingestion study.
     -- An increased incidence of ear canal tumors and mammary
        region masses was detected at the 80-ppm level.
     -- An apparent increase in the subcutaneous masses of the
        mammary region was detected in the rats in the 20-ppm
        group.
     -- No ear canal or central nervous system tumorF were
        found in the 20-ppm group.


                              9
B-164031(2)



'ccording to an FDA Division of Toxicology official, although
an inhalation study is not directly relevant to ingestion
 -oxicity, the findings of tumors similar to those in the
feeding study and tumors other than of the lung or respira-
tory tract indicate the systemic nature of the apparent car-
cinogenic effect of acrylonitrile after absorption.

     By letter dated May 23, 1977, the E. I. duPont Company
providedd3DA with preliminary results of an ongoing epidemi-
ology st'ldy of workers exposed to acrylonitrile in a textile
fibers lant. The preliminary findings revealed a possible
relationship between exposure of in-plant workers to acrylo-
nitrile and cancer. A higher-than-expected incidence of
various types of cancer was found among the 470 males sur-
veyed.  These men began working in the polymerization area
of the plant between 1950 and 1955 and are either still
actively employed by or retired from the company. Medical
data analyze3 through December-31, 1975, showed that 16 cancer
cases among his group were found, compared with an expected
company rate of 5.8 and a national rate f 6.9.    The cases
include six lung cancers (1.5 expected) and three colon
cancers (0.5 expected).   All cases of cancer occurred in the
group naving initial exposure during startup of the plant in
1950 through .1952.

Migration/extraction stidies

     In 1975 FDA began o develop methodologies for measuring
acrylonitrile migration. The team leader responsible for
FDA's efforts stated in written testimony for a public hear-
ing (see pp. 16 and 17) that FDA had developed and confirmea
an extraction test method capable of measuring acrylonitrile
at levels of 0.04 pm.    e said that FDA's method is based on
one submitted by duPont which uses gas liquid chromatography.

     The team leader s.d that two migration studies on
acrylonitrile bottis have been made with FDA's method. On
August 6, 1976, FDA began a migration study using two test
groups of acrylonitrile bottles--four 32-ounce bottles from
Monsanto and four 16-ounce bottles fron Borg-Warne:--which,
according to FDA's analysis, contained 9 to 10 ppm and 25 ppm
residual acrylonitrile, respectively.

     Two of the bottles from each company were filled with
the food simulant 3 percent acetic aid and two with food
simulant 8 percent ethanol, and each was heated to 120 degrees
Fahrenheit. Food simulants are often used in place of food


                               10
B-164031(2)



products during migration te.ts ir order to simplify detection
and measurement by eliminating many of the unknown substances
that might be present in food.  Acetic acid is generally ac-
cepted as a substitute for carbonated beverages and ethanol
as a substitute for alcoholic beverages.

     Test samples were withdrawn from the bottles over a period
of about 5 months. The following table presents the average
concentrations of acrylonitrile found on January 12, 1977, in
samples drawn from the bottles during the test period.

                                  Acrylonitrile found
          Solvent         Monsanto bottles Borg-Warner bottles

3 percent acetic acid          0.062 ppm          0.112 ppm
8 percent ethanol               .078 ppm           .126 ppm

An FDA project manager involved in regulating acrylonitrile
told us that the Borg-Warner bottles used in this test were
early prototype bottles, not those made from copolymers that
complied with the final Borg-Warner regulation.

     According to the team leader, the second FDA study used
empty and filled Monsanto bottles obtained from the Coca Cola
Bottling Company of New York on August 9, 1976. In one por-
tion of the study, two empty bottles, whose walls were found
to contain approximately 7 to 8 ppm of residual acrylonitrile,
were filled on February 14, 1977, with 3 percent acetic acid
and stored at 120 degrees Fahrenheit. After various periods
of time, samples of the acid were withdrawn and tested. The
average amount of acrylonitrile found in the acid samples is
indicated below.

                                           Average amount of
                                             acrylonitrile
                                              (estimated)
 Date                   Elapsed time           (note a)
2/24/77                   10   days            0.005   ppm
3/ 1/77                   14   days             .005   ppm
3/15/77                   28   days             .005   ppm
4/12/77                   56   days             .018   ppm
5/10/77                   84   days             .022   ppm

a/These values are estimated because the gas liquid chromoto-
  graphy method is not capable of accurately quantifying con-
  centrations of acrylonitrile less than 0.04 ppm.


                                  11
B-164031(2)



     In another phase of the study, four of the bottles
filled with Coca Cola were maintained at room temperature
from the time of collection until January 1977.  FDA extrac-
tion studies showed corcentrations of acrylonitrile in the
Coca Cola ranging from 0.013 to 0.029 ppm.

     The FDA laboratory confirmed these findings with mass
spectromeLric and polarographic techniques.  Both techniques
confirmed the levels of acrylonitrile found in the Coca Cola
using the gas liquid chromotography method.  The laboratory
team leader believed that FDA's studies unequivocally estab-
lish that acrylonitrile beverage containers can reasonably
be expected to become, and in fact do become, a component of
the beverage packaged in them.

DECISION TO BAN USE OF ACRYLONITRILE

     An FDA project manager involved in regulating acryloni-
trile told us that, on the basis of the positive findings of
the teratogenic study (see p. 8), FDA officials decided in
November 1976 to revise the interim regulation for acryloni-
trile regarding its use in making plastic bottles for car-
bonated beverages and beer by reducing its permissible migra-
tion level from 0.3 ppm to 0.05 ppm. This evel approximated
the lowest attainable level of measurable detection. A draft
revision to the interim regulation dated November 1976 was
prepared.

      After receiving the interim report on Dow's 2-year
chronic-feeding study in January 1977, FDA officials decided
that the use of acrylonitrile in making plastic bottles for
carbonated beverages and beer should be banned and that all
other crylonitrile uses should be restricted to a maximum
permissible migration level of 0.05 ppm. According to FDA's
Office of General Counsel, this decision was based not only
on the findings of the teratology and chronic-feeding studies
but also in combination with (1) data developed by FDA extrac-
tion studies indicating that migration of acrylonitrile occurs
in beverage containers containing the substance, (2) indica-
tions that the containers being manufactured might not meet
the proposed 0.05-ppm limit on acrylonitrile migration, and
 (3) the possibility of much greater consumer exposure to
acrylonitrile from caroonated beverage containers than from
other food-contact uses.

      On February 15, 1977, FDA published its environmental
 impact statement in the Federal Register which included an

                              12
B-164031(2)



announcement of the agency's intentions to (1) suspend the
use of acrylonitrile in making plastic bottles for carbonated
beverages and beer and (2) reduce the allowable amount of
acrylonitrile migration from 0.3 ppm to 0.05 ppm for all
other food-contact articles.

     At that time the Coca Cola Company was test marketing
its product packaged in acrylonitrile bottles manufactured by
Monsanto Company.  On February 14 ad 16 and March 3, 1977,
Monsanto met with FDA officials and presented data in an at-
tempt to convince FDA to withdraw its intended ban on the use
of such bottles. Monsanto's primary points were that (1) the
migration test procedures which require storing the bottles
at 120 degrees Fahrenheit until no further migration occurs
were unreasonable since sealed soft drink bottles deform at
that temperature and the contents are ruined, (2) in lieu of
a ban, a standard should be set allowing no acry onitrile
migration to soft drinks when the bottles are stored for
30 days at 120 degrees Fahrenheit (equivalent to 2 years of
storage at room temps-ature) and tested with a proposed test
method sensitive to 0.01 ppm, and (3) the safety data sub-
mitted to date plus the metabolic studies in progress would
enable FDA to limit acrylonitrile consumption by setting a
migration staneard for acrylonitrile bottles.

      After evaluating Monsanto's data, FDA officials still
believed that the safety of acrylonitrile was sufficiently
questionable to warrant banning its use in plastic bottles
for carbonated beverages and beer.   FDA advised Monsanto
that (1) migration testing at 120 degrees represents only
a slight exaggeration of possible storage and transportation
temperatures and would cover all eventualities in the storage
and shipments of the final product, (2) it could not guarantee
that 120 degrees for 30 days represents the point at which all
potential migration of acrylonitrile has occurred, nor could
it. determine, with available data, that consuming large   -
volumes of a beverage containing up to 0.01 ppm acrylonitrile
was safe, and (3) it could not set a no-effect level for
acrylonitrile from the available data and the metabolic
studies need to be completed before conclusions on sch a
level could be drawn.

     On March 11, 1977, FDA published two notices in the
Federal Register. One notice proposed several amendnents
to the interim food additive regulation for acrylonitrile,
including (1) the reduction of the maximum permissible
StDunt of the substance allowed to migrate from packaging


                             13
B-164031(2)



material other than beverage containers from 0.3 ppm to
0.05 ppm, (2) the deletion of duplicative requirements for
submission of migration data on repeated-use articles, (3) a
requirement for higher test temperatures, if needed, to simu-
late actual use conditions, and (4) a requirement for addi-
tional teratogenic tests in a species other than rats.

     The other notice suspended the food additive regulations
that permitted acrylonitrile to be used to make beverage con-
tainers. The notice stated that the regulations were being
suspended pending full review of the objections and request
for hearing filed by the National Resources Defense Council
on July 13, 1976.   (See p. 6.) The notice also said the
Commissioner anticipated that, when the Dow Chemical chronic-
feeding study was completed or possibly sooner, it would be
appropriate to convene an evidentiary hearing to resolve the
factual issues raised in the Council's objections.

FDA'S ACTION STAYED

      )n March 7, 1977, Monsanto Company filed a motion in
the   .S. court of appeals requesting a review of FDA's sus-
pens on of its regulation authorizing the use of acryloni-
trile in bottles intended to hold soft drinks. On March 1,
the court sustained Monsanto's motion pending its review of
the merits and ordered that oral arguments be presented on
March 16. The court granted the Natural Resources Defense
Council permission to present oral arguments.

     In its motion Monsanto maintained that DA had not
followed the procedures required by section 409 of the Fed-
eral Food, Drug, and Cosmetic Act in suspending the regula-
tion. Monsanto said that the regulation issued to it permit-
ting the use of acry.onitrile in beverage bottles could not
be suspended without prior notice and   hearing. According
to Monsanto the primary issue was whether FDA could remove
a previously lawful product from the market without notice
and opportunity for affected parties to contest the reason-
ableness, necessity, or legality o the action.

     In an affidavit submitted with its motion to review FDA's
suspension of the regulation, Monsanto's group vice-president
stated that the most sensitive validated test methods avail-
able showed that no migration occurred in Monsanto bottles
under the intended conditions of use. He said that if
Monsanto's motion were not granted, the company would suffer
irreparable harm in at leas'. the following ways.


                             14
B-164031(2)



     -- It would be unable to sell existing inventories of
        about 15 nmillion bottles valued at more than
        $2 million.

     -- It would be unable to resume operations at some plants
        and facilities, and hundreds of jobs would be perma-
        nently lost.

     -- A arge part of its capital investment in the facili-
        ties producing the bottles, exceeding $60 million,
        would be lost.

     -- It would lose potential sales of more than $30 million
        in 1977 and projected sales of more than $100 million
        annually by 180.

     FDA argued that it had followed the administrative pro-
cedures outlined in section 409 of the act, which was its
basis for the suspension.   FDA said that its authority to
suspend the regulation was derived from that regulation's
dependence on the results of ongoing efforts to establish
safe conditions for use of acrylonitrile, as outlined in itF
interim regulation.   FDA also argued that (1) Monsanto was
given the opportunity, through FDA's advance notice of its
intentions in the environmental impact statement, to present
data and arguments to the agency before the suspension was
issued and (2) more importantly, Monsanto would have further
opportunity at an evidentiary hearing to challenge FDA's judg-
ment about the lack of safety of acrylonitrile in beverage
bottles.   FDA said that Monsanto's argument that a hearing
should precede rather than follow the suspension is more
appropriately directed at the Congress because section 409(e)
provides that food additive regulations are effective at the
time of publication in the Federal Register but may thereafter
be stayed if objections are filed and a hearing is sought.
FDA noted that the Congress has not provided for delaying
regulations or for an automatic suspension upon receipt of
objections.

     FDA said that, although it did not dispute Monsanto's
assertion that the administrative suspension would result in
loss of sales for Monsanto and "economic dislocation" of
those workers laid off, it did not believe these injuries
amounted to "irreparable injury." FDA said that Monsanto's
claim was mitigated by the fact that it had been on notice
since February 975, when the regulation for its product was
issued, and in act since November 1974, when the proposed


                             15
B-164031(2)



interim regulation for acrylonitrile notified all interested
parties that the safety of the substance was an open question
and that its use could be affected at any time by FDA action.
According to FDA, the very real questions about the safety
of the bottle could not allow Monsanto to assert its economic
losses as a basis for continuing to expose the public to a
potential health hazard.

     The court determined that for FDA to order a suspension
of regulations based on objections filed 8 months earlier by
the National Resources Defense Council constituted "arbitrary
and capricious administrative action" and that the requirement
of section 409(f)(1) of the Federal Food, Drug, and Cosmetic
Act that a hearing be held "as promptly as possible," com-
pelled FDA to process objections more expeditiously than it
had in this case.

     On March 18, 1977, the court ordered that FDA's suspen-
sion of the regulation issued for Monsanto's product be
lifted until May 18 and that FDA hold the required public
hearing promptly and issue final findings and conclusions by
that date. On April 11, 1977, all parties involved in the
case requested a 120-day extension, which the court granted,
thus setting a new date of September 19 for issuance of final
findings and conclusions.

     The hearing process required by the court has been com-
pleted. Five parties participated:   FDA, Monsanto Company,
Borg-Warner Corporation, Vistron Corporation, and the Natural
Resources Defense Council. The major issues addressed during
the hearing, which was conducted by an administrative law
judge, were (1) whether or not acrylonitrile is a food addi-
tive as defined by law and (2) whether or not the safety of
acrylonitrile used to fabricate beverage containers has been
shown.

     Cross-examination of witnesses took place from
June 20 through 27, 1977, and briefs were submitted to
the administrative law judge for a decision. On August 4
the judge rendered an initial decision that acrylonitrile
does migrate to food, thus making it subject to regulation
under the act, and that it has not been shown to be safe for
use in food.

     The FDA Commissioner reviewed the initial decision along
with exceptions filed by the manufacturers. The Commis-
sioner's final decision, effective September 19, 1977,


                             16
B-164031(2)



affirmed the initial decision of the administrative law
judge by finding "* * * that acrylonitrile copolymers used
to fabricate beverage containers are food additives and
that they have not been shown to be safe."  The Commis-
sioner s decision concluded the hearing process required
by the court of appeals, and an official of FDA's Office
of General Counsel said that a copy of the decision was
filed with the court on September 20.

     On September 23 FDA published in the Federal Register,
to be effective 90 days from that date, a final order ter-
minating all regulations which permit the use of acryloni-
trile in making beverage containers.

      An official of FDA's Office of General Counsel said
 that interested parties can now file objections to the
·decision with the court, which could re iew the basis for
 FDA's decision. He added that 6 months or more may be
 required to complete a court of appeals review of the
 decision. At that time the parties wor ld be required to
 accept the court's decision or appeal the case to the
 U.S. Supreme Court.



     As requested by your office, we have not obtained the
Department of Health, Education, and Welfare's written com-
ments on the matters in this report. However, we have dis-
cussed these matters with FDA officials and have considered
their comments.

                                 Sincerely yours,




                       ACTING    Comptroller General
                                 of the United States




(10878)                         17