-------
is 0.3 mg/kg/day  for  1 year and  for monkeys the NOEL is  less  than
1.0 mg/kg/day for 1.5 years.

5.5. Mechanisms of Neurotoxic Action
     The current  status of our knowledge concerning the
biochemical mechanism of aery1amide's neurotoxicity has  recently
been reviewed by  Miller and Spencer (1985), O'Donoghue (1985),
and others.
     Two potential mechanisms for acrylamide's peripheral
neurotoxicity are under investigation.  One involves the binding
of aery1amide to  neurofilaments  and the other describes  the
.inhibition of neural,  enzymes needed for energy production.
Recent work lends^support to each hypothesized mechanism.  In a
study by Sickles  (1987), acrylamide demonstrated significant
reductions in NADH-tetrazolium reductase (TR) activity in ligated
axons, Purkinje neurons, and dorsal root ganglion neurons.  In
contrast, no significant reductions in activity were noted in
non-neural tissue enzymes.  Methylene-bis-acrylamide which is
non-neurotoxic had no effect on  enzyme activity levels except in
Purkinje neurons.  Sickles (1987) concluded that the data
support, in general,  the hypothesis that specific inhibition  of
neural NADH-TR activity by acrylamide is the primary site of
action in producing neuropathy.  On the other hand, binding to
cytoskeletal elements of the nervous system has been reported by
Lapadula et al.  (1986).  A series of in vivo and in vitro
experiments were  done using mice and brain microtubule and spinal

                               5-20

-------
cord cytoskeletal preparations from rats.  Binding to
cytoskeletal elements occurred in the in vitro studies.  In the
in vivo studies, binding to cytoskeletal proteins was similar to
that in the in vitro studies and occurred in all neural tissues.
According to the authors, these data demonstrate that aery1amide
may produce neuropathy by binding directly to cytoskeletal
elements.
     Aery1amide has been shown to interfere with retrograde
axonal transport.  How these biochemical effects singly or in
combination relate to axonopathy needs further investigation
before the mechanism of aery1amide's neurotoxicity can be
elucidated.
                              5-21

-------
            6.   DEVELOPMENTAL AND REPRODUCTIVE  EFFECTS
6.1. Summary and Conclusions
     Aery1amide produces both fetal and postnatal developmental
effects in mouse and rat offspring following administration to
pregnant dams,  it produces neurotoxic effects (tibial and optic
nerve degeneration) in the neonates at levels that are not toxic
to the dam.  In addition/ aery1amide changes intestinal enzyme
levels in the fetus at dose levels where no maternal toxicity is
apparent, but the toxicological significance of these changes is
not clear.  There is concern for the conceptus following maternal
exposure during gestation since these data show that aery1amide
can have a direct effect on the conceptus.  A thorough teratology
study would provide needed data on the structural teratogenic
                  •\
effects of aery 1 amide on the conceptus.  Due to the nature ,pf
aery1amide toxicity, a state-of-the-art behavioral teratology-
study is also needed to provide information on effects in the
neonate.
     Aery1amide also has an adverse effect on reproduction.  A
dominant lethal effect and decreased copulatory performance have
been observed in rats.  In one assay, rat testosterone levels
were depressed, but the functional significance of this finding
is uncertain.  In the mouse, decreased fertility was observed
following exposure of male mice to aery 1 amide and an increased
resorption rate; resulted from the exposure of male or female
mice.  Degeneration of testicular epithelial tissue and a
dominant lethal effect have also been observed in other mouse

                               6-1

-------
studies.  The available data indicate that aery1amide can act
directly on the reproductive system, rather than indirectly
through stress or other systemic effects.
     Developmental toxicity data indicate a lowest observed
effect level (LOEL) of 20 mg/kg/day.  Since this was the lowest
dose level tested in the study, a no observed effect level (NOEL)
was not established (Table 6-1).  The LOEL and NOEL for
reproductive effects (Table 6-2), are 8.8 and 4.2 mg/kg/day,
respectively.

6.2. Developmental and Female Reproductive Effects
     Several investigators have described the effects of
acrylamide on the*developing conceptus.  Pregnant rats (Porton
                  i
strain) were given acrylamide either as powder in the diet ,pr by
iv injection (Edwards 1976).  In Group I, 8 rats were given 200
ppm acrylamide in the diet (about 20 mg/kg/day) from the day of
mating until parturition.  In Group II, 6 rats were given 400 ppm
(about 40 mg/kg/day) acrylamide in the diet from the day of
mating until gestation day 20, and then were subjected to
Caesarean section,  in Group III, 4 rats were given a single iv
dose of acrylamide (100 mg/kg) on day 9 of gestation.  The only
difference between the control and Group II was a slight decrease
                               6-2

-------
                                                        Table 6-1

                                                  Sunmary of Acrylamide

                                      Developmental and Female Reproductive Toxicity
STUDY
TYPE (Ref)
Species, Route
  dose levels
   Maternal
LOEL       NOEL
 Developmental
LOEL      . NOEL
                  Observed Effects
           Maternal.   ,       Developmental
Developmental
Toxicity

(Edwards, 1976)
Porton Rat
Oral-diet
0 ppn
200 ppm or 20 mg/kg/d
gd 0-22
400 ppm or 40 mg/kg/d
gd 0-20
100 mg/kg (i.v.) gd 9
 200
 ppm
400
ppm
                    Ataxia;
                    gait
None.  It was
shown, however,
that acrylamide
readily crosses
the placenta.
Developmental
Toxicity
(1-Cten. Repro.)

(American
Cyanamid, 1980)
Reproduction
and Fertility

(Zenick et al.,
1986)
Sprague-Dawley Rat       50        25
Oral—diet                PE*&
0, 25, 50 ppm or
0, 2.5, 5.0 mg/kg/d
3 weeks prior to
mating gd 1-19
Long Evans Rat           50        25
Oral-drinking            ppm       ppm
water
0, 25, 50, loo ppm
                   dose       dose     Decreased body
                   unspeci-   unspeci- weight gain;
                   f ied       fied     slight alopecia

                                       Nerve fiber
                                       degeneration in
                                       sciatic and optic
                                       nerves
                                        tfallerian de-
                                        generation of
                                        tibial nerve,
                                        unilateral optic
                                        nerve degene-
                                        ration (dose
                                        level unspeci—
                                        fied)
                   50 ppm     25 ppm
                                            aased birth
                                                          weight
         decreased body
         weight (50 ppm);
         decreased fluid
         intake (50 ppm);    Decreased body
         increased M»*mtitih  weight gain
         splaying (100 ppm); thru day 42

-------
            Table 6-1  (continued)

            Sunmary of Acxylamide

Developmental and Female Reproductive Tendedty
1
STUDY
TYPE (Ref)
Developmental/
Neonatal
(Halden et al.,
1981)
Iwo-Generation
Reproduction
(Naloo Chemical
00. 1987)
Species, Route
dose levels
Fischer 344 Rat
Oral-gavage
20 mg/kg/d, gd 7-16
Fischer 344 Rat
Oral-drinking water
0, 0.5, 2.0, 5.0 mg/
kg/d for 10 weeks
throughout gestation
and lactation, 11
weeks for second
generation.
Maternal Developmental Observed
LOEL NOEL LOEL NOEL Maternal
— - — - 2.0 mg/kg — — None reported
2.0 0.5 5.0 mg/kg 2.0 m/kg Increased
mg/kg mg/kg peripheral
neuropathy
Decreased body
weight
Decreased body
Effects
Developmental
Changes in various
iHt^w?r* i ^Tft • flini?rvifto
levels measured
in the neonate
(see text).
Increased
resorptions
per litter
Decreased
litter size
                                             weight gain

                                             Decreased No.
                                             of litters
                                             (fecundity
                                              index)'

                                             Increased pre-
                                             implantation
                                             loss (5.0 mg/

-------
Reproductive
Tcodcity
Toodcity Assay
(Sakamoto 6
Hashimoto, 1986)
                                                  Table 6-1 (continued)

                                                  Summary of Aczylamide

                                      Developmental and Female Reproductive Toodcity
STUDY
TYPE
(Ref)
species
dose
, Route
levels
Maternal
IOEL NOEL
Developmental
LOEL NOEL
Observed
Maternal
Effects
Developmental
ddY Mouse
Oral-drinking water
0-5mM, 4-6 weeks.
1.2mM     0.9 mM  -
  or        or
0.447     0.335
 mg/kg/d   mg/kd/d
1.2mM     Increased No.
          resorptions
          per
None reported

-------
   Summary    of  Acrylamide





Male   Reproductive      Toxicity





           Table   6-2

STUDY
TYPE (Ref)
Dominant
Lethal and
(2-Gen. Repro.)

(Nalco Chemical
Co., 1987)









Testicular
Effects

(Hashimoto et
al., 1981)


Dominant Lethal


Smith et al..
(1986)




Species, Route
and Dose levels
Fisher 344 Rat
Oral-drinking
water

0, 0.5, 2.0, 5.0
mg/ kg/day for
10 weeks.
11 weeks
for second
generation

^
1


ddY Mouse
Oral-gavage

0, 35.5 mg/kg
2/wk for 8-10 wk.
Av. daily dose =
10.1 mg/kg
Long Evans Rat
Oral-drinking
water
0. 15, 30, 60 ppm
or
0, 1.5, 2.8, 5.8
mg/kg/day for 80
days
1

LOEL
2.0
mg/kg/d













35.5
mg/kg





2.8
mg/kg/d



;


i

NOEL
0.5
mg/kg/d




















1.5
mg/kg/d






I

Observed Effects
Increased peripheral
neuropathy

Decreased body
weight

Decreased body
weight gain

Decreased No. of
litters (fecundity
index)
Increased pre-
i replantation ftss
(5.0 mg/kg/d)
Degeneration of
testicular
epithelia
Weakness and ataxia
in hind limbs


Increased pre-
implantation loss
(high dose only)

;
increased post-
i implementation
loss
              6-6

-------
   Summary    of  Acrylamide





Male   Reproductive      Toxicity





   Table   6-2   (continued)
1
STUDY Species, Route
TYPE (Ref) and Dose levels
Reproductive ddY Mouse
Toxicity Assay Oral -drinking
(Sakamoto & uater
Hashimoto, 1986) O.Snfl, 4-6 wks
4.2, 8.8, 13.1,
17.4 mg/kg/day
for
18.7 mg/kg/day
for

Reproduction Long Evlhs Rat
and Fertility Oral-drinking
water
(Zenick et al., male: 0, 50, 100,
1986) 200 ppm or
0, 4.2, 7.9, 11.6
mg/kg/day for 10
weeks
Dominant Long Evans Rat
Lethal Oral-drinking
water
(Sublet et al., 0, 5. 15, 30, 45,
1986) 60 mg/kg/d for 5
days


1

LOEL
8.8
mg/kd/d








100 ppm







30
mg/kg/d






1

NOEL
4.2
mg/kd/d








50 ppm















1

Observed Effects
Decreased fertility
rate
Decreased No.
fetuses/dam
Increased No.
resorpt i ons/dam




Decreased copulatory
performance increased
hindlimb splaying)
-
^



Increased pre- implan-
tation loss;

Increased post- implan-
tation loss;

Effects seen primarily
in weeks 1-3 post mating
              6-7

-------
   Summary    of  Acrylamide





Male   Reproductive      Toxicity





   Table   6-2   (continued)

STUDY
TYPE (Ref)
Testosterone
Assay
(Ali et al..
1983)

Dominant Lethal
Shelby et a I.,
1986)


Spermhead
Abnormality
(EPA, unpublished)
data)



1
Species, Route
and Dose levels
Fischer 3U Rat
Intraperitoneal
0, 10, 20 mg/kg/
day for
20 days
C3H X101 hybrid
mouse
i.p. 125 mg/kg
^
or A
5 X 50 mg/kg/day
•
B6C3F1 mouse
25, 50, 100 mg/
kg/day X 5 days




1

LOEL
1
20 mg/kg |




50
mg/kg/d -



100
mg/kg/d





1

NOEL
1
10 mg/kg




	











1

Observed Effects
Dose-dependent decrease
of testosterone and
pro I act in


Increased post-
implantation loss


-
Decreased testes:
body ut. ratio
(100 mg/kg/d)

Increase incidence
of spermhead
abnormalities.
              6-8

-------
in fetal weight of the latter which might be attributed to the
lower maternal food intake of the treated animals.  No
macroscopic abnormalities of organ or skeletal structure were
seen in fetuses from Group II.  The litters in Groups I and III
gained weight normally until weaning/ and no differences in gait
(a neurotonic effect) were observed.  No abnormalities were found
in the brains/ spinal cords or sciatic nerves of these rats at 6
weeks of age/ either macroscopically or by light microscopy in
this relatively insensitive (too few animals) study.  Edwards
(1976) also showed that acrylamide was present in fetal tissues
at levels similar to that of dams.  These findings have been
corroborated in other species by o.ther investigators using 14C-
labeled acrylamide (Ikeda et al./ 1983, 1985; Marlowe et al./
1986).                                                     i
     Walden et al. (1981) monitored the concentrations of
selected intestinal enzymes in rats exposed to acrylamide (20
mg/kg/day) io utero and during lactation.  While significant
changes in the levels of acids and alkaline phosphatases/ beta-
glucuronidase/ citrate synthetase/ and lactate dehydrogenase were
observed/ a consistent pattern of toxicity did not emerge/ and
the toxicological importance of these findings is not apparent.
     Zenick et al. (1986) exposed female Long-Evans rats to
acrylamide in drinking water at levels of 0, 25, 50 or 100 ppm
for 10 weeks.  Exposure began two weeks prior to mating and
continued throughout mating/ gestation, and lactation, Hindlimb
splaying appeared in the 100 ppm group during the first to second

                               6-9

-------
week of pregnancy.  Body weight and fluid intake were also
depressed.  Dams in the 50 ppm group showed depression of these
parameters during the last two weeks of lactation.  Acrylamide
did not significantly affect mating performance, pregnancy rates/
litter size or survival.  However, acrylamide did significantly
depress pup body weight at birth  (100 ppm group) and weight gain
during lactation through post-weaning, day 42 (50 and 100 ppm
groups).  Vaginal patency was delayed in the 100 ppm group only.
These data indicate that acrylamide has a deleterious effect on
pup development when dams are exposed at levels of 50 ppm and
above, suggesting that 25 ppm is a NOEL for this endpoint.
     In a modified one-generation xeproduction study, groups of
20 female Sprague*Dawley rats were treated with 0, 25, or 50 ppm
of dietary acrylamide for 2 weeks before mating (Biodynamics,
1979; Spencer and Schaumberg, 1977).  After mating, the rats were
maintained on the 25 or 50 ppm acrylamide diets during 19 days of
gestation.  Although the rats on the 50 ppm of dietary acrylamide
gained less weight, there were no statistically significant
effects due to acrylamide treatment on maternal body weight,
maternal mortality, mating or pregnancy rate.  The pups did not
display any gross malformations at birth and there were no
statistically significant differences in litter size, pup weight
or viability between controls and groups exposed to acrylamide.
At weaning (day 21 of age), 4 pups from the control group (2 of
each sex) and 8 pups from each of the treatment groups (4 of each
sex) were examined histologically for neurotoxicity.  Four of the

                              6-10

-------
pups exposed to aery1amide in utero (dose unspecified) displayed
Waller!an degeneration in the tibial nerve, compared with one of
the pups from the control group.  Three of the pups exposed to
aery1amide in utero (dose unspecified) had unilateral optic nerve
degeneration; no optic nerve degeneration was observed in the
control group.  There was some fiber degeneration in the sciatic
and optic nerves of the treated, but not the control pups.
     Thus, it appears that in utero exposure to acrylamide
results in neurotoxic manifestations in the offspring, possibly
at a dose which produced no overt signs in the mother.  Since
these neurotoxic manifestations are degenerative in nature, they
cannot be considered to be an adverse effect on the development
of the conceptus.* Rather, it appears that these effects occurred
during late gestation (after major organogenesis)  or in the*post-
natal, pre-weaning period.
     The available pharmacokinetics data suggest that the half-
life of acrylamide metabolites is up to 8 days (see section 4).
Since dosing of the dams continued until gestation day 19, and
acrylamide is known to cross the placenta (Ikeda et al., 1983,
1985; Marlowe et al., 1986), it is expected that significant
quantities of acrylamide or its metabolites could reach the late-
gestation fetus as well as the neonate during the first few post-
natal days.  These considerations could provide an explanation
for the presence of degenerative changes in the neonate following
exposure of the dam during gestation.
                              6-11

-------
     The apparent discrepancy between these results  (fetal
effects at 25 or 50 ppm) and those reported by Edwards  (1976;
little or no direct fetal effects at 400 ppm) is striking.  The
major protocol difference is that American Cyanamid Company
(1980) dosed animals for 2 weeks prior to mating, whereas Edwards
(1976) began oral dosing from gestation day 1.  Dosing for the 2
additional weeks may have resulted in more aery1amide or its
metabolites being available via the milk during the pre-weaning
period, thereby raising aery1amide concentrations in the neonate
to toxic levels.
     Zenick et al. (1986) show that effects in the neonate were
observed at dose levels similar to those used by American
Cyanamid Company 41980).  At the high dose, maternal toxicity was
                  \
clearly more severe in the Zenick study than in the American
Cyanamid study.  While the nature and extent of the neonatal ~
effects reported in Zenick et al. suggest that they result from
maternal toxicity, the neurotoxic effects reported in American
Cyanamid suggest a primary effect on the fetus.  Although the
American Cyanamid study is an incomplete protocol and is not
fully reported, this apparent discrepancy in results should not
be ignored.  Strain-specificity in rats, observed in acrylamide-
induced male reproductive toxicity (see section 6.3), may be a
possible explanation.  Further study is needed before definite
conclusions can be reached.        :
                               6-12

-------
6.3. Male Reproductive Effects
     All et al. (1983) studied the effect of acrylamide on the
levels of circulating testosterone/ growth hormone, and prolactin
in male F344 rats.  Animals were injected ip, daily for 20 days,
with an aqueous solution of acrylamide at doses of 0, 10 or 20
mg/kg.  Twenty-four hours after the last treatment, the animals
were decapitated and hormones were assayed by radioimmunoassay.
A dose-dependent depression of testosterone and prolactin
concentrations was observed, with statistical significance seen
only at the 20 mg/kg dose (P<0.05).  Growth hormone concentration
was not altered.  Decreased levels of testosterone may lead to
decreased testicular function resulting in impotence, reduced
seminal volume, and sterility.
     The toxic effect of acrylamide and several related compounds
on the testes was investigated by Hashimoto et al. (1981).  Male
mice (ddY strain, 5 to 6 weeks of age, weighing about 29 g)  were
distributed randomly into groups of 5 to 7 animals.  Groups were
dosed orally with test compounds, twice weekly for 8 to 10 weeks,
at a concentration of 35.5 mg/kg.  In order to examine the effect
of metabolic activation, a parallel study was conducted in which
mice were injected ip with phenobarbital (PB) at a dose of 50
mg/kg for 5 successive days per week, from 1 week prior to the
study start until the last week of the study.
     Acrylamide was neurotoxic at 35.5 mg/kg, producing ataxia
and weakness in the hindlimbs.  It also produced both testicular
atrophy and significant reduction in testes weight, with

                               6-13

-------
degeneration of the epithelial cells of the seminiferous tubules.



However, the interstitial cells were normal.  Testicular damage



appeared to be completely prevented or reduced by PB



pretreatment, as demonstrated in the parallel study.  The authors



concluded that PB may enhance the inactivation of aery1amide.



Since methacrylamide (another neurotoxic compound) did not



produce any testicular effects when given at a neurotoxic dose,



the data suggest that neurotoxicity and testicular toxicity may



occur by different mechanisms.



     In a subsequent study, Sakamoto and Hashimoto (1986)



demonstrated that acrylamide decreased the fertility rate in male



ddY mice.  Increased resorptions per dam resulted when treated



male or female mice were mated to unexposed animals.  These
                  i


effects on reproduction occurred at levels above the threshold



for neurotoxicity.  The dosing regimen was significantly



different from that used in the earlier study.  Animals were



given acrylamide in the drinking water 4 to 6 weeks prior to



mating.  In the earlier report, twice weekly oral intubations



were given for 8 to 10 weeks.  While pharmacodynamics and other



factors must be evaluated before a firm conclusion can be



reached, it appears that the reproductive organs are not more



sensitive to acrylamide toxicity than is the peripheral nervous



system.



     The reproductive effects of acrylamide also have been



evaluated in a dominant lethal/two-generation reproduction assay.



Acrylamide was administered to male and female F344 rats in the





                               6-14

-------
drinking water for 10 weeks (Nalco Chemical Co., 1987).  Dose
levels were 0, 0.5, 2.0 and 5.0 mg/kg/day.  Following mating of
these treated males to the treated females, the treated males
were remated to untreated females (mating periods unspecified).
Following mating, administration to treated females continued
throughout gestation and lactation.  The same basic protocol was
followed for the second parental generation except that treatment
was for 11 weeks.
     The dominant lethal results are as follows:
     1.  Although the number of eggs ovulated per untreated dam
was equal among all dose groups, there was a significant
reduction in total implants per litter at 5.0 mg/kg/day.
Therefore pre-implantation loss was significantly increased.
     2.  There was also a significant increase in non-viable
implants per litter, i.e. resorptions, at 5.0 mg/kg/day.
Accordingly, post-implantation loss was increased.
     The reproductive results (first generation) are as follows:
     l.  Although males given aery1amide were able to
successfully mate with females (no effect on mating index), the
litter size and the number of litters was reduced at 5.0
mg/kg/day compared to controls.  Since both sexes were dosed, it
is not clear whether this is a male- or female-mediated
observation.  Although the similarity of these results with those
observed in the dominant lethal assay suggest that this is a
male-mediated effect, this does not preclude the possibility of
an effect in females as well.

                               6-15

-------
     2.  Additionally, parental animals in the 2.0 and 5.0
mg/kg/day groups displayed evidence of ataxia.  Animals in these
groups had lower body weight gains than the controls.
     The results of the second generation are as follows:
     1.  No effect on mating, number of females pregnant, or
gestation length.
     2.  At 5.0 mg/kg/day the number of implantations per dam and
the number of live pups per litter were significantly reduced.
Post-implantation losses were significantly increased.
     3.  Gestation weight gain was decreased at 2.0 and 5.0
mg/kg/day.  Lactation weight gain was increased at 5.0 ug/kg/day.
     Acrylamide-induced peripheral, neuropathy was produced in
male rats at 5.0 mg/kg/day.  Thus, it appears that aery1amide is
reproductively toxic in the rat at doses which may also be i
neurotoxic.
     Shelby et al. (1986) administered acrylamide either as a
single ip dose at 125 mg/kg or as 5 daily doses at 50 mg/kg/day
to hybrid mice (C3H x 101) in a dominant lethal assay.
Subsequent matings to T-stock and C3H x 101 females showed
significant post-implantation loss following both dosing
regimens, from matings between days 4.5 and 11.5 after treatment.
This study suggested that late spermatids and early spermatozoa
were affected.
     Zenick:et al. (1986) exposed male Long-Evans rats to
acrylamide in drinking water at levels of 0, 50, 100 or 200 ppm
(0, 4.2, 7.9, and 11.6 mg/kg/day, respectively) for 10 weeks.  In

                              6-16

-------
males, copulatory behavior, seminal parameters, fertility
(controls and 100 ppm only) and fetal outcomes were evaluated
using untreated females.  Hindlimb splaying was apparent in the
200 ppm group by week 4 and less severe splaying appeared in the
100 ppm group at week 8.  Disruptions in copulatory behavior
preceded the appearance of this atazia.  These disruptions in
mating performance interfered with ejaculatory processes and
subsequent transport of sperm, since semen was found in the
uterus of only one of the females mated with the 100 ppm group at
week 9.  Moreover, only 33% of the females mated (week 10) to the
100 ppm group were pregnant.  Post-implantation loss was also
significantly increased.  These data indicate that aery1amide had
a deleterious effect on copulatory behavior, fertility, and fetal
survival.  As fertility and fetal outcomes were not evaluated at
50 ppm, the NOEL in the male could not be determined.  Since '
effects were dependent upon cumulative dosing (effects at 1/2
dose level took twice as long to be observed) it is possible that
neurological effects were occurring, but not yet evident, at
these low levels (see section 5).
     To determine the coexistence of neurological and
reproductive effects, a male reproductive study was conducted at
lower drinking water doses (0, 15, 30 and 60 ppm or 0, 1.5, 2.8,
and 5.8 mg/kg/day, respectively) with Long-Evans rats (Smith et
al., 1986).  Males were exposed to acrylamide for a total of 80
days.  After 72 days, treated males were mated with untreated
females until each had impregnated two females or until day 80.

                               6-17

-------
Females were sacrificed on day 14 of gestation and examined for
corpora lutea and for living and dead fetal implants.  Fertility
rate and pre- and post-implantation losses were measured.  In
addition, half of the males were sacrificed and cytological
preparations of the testes and the sacral, sciatic,and tibial
nerves were made.  The remaining males were sacrificed 12 weeks
later.
     There were no effects on body weight, food and fluid intake,
or fertility.  Pre- and post-implantation losses were both
evident at the high dose (60 ppm), while post-implantation loss
was also evident at the mid-dose (30 ppm).  No gross or
histopathological signs of nerve damage were observed at any
dose.  Upon sacrifice immediately following treatment, testicular
preparations showed no increase in aberrations.  In animals*
sacrificed 12 weeks later, a total of 4 reciprocal translocations
were observed in the treated animals (1, 1, and 2 in the 15,  30,
and 60 ppm groups, respectively).
     In a subsequent study using higher doses, Sublet et al.
(1986) observed effects at times that suggest that spermatozoa
and spermatids represent the germ cell stages most sensitive to
aery1amide's effects.  These investigators have shown that the
reproductive toxicity of aery1amide in the Long-Evans rat is a
primary effect on testicular tissue and its contents, since the
effects were observed in the; absence of neurotoxicity.  These
findings are strengthened by the results of Marlowe et al. (1986)
who studied the distribution of 14C-labeled aery1amide in male

                               6-18

-------
Swiss-Webster mice by whole-body autoradiography.  They observed
that significant amounts of aery1amide appeared in the testes one
hour after oral administration.  Radioactivity was present up to
9 days following injection, indicating that aery1amide has ample
opportunity to exert a direct toxic effect on the testes.
     Mccollister et al. (1964) exposed male Dow-Wister rats
(number unspecified) to 800 ppm of aery1amide in the diet for 10
days, followed by 400 ppm for 9 days.  Following 51 days of
recovery, gross and microscopic examination revealed marked
tubular degeneration in the testes of the two remaining rats (the
others had died during and after dosing).  In addition, histolo-
gical examination revealed no effect on the testes at 30 ppm.
The testes of intermediate dose animals (70, 90, 110 and 300 ppm)
apparently were not examined histologically, so a thresholds dose
has not been defined.  These data provide limited information-on
reproductive toxicity, as testicular effects were noted only at
lethal doses.
     A sperm-head abnormality study was conducted in B6C3F1 mice
exposed to aery 1 amide by 5 daily doses at 100 mg/Jcg ip (J. Meier,
USEPA, Cincinnati, OH, personal communication).  Increased
incidences of sperm-head abnormalities and decreased testes
weight to body weight ratios were observed.
                               6-19

-------
                      7.  GENOTOXIC EFFECTS


7.1. summary and Conclusions



     For years, the neurotoxicity of acrylamide has been the


primary basis for its human health concern.  Consequently, it


appears, that neither the mutagenic nor the carcinogenic potential


of this important industrial chemical has been extensively


examined until recently.  Despite this relatively late interest


in aery1amide's potential mutagenicity and carcinogenicity,


evidence for aery1amide's genotoxic potential has appeared in the


literature.


     The major concern for acrylamide's genotoxicity centers on


its clastogenic activity.  This effect appears more pronounced in


the germ cells as»compared to somatic cells.  The interaction
                  t

with germinal tissues suggests the possible heritability of^


acrylamide-induced DNA alterations.  If the transmission of  -


acrylamide-altered DNA is demonstrated/ this could have


implications for a human health concern for future generations.


Other reports support the evidence that acrylamide can interact


with DNA, e.g., DNA binding and induction of aneuploidy and


polyploidy.  On the other hand, the weight of evidence from


available reports suggests acrylamide may not produce detectable


gene mutations.  Since most of this evidence is derived from in


vivo studies, it is not clear at this point whether or not


acrylamide requires metabolic activation to exert its genotoxic :


effects.  The genotoxicity results available to EPA to date are


summarized in Table 7.1




                               7-1

-------
     The revised guidelines for mutagenicity risk assessment
published by EPA (51 FR 34006, 1986) allow for an evaluation of
the potential genetic risk associated with human exposure to
aery1amide.  The genotoxicity data for aery1amide suggest an
intrinsic mutagenic (clastogenic) activity and provide sufficient
evidence for chemical interaction in the mammalian gonad.  The
body of evidence suggests that aery1amide may induce alterations
in the genome of germinal cells.  These data provide a fairly
strong weight of evidence bearing on the potential for human
germ-cell mutagenicity and its heritability.  Heritable
translocation results increase the strength of a human germ-cell
mutagenicity concern to a level that would only be further
strengthened by direct human evidence (the highest level of
evidence for human mutagenicity).  Data from a rigorous, well-
conducted heritable translocation assay may allow for a possible
quantitative risk assessment that may be tied to human exposure
data concerning acrylamide.
                               7-2

-------
                    Table 7-1  Summary Table of Genotoxic Results for Acrvlamide
  A3 sav

Gene Mutations

 Salmonella
  strains TA97,TA98,TA100,
  TA102,TA1535,TA1537,
  TA1538 + activation
Result
                                            a)
              Highest
           Dose  Tested
negative    30 mg/plate
   Reference
                              Lijinsky & Andrews, 1980
                              American Cyanamid, 1983a
                              Bull et al., 1984a
                              Institut d'Hygiene et
                                d'Epidemiologie, 1985
                              Hashimoto & Tanii, 1985
                              NTP, 1985
                              EPA, unpublished results
 Mouse Lymphoma - no activation

 CHO/HPRT, ± activation

 Drosophila sex-linked recessive
 lethal - feeding

Chromosomal effects - in vivo

 Aberrations in mouse bone
 marrow - feeding (in diet)
        - i.p. injection

 Aberrations in mouse
 spermatogonia - feeding (in diet)
               - i.p. injection
positive   850 ug/ml
                  i
negative  1200 ug/ml

negative   100 ppm
negative
negative
positive
positive
            500 ppm
            100 mg/kg
            500 ppm
            100 mg/kg
                              Moore et al.,  in press

                              American Cyanamid,  I983b

                              American Cyanamid,  I985a
Shiraishi, 1978
Shiraishi, 1978
Shiraishi, 1978
Shiraishi, 1978

-------
              Table 7-1  Summary Table of Genotoxic Results for Acrvlamide (continued)
 Chromosomal effects - in vivo

 Dominant lethal
 in rats - drinking water

 in mice - single or
           multiple
           i.p. injections

 Mouse micrbnucleus - oral gavage
                    - oral gavage

 Heritable translbcations
                in mice - multiple
                i.p. injections

Chromosomal effects - in vitro

 Human blood lymphocytes
 - aberrations
 - 8CB

 CHO/8CE - 2 hr. exposure, +
           activation
           24 hr. exposure, no
           aestivation

 Micronucleus - Chinese hamster
                fibroblasts -
                3 hr exposure
              Highest
Result10    Dose Tested
                    Reference
positive


positive
negative
negative
positive
positive
positive
negative
positive

positive



negative
5.8 mg/kg/day    Smith et al., 1986
125 mg/kg
5X50 mg/kg/day

2x75 mg/kg/day
5X100 mg/kg/day
5x40 mg/kg/day
5x50 mg/kg/day
50 ug/ml
so ug/ml
3.25 mg/ml

0.175 mg/ml
10"3 M
                 Nalco Chemical Co., 1987
                 Shelby et al., 1986

                 American Cyahamid, 1983c
                 EPA, unpublished results
                 Shelby et al.,  1987
                 Shelby et al.,  1987
                 Institut d'Hygiene et
                  d'Epidemiologie,  1985
                 EPA,  unpublished results

                 EPA,  unpublished results
                Zaichkina and Ganassi,
                         1984
 Mouse Lymphoma - no activation
positive ,    850 ug/ml
                 Moore et al.,  in press

-------
             Table 7-1  Summary Table of Genotoxic Results for Acrvlamide (continued)
                                                   Highest
Chromosomal effects - in vivo
Other effects
Mouse spermhead
UDS/primary rat

abnormality -
oral gavage
hepatocytes
Transformation with: C3H/10T 1/2,
no activation
NIH/3T3 cells,
no activation
BALB/C-3T3 cells,
+ activation
Result*0
weak
positive
positive
negative
positive
positive
p
positive
Dose Tested
5X100 mg/kg/
day
3.3 mg/ml
10~2 M
200 ug/ml
200 ug/ml
800 ug/ml
Mitotic recombination in
          Saccharomvces - D7,
          ± activation
negative
500 ug/ml
                                                                       Reference
                                                                   EPA, unpublished results

                                                                   American Cyanamid, I983e, 1985b
                                                                   Miller and McQueen, 1986
                                                                   Banerjee and Segal, 1986

                                                                   Banerjee and Segal, 1986

                                                                   American Cyanamid, 1985c, d
Institut d'Hygiene et
 d'Epidemiologie, 1985
Amplication of SV40 DMA inserts of
SV40 transformed
 Chinese hamster cells

Alteration of transfection of
E. coli CR63 cells
 weak
 positive
150 ug/ml
vanhorick and Moens, 1983
 positive     10 ug/O.l ml   Vasavada and Padavatty, 1981
a) results are determined for all testing conditions listed

-------
7.2. Gene Mutation Assays

7.2.1.  Salmonella Assays
     Acrylamide has been tested for mutagenicity by several
investigators in the Salmonella/mammalian activation assay.  The
evidence indicates aery1amide does not induce gene mutations in
this assay (Lijinsky and Andrews, 1980; American Cyanamid Co.,
1983a; Bull et al., 1984a; Institut d'Hygiene et d'Epidemiologie,
1985; Hashimoto and Tanii, 1985; NTP 1985).   Acrylamide did not
produce Salmonella revertants in strains TA98, TA100, TA1535,
TA1537, and TA1538 with or without activation.  In some
instances, concentrations up to 30-mg/plate  were used and both
plate incorporation and liquid suspension procedures were
performed.                                                 %
     Furthermore, aery1amide was tested in the plate
incorporation method with the newer Salmonella strains TA97 and
TA102 (J. Meier, USEPA, Cincinnati, OH, personal communication).
The results were negative up to a dose of 30 mg/plate, with or
without activation.

7.2.2.  Eukaryotic Gene Mutation Assays
     Acrylamide was tested for its ability to induce mutations at
the hypoxanthine guanine phosphoribosyl transferase (HPRT) locus
in cultured Chinese hamster ovary (CHO) cells at doses up to 1200
ug/ml with and without metabolic activation  (American Cyanamid
Co., I983b).  Although increased mutation frequencies were

                               7-6

-------
observed at 300 ug/ml with activation and at 37.5 ug/ml without
activation, a dose-response relationship was not observed.
However, because these increases appear at unique doses, an
additional experiment should be performed to confirm these
results before suggesting potential mutagenic activity by
aery1amide in this assay.
     Aery1amide was tested without exogenous activation in mouse
lymphoma L5178Y TK-f/- cells for mutation at the thymidine kinase
                                                             /
locus (Moore et al., 1987).  Acrylamide induced a dose-dependent
increase in mutation frequency up to concentrations of 850 ug/ml.
All doses gave survivals above 10%.  It appears that the
increased mutation frequency is due almost exclusively to
induction of smalt, colonies.  Moore et al., (1985) have suggested
that the mouse lymphoma assay may be capable of evaluating "hot
only gene mutations, but clastogenic events as well.  Small
colony formation appears to represent chromosomal alterations
whereas large colonies represent gene mutations (Hozier et al.,
1985).  These data with mouse lymphoma cells, therefore, may
represent a clastogenic event (see Section 7.3.2).
     A Drosophila sex-linked recessive lethal assay was performed
by feeding 100 ppm aery1amide to Canton-S wild type males for 72
hours (American Cyanamid Co., 1985a).  These were mated to Base
stock females for a total of three broods.  The dose of 100 ppm
caused 100% sterility in broods 2 and:3.  Therefore, 100 ppm
acrylamide was fed to males for 24 hours.  Sterility was still
noticeable, but the assay was able to be performed.  Over 5,000

                               7-7

-------
lethal tests each were examined in the treated and the control
groups over two experiments.  No increase in the percent lethals
was observed over controls indicating that aery1amide did not
induce sex-linked recessive lethals in post-meiotic germ cells of
treated males.

7.2*3.  S'm|imary of Gene Mutation Assay Results
     Aery 1 amide does not appear to induce gene mutation in three
   >
types of gene mutation assays examined:  the Salmonella/mammalian
activation assay, the CHO/HPRT mutation assay, and the Drosophila
sex-linked recessive lethal assay.  The mouse lymphoma mutation
data suggest that aery 1 amide may i-nduce mutations in an
eukaryotic gene mutation assay, but these data may reflect a
clastogenic event due to the predominant formation of small'1
colonies.  It may be relevant that the CHO sub-clone generally
used in the CHO/HPRT mutation assay does not appear to be
sensitive to clastogens (Hsie et al., 1986).  The remaining assay
that would confirm the apparent lack of gene mutation activity by
acrylamide, as well as possible heritability resulting from such
activity, would be the mouse specific locus assay.  However,  due
to the weight of evidence indicating no mutational activity,  this
assay is not recommended for further assessment regarding gene
mutations.
                               7-8

-------
7.3. Chromosomal Assays

7.3.1.  In Vivo Effects
     The first evidence demonstrating that aery1amide induced
chromosomal effects was reported by Shiraishi  (1978).  Acrylamide
was administered to male mice either 1) in the diet  (500 ppm) for
1, 2 or 3 weeks exposure, or 2) by a single ip injection of 50,
100, 150 or 200 mg/kg with animals held for 12 and 24 hours and
11 and 12 days post-exposure (LD50 determined to be 200 mg/kg;
only doses of <100 mg/kg provided data).  No increase in
frequency of chromosome aberrations was observed in bone marrow
under any conditions.  However, significant increases in
chromosome breaks \(up to 19% vs. 2.4% controls) and clear
chromatid exchanges (up to 6% vs. 0% controls) were observed in
spermatogonia over 2 to 3 weeks of feeding.  Increases in breaks
were also observed 11 to 12 days post-injection (up to 12.5% vs.
3% controls).  There were, however, noticeable decreases in
mitotic activity in both bone marrow and spermatogonia.  The
total number of aneuploid and polyploid cells increased with time
after treatment in both these cell types.  Acrylamide also
induced significant increases in aberration frequency in primary
spermatocytes under both treatment conditions.  The aberrations
included univalents, fragments such as breaks and deletions, and
rearrangements such as chain quadrivalents and ring quadriva-
lents.  sister chromatid exchange (SCE)  frequencies in bone
marrow and spermatogonia were not increased over controls under

                               7-9

-------
any treatment condition.  Acrylamide significantly reduced testes
weight suggesting possible cell killing.  These results suggest a
genotoxic hazard to the male germ cell posed by acrylamide.  This
report is noteworthy as it demonstrates that a chemical is
clastogenic to germ cells, but not to the examined somatic cells
in the same test animal.
     Further evidence for germ cell genotoxicity in the absence
of cell atresia, predominantly through chromosomal effects, is
found in several dominant lethal studies in rodents.  Smith et
al. (1986) exposed male Long-Evans rats to 15, 30 and 60 ppm (0,
1.5, 2.8 and 5.8 mg/kg/day, respectively)  acrylamide in their
drinking water.  They were treated.for ten weeks and then mated
at the end of this; period.  There were no observable differences
in weight gain or fluid consumption from controls and the  *
fertility rate was comparable in all dose groups.  An increased
pre-implantation loss was seen only at the highest dose where an
approximately 2.5-fold elevation over the negative control was
observed.  An increased post-implantation loss was observed at
the two highest doses with up to a six-fold increase over the
negative control.  Also, no peripheral neuropathies as determined
by light microscopy of the sciatic nerve nor hindlimb splaying
(characteristic of acrylamide neuropathy)  were observed.
     These investigators also performed a preliminary dominant
lethal study using a 5-day treatment regimen followed by serial
matings  (Sublet et al., 1986).  Rats were dosed with 0, 5, 15,
30, 45 and 60 mg/kg/day (100 mg/kg/day was lethal) and then mated

                              7-10

-------
at 1, 2, 3, 4, 7 and 10 weeks after exposure.  Significant pre-
implantation losses were observed above 5 mg/kg and post-
implantation losses at doses greater than 15 mg/kg.  The results
from these two studies appear to suggest that acrylamide affects
the spermatid-spermatozoa stages.
     The Smith et al. (1986) study also examined cytogenetic
endpoints.  After breeding the male rats for the dominant lethal
study, one-half of the males in each group were sacrificed and
testicular preparations were made.  The remaining males were
sacrificed 12 weeks later for analysis of possible reciprocal
translocations in spermatocytes derived from treated sperma-
togonial stem cells.  There was no. increase of aberrations in
animals at the first sacrifice.  From preliminary analyses, they
have noted the presence of 4 reciprocal translocations found in
treated males and none in controls at 12 weeks.  Although no '
statistics were applied, these investigators concluded that this
is suggestive, but not conclusive, evidence of an acrylamide
effect.
     Smith et al. (1986) reported negative findings, based on
preliminary observations, for chromosome aberrations in
differentiating spermatogonia at high, acute doses of acrylamide.
One dose of 125 mg/kg (ip) was given to six C3H x 101 mice and
spermatogonial cell preparations were made.  No chromosomal
aberrations were observed in a total of 10o metaphase cells each
from the control and treatment groups.
                              7-11

-------
     Shelby et al. (1986) reported positive results in a mouse
dominant lethal study after aery1amide exposure.  Male (C3H x
101) Fv hybrid mice were exposed with either a single 125 mg/kg
ip dose (maximum tolerated dose) or 5 multiple 50 mg/kg/day ip
doses of acrylamide.   Males were mated with either T stock
females or (SEC x C57BL) F1  hybrid  females.   An observed
increased frequency of dead implants was apparent from matings
4.5 and 11.5 days after treatment,  with the multiple treatment
more effective than the single dose.  These results were seen at
times that would suggest an acrylamide-induced effect on the
spermatid-spermatozoa stages.
     Another dominant lethal study, was performed with F344 rats
(Nalco Chemical CO,.,  1987).   Male rats were exposed to acrylamide
in drinking water at doses of 0, 0.5/ 2.0 and 5.0 mg/kg/day*for
10 weeks prior to breeding.   A dominant lethal effect was seen at
the top dose only.  Mating performance was unaffected although
animals had lower body weight gains compared to controls.  At 5
mg/kg/ the number of corpora lutea  per dam were unaffected, the
number of implants per litter were  significantly reduced causing
a significant increase in pre-implantation loss compared to
controls (24.9% vs. 14.3% control), and the number of viable
implants per litter were significantly reduced, causing a
significant increase in post-implantation loss compared to
controls (14.3% vs. 6.2% control).   These studies provide
evidence that acrylamide can reach  the germ cells via an in vivo
exposure and produce a genotoxic effect.

                              7-12

-------
     Confirmation of the ability of aery1amide to induce
heritable germ cell mutagenic effects comes from a study of
treated male mice.  Shelby et al.  (1987) administered aery1amide
intro-peritoneally (either 40 mg/kg/d or 50 mg/kg/d for 5d),
mated them and screened male progeny for heritable translocations
by noting reductions in fertility.  A number of presumptive
carriers of translocations were confirmed by cytogenetic
analysis.  A highly significant increase in translocations was
noted in the male progeny/ being 39%, 24% and 0.2% in the 50
mg/kg/d, 40 mg/kg/d and historical control groups/ respectively.
In summary/ the dominant lethal studies in rats and mice and the
mouse heritable translocation study demonstrate the ability of
aery1amide to produce transmissible germ cell mutagenic effects.
     Aery 1 amide has been tested in two studies to assay itsr-
ability to induce chromosome breakage in mouse bone marrow cells
as measured by an increase in micronuclei.  In the first study
(American Cyanamid Co., 1983c)/ male and female CD-I mice were
exposed to acrylamide by gavage under two regimens.  Mice were
exposed to either 1)  75 mg/kg and sacrificed and cells harvested
30 and 48 hours later/ or 2)  2 x .75 mg/kg (24 hours apart) and
sacrificed and cells harvested 48 and 72 hours from initial
treatment.  The 75 mg/kg dose produced piloerection,
hypersensitivity and abnormal gait.  Lethality was seen at doses
of 100 mg/kg and above,  one thousand polychromatic erythrocytes
were scored from each of eight animals (4 of each sex).  Although
there were small/ absolute increases in micronuclei numbers with

                              7-13

-------
either treatment regime, these increases did not appear
biologically significant.
     Another mouse micronucleus study also examined the effect of
aery1amide on CD-I mice (J. Meier, USEPA, Cincinnati, OH,
personal communication).  Ten animals (5 of each sex) were
exposed to 5 consecutive daily doses of 25, 50 or 100 mg/kg/day
by gavage.  Bone marrows were harvested 6 hours after the last
dose.  The negative results corroborate the ones from the
previous study.  However, there did not appear to be any gross
signs of toxicity in contrast to many reports at the 100
mg/kg/day dose.  These micronucleus results are consistent with
the Shiraishi (1978) study where no chromosomal aberrations were
found in bone marfpw.
                                                           ^
7.3.2.  In Vitro Effects
     Aery1amide has been tested for chromosomal effects in
cultured mammalian cells.  Acrylamide was examined for its
ability to induce chromosome aberrations and sister chromatid
exchanges (SCE) in peripheral blood lymphocytes from human donors
(Institut d'Hygiene et d'Epidemiologie,-1985).
Phytohemmaglutinin-stimulated lymphocytes were exposed to
aery1amide at concentrations of 0.1 to 50 ug/ml for 72 hours.
Acrylamide induced significant increases in aberrations (breaks,
fragments, dicentrics, rings and minute chromosomes were scored).
The frequency of SCE/cell was slightly enhanced over baseline,
but neither a doubling at any dose nor a dose-response was

                               7-14

-------
observed.  Acrylamide also increased the incidence of aneuploidy
and polyploidy.  The observed effects were independent of blood
donor.
     The clastogenic response to acrylamide was examined in mouse
lymphoma cells in culture (Moore et al., in press).  Acrylamide
induced both chromatid and chromosome breaks and rearrangements.
A clear dose-response for clastogenicity was observed at
concentrations up to 850 ug/ml (64 aberrations/100 cells) without
metabolic activation.  These results appear to support the
hypothesis discussed earlier (see section 7.2.2) that mouse
lymphoma cells may be capable of detecting clastogenic and
mutagenic events by discrimination, of colony size after mutant
expression.      *
     Acrylamide was examined for its ability to induce SCE in CHO
cells in culture.  CHO cells were exposed to acrylamide for  '
either 24 hours (doses up to 0.175 mg/ml)  or 2 hours (doses up to
3.25 mg/ml) (J. Meier, USEPA, Cincinnati,  OH, personal communica-
tion).  There was a significant increase in SCE/cell, but no
doubling of control values,  after the 24-hour exposure without
activation (tests with microsomal activation were not conducted).
For the 2-hour treatment, a significant doubling of control
values was seen with or without activation.  The level of induced
SCE/cell under activated conditions was not greater than that
seen under nonactivated conditions.  This  suggests a possible
direct acting mechanism for acrylamide.  In another SCE study
(American Cyanamid Co., 1983d), the reported results were

                              7-15

-------
negative, but the negative control values were exceptionally
high.  While the effects at all doses were similar to the
baseline values/ this study should be repeated with acceptable
negative control values to be considered valid.
     A report, translated from Russian, suggests that aery1amide
does not induce micronuclei in cultured Chinese hamster
fibroblasts (Zaichkina and Ganassi, 1984).  Aery1amide was tested
at only one concentration (10~3M) for a 3-hour exposure and the
cells were apparently fixed after a 24-hour culture period.
However, this study is too limited to reach definitive
conclusions.

7.3.3.  Summary of Chromosomal Assay Results
     The studies examining the chromosomal effects of acryl*amide
indicate its clastogenic potential.  This effect appears to be
more pronounced in the germ cells than in somatic cells.  The
suggestive reciprocal translocation results raise the possibility
of acrylamide-induced alterations to DNA that may be transmis-
sible to future generations.  Indeed, aery1amide is currently
being tested in a mouse heritable translocation assay to test for
its possible heritable effects (Shelby and Generoso, personal
communication).
     In vivo as well as the in vitro results suggest aery 1 amide
may induce aneuploidy.  Aneuploidy and polyploidy are consistent
effects in studies which scored for these endpoints.  However,
the reviewers in the Aneuploidy Program sponsored by.the U.S. EPA

                               7-16

-------
suggest that this does not permit the distinction of an exclusive
aneuploidy-inducing ability by acrylamide as the aneuploidy and
polyploidy numbers were reported as one figure  (Cimino et al.
1986).  Acrylamide should be tested for its ability to produce
aneuploidy using appropriate tests and measurements.
     Since most of the clastogenic activity was noted in in vivo
studies, it is uncertain whether metabolic activation is required
to exert its genotoxic effects.  The in vitro results suggest a
possible direct acting mechanism for acrylamide.  The CHO/SCE
study found increases in SCE/cell under nonactivated conditions,
albeit at relatively high concentrations.  In vitro studies with
human lymphocytes in which clastogenicity was observed were
performed without*activation.  However, the metabolic activating
capability of lymphocytes has not been totally explored anoS. it is
known that lymphocytes are capable of activating chemicals such
as cyclophosphamide to genotoxic forms (Waalkens et al., 1981).
The mouse lymphoma results demonstrate that acrylamide induces
aberrations without activation.  However, this direct acting
mechanism may not be entirely clarified by these results.  These
mouse lymphoma cells have been found to detect the mutagenicity
of 2-acetylaminofluorene without addition of exogenous exzymes
(M. Moore, USEPA, Research Triangle Park, NC, personal
communication).  Since 2-acetylaminofluorene is believed to
require metabolic activation to exert its genotoxic effect, these
cells may retain residual exogenous metabolic capability.
                               7-17

-------
Whether this capability is class specific or not has not been
addressed.

7.4  Other Genotoxic Effects
     There are several reports indicating that acrylamide is
capable of binding to DMA (see Section 4).  Briefly/ it has been
shown that radiolabeled acrylamide binds to protein and nucleic
acids in rodents (Hashimoto and Aldridge, 1970; Carlson and
Weaver/ 1985).  Furthermore/ there is evidence to suggest that
acrylamide is capable of direct DNA allcylation via Michael
addition (Solomon et al. 1985).  These investigators noted that
acrylamide reacted most strongly with adenine in in vitro
reactions with calf thymus DNA.
     Two studies were performed to evaluate the effect of  i
acrylamide on mouse sperm-head morphology (J. Meier/ USEPA/
Cincinnati/ OH/ personal communication).  In both studies B6C3F1
mice were exposed to acrylamide at 0, 25/ 50 and 100 mg/Jcg/day
for 5 days by gavage.  Significant mortality was found at the top
dose.  Animals were sacrificed at 3 weeks and 5 weeks.  There was
a significant drop in the testes weight:  body weight ratio at
the upper doses.  However/ the number of sperm per mg of cauda
epididymis remained similar between control and treated groups.
At 3 weeks/ significant increases in the number of abnormal
sperm-heads at 50 and 100 mg/kg/day were noted/ including banana/
blunt hook/ amorphous, pinhead and two headed morphologies.
There was a large increase in abnormal spermheads at the 100

                               7-18

-------
mg/kg/day dose, but it is not clear whether this may be due to
cytotoxicity or genetic effects at this relatively high dose.  At
5 weeks, a response was apparent/ but increased numbers of
abnormal sperm seen at the top dose only.  Again/ whether or not
this is due to cytotoxicity is uncertain.
     There are conflicting reports regarding the ability of
acrylamide to induce unscheduled DNA synthesis (DOS) in primary
rat hepatocytes.  Acrylamide was tested in two UDS assays by
American Cyanamid Company (1983e/ I985b).  It induced a dose-
dependent increase in UDS up to 600 ug/ml (American Cyanamid Co.,
1985b).  However/ only the top dose produced an increase greater
than 5 net nuclear grains (9.1 + 0..9 with triplicate cultures)
with 88% of the cetlls in repair (>4 net nuclear grains).
Cytotoxicity was seen at doses of 750 ug/ml and higher.  Im the
other UDS assay (American Cyanamid Co./ 1983e)/ higher doses Were
apparently able to be assayed and large increases in UDS were
observed (e.g. >40 net nuclear grains at doses of 1 and 3.3
mg/ml).  Doses of 10 mg/ml and higher appeared cytotoxic.
Another report suggests that acrylamide is negative in a
similarly performed assay (Miller and McQueen/ 1986).
Concentrations of acrylamide up to 10'2 M induced no apparent
increase in DMA repair.  This concentration is similar to the
doses tested in the previous reports/ 600 to 1/000 ug/ml.   Higher
concentrations in this study were cytotoxic to the hepatocytes.
In support of the negative finding/ density gradient studies
confirmed the lack of induced DNA repair as [3H] thymidine was

                              7-19

-------
not found to incorporate into nonreplicative DNA after aery1amide
exposure (Miller and McQueen, 1986).  These same investigators
further report that aery1amide did not interfere with the repair
of uv light-induced DMA damage.  The conflicting UDS data
preclude an unequivocal determination of the effect of aery1 amide
on DNA damage and repair.  It should be noted that the
variability of the UDS hepatocyte test is affected by several
factors including the functional state of the animals and the
isolated cells (Lonati-Galligani et al., 1983).
     Acrylamide has been tested in several in vitro
transformation assays.  It exhibited dose-dependent
transformation in both NIH/3T3 mouse fibroblast cells and C3H/10T
1/2 cells up to 100 and 150 ug/ml, respectively (Banerjie and
Segal, 1986).  All morphologically transformed cells were  i
subsequently shown to grow in soft agar, which suggests malignant
potential by the transformed cells.  Two additional in vitro
transformation assays with BABL/C-3T3 cells examined under
activated conditions support the published finding (American
Cyanamid Company, 1985c, 1985d).  These two assays examined
acrylamide's transforming potential over two dose ranges (10 to
100 ug/ml and 100 to 800 ug/ml) in the presence of exogenous
metabolic activation.  Acrylamide apparently produced little
cytotoxicity, but induced a dose-dependent increase in
morphological transformants (e.g., type III foci)  within each
study.
                              7-20

-------
     Aery1amide was not  able to  significantly increase  the
mitotic recombination frequency  at the trp 5 locus in
Saccharomyces cerevisiae D7 (Institut d'Hygiene et d'Epidemio-
logie, 1985).  Saccharomyces were exposed to aery1amide
concentrations of l to 500 ug/ml with or without exogenous
activation,  for either 2 hours with shaking at 37'C or  17 hours
with shaking at 29*C.  The top dose was slightly toxic  to the
yeast/ but the conversion frequency was not increased with either
treatment.
     Aery1amide is a weak inducer of the amplification  of 8V40
DNA inserts  of 8V40 transformed Chinese hamster cells (Vanhorick
and Moens, 1983).  This  effect might result from a weak DMA-
damaging activity*demonstrated by the ability of high
concentrations of aery1amide to inhibit the DNA synthesis rate of
the cells.   Furthermore, acrylamide synergistically increased*the
enhancement  of 8V40 DNA amplification by 6 established
carcinogens  in a dose-dependent manner.  This potentiation of
effects by other chemicals may need to be taken into account when
assessing the risk of acrylamide.  Amplification of 8V40 DNA may
provide intrachromosomal sites for DNA breakage (DNA-damaging
activity).  The polycopy of the 8V40 inserts of the infected
cells is a suggested model for replicon amplification induced by
carcinogens.  The authors suggest this amplification may be a
molecular mechanism for carcinogenesis initiation.       :
     Acrylamide was tested for its ability to alter the
transfection of E.  coli CR63 cells with E.  coli B lysate of

                              7-21

-------
colitis bacteriophages (Vasavada and Padayatty, 1981).
Transfaction by phage DNA is inhibited by alteration(s) to the
transfecting DNA (e.g., mutation, DNA cleavage).  Aery1amide-
treated phage DNA (10 ug aery1amide in 0.1 ml H20) produced a 50%
inhibition of trahsfection ability as compared to non-treated
phage control.  Aery1amide does not appear to interfere with the
transfection process as transfection of E. coli CR63 cells pre-
treated with acrylamide was not reduced when transfeeted with
non-treated phage DNA.  The authors suggest acrylamide is
interacting with phage DNA and producing alterations that change
the transfection efficiency.
     Evidence presented in this section on other genotoxic
effects suggests that acrylamide is capable of interacting with
DNA and producing a biological consequence.  The transformation
results and possibly the amplification of SV40 inserts suggests
acrylamide produces effects that may have implications for its
carcinogenic potential.
                              7-22

-------
                     8.   CARCINOGENIC EFFECTS
8.1. Summary and Conclusions
     Administration of acrylamide in the drinking water to F344
rats for 2 years caused a statistically significant increase in
the incidence of tumors (benign and/or malignant) in dosed
animals at the following tumor sites:  CNS uterus, mammary and
thyroid gland (females); scrotum, thyroid, and adrenal gland
(males).  Furthermore, in a series of one year limited bioassays
in mice, acrylamide was shown to:  (1) be a skin tumor initiator
in SENCAR mice (gavage, i.p., and dermal) and Swiss ICR mice
(gavage); and (2) induce lung adenomas in strain A/J mice (oral
                      /•
and i.p. routes), Swiss ICR mice (gavage, and SENCAR mice (i.p.).
     Additional support for the conclusion that acrylamide is a
               »
carcinogenic agent is provided by acrylamide's genotoxicity.
Acrylamide has been shown to be a clastogenic agent both in vivo
and in vitro.  Preliminary reports suggest that acrylamide binds
to DNA, induces DNA damage and repair effects, and causes in
vitro cell transformation.  Acrylamide also has been shown to
cause dominant lethal effects.
     OTS has concluded that there is a "sufficient" weight of
evidence from long-term studies in animals and from additional
supporting studies to identify acrylamide as a "Group B2"
carcinogen — probable carcinogen in humans, as defined in EPA's
Chemical Carcinogen Risk Assessment Guidelines.  Although the
chemical has only been studied in one long-term bioassay in rats,
it induced-statistically significant increased incidences of

                               8-1

-------
tumors in multiple tissues in both sexes.  Further, several of
the tumor incidences were dose-dependent.  Additional evidence
for the conclusion is provided by positive data from a series of
limited bioassays in mice, from a series of genotoxicity studies,
and from in vivo and in vitro DMA binding studies.
     The International Agency for Research on Cancer (IARC) has
reviewed the evidence regarding the carcinogenic potential of
acrylamide, and has concluded that acrylamide is a Group 2B
carcinogen—sufficient evidence for carcinogenicity in animals,
with no data or inadequate data on humans available (IARC, 1986).

8.2. Epidemiology Studies
     American Cyanamid studied the relationship of worker
exposure to acrylamide and cancer mortality at its Warners Plant
(Collins, 1984).  The investigation by American Cyanamid yas
based on analyses of two study groups:  a long duration exposure
cohort (10 individuals) and a short duration/intermittent
exposure cohort (52 individuals).  A standardized proportionate
mortality ratio (SPMR) was utilized to analyze the data because
information was not available for the entire population at risk.
Results from the study indicated no significant excesses of all
types of cancer mortality had occurred among the exposed
employees.  Although mortality from cancers of the lung and
central nervous system appeared to be slightly elevated, these
findings were based on small numbers of cases and the SPMRs were
not significantly different from expected values.  Many factors

                               8-2

-------
limited the interpretation of the study.  These limitations
included:  underrepresentation of the employee population
potentially at risk for exposure-related effects, incomplete
ascertainment of causes of death for cohort members,  incomplete
acrylamide exposure data, and the small size of the cohort
(Pickrel et al., 1986).
     Because of the limitations noted, the Warners Plant cohort
study does not provide an adequate assessment of the mortality
experience of workers exposed to acrylamide.  The investigation
by Collins (1984), admittedly, was limited only to information
abstracted from available company records.  As a result of this
                     /
restriction, the cohort may have been incomplete and biased with
respect to employee acrylamide exposure and cancer mortality.
               »
The analysis of 'two separate exposure cohorts may have also
precluded a proper interpretation of workers* cancer mortality
patterns.  Analyses of combined data for two cohorts indicated
that significant excesses of lung cancer mortality may have
occurred (Obs/Exp = 8/4.64, SPMR = 1.72, p = 0.6).  However, the
relatively small cohort size and low level of statistical power
associated with the study precluded further interpretation of
these results.
     In a study by Sobel et al. (1986), the mortality experience
of 371 employees assigned to acrylamide monomer and
polyacrylamide operations was examined.  Emphasis was given to
those cancer sites identified from animal studies (see Johnson et
al. , 1986)-; i.e., tumors of the central nervous system, thyroid

                               8-3

-------
gland, other endocrine glands, and mesotheliom'as.  Whereas 38
deaths were expected, a total of 29 deaths was observed up until
1982.  No deaths were classified for the sites identified from
the animal studies.  However, 11 deaths due .to malignancies were
noted versus 7.9 expected (an SMR of 139). When those employees
with previous dye exposure were excluded, the obs/exp ratio
changed to 4/6.5 for an SMR of 61.   In Collins (1984) a slight
excess of lung and central nervous system deaths was observed.
While an excess of deaths from respiratory system cancers was
seen in Sobel et al., obs/exp, 4/1.9, SMR 202, a deficit resulted
when employees with previous dye exposure were removed (obs/exp,
1/2.4).
     American Cyanamid followed up this work  with a more detailed
               »
historical prospective cohort mortality investigation, similar to
that of Sobel et al. (1986).  It was a more involved analytic
follow-up to the small proportionate mortality ratio analysis of
Collins (1984).  The study was performed at three U.S. plant
locations - Warners, Fortier, and Kalamazoo - and at the Botlek
facility in the Netherlands.  Employees were  engaged in different
jobs with potential exposure to either or both production of
acrylamide monomer or its polymerization into polyacrylamide.
     The maximum number of years for the worker(s) with the
longest duration of exposure contrasts with the "average" number
of years of duration at the plants.  These figures, along with
the estimates of numbers of departments, job  descriptions and
exposed workers are listed below in Table 2.   The table helps to

                               8-4

-------
show that the great majority of person-years of exposure occurred
at the Warners plant, where the average duration of exposure is
higher only.-than that at the Botiek plant.  It also shows that,
although all the plants produced acrylamide,. the industrial
hygienists have categorized many more jobs as "exposed" at
Warners.  It is possible some of these jobs had minimal exposure,
which, if true, would dilute the comparisons.
Table 2.  Estimates of Exposed Jobs, Titles, Workers, and
          Years of Exposure
                          Job
Plant
Warners
Fortier
Kalamazoo
Botiek
Departments Titles Workers
9
7
5
' 9
70
8
11
31
1355
634
37
267
Maximum
30
18
15
18
Average
6.
7.
10.
5.
3
2
3
7
     There was a statistically significant excess of total cancer
for the combined Warners cohort (448 obs / 412.6 exp / SMR 109/ p
= 0.04), a difference of 35.4 cases.  No information was given
that might have contributed to understanding this excess.  When
the entire cohort was separated into exposed and unexposed
subcohorts, there was an excess of 35.6 total cancer cases in the
unexposed group.  It may be due to another occupational exposure.
     Twenty-five cancer sites were examined, but because of the
low number of expected deaths this much analysis need not have
been performed.  Just considering the SMRs among the exposed
subcohorts"combined from the four plants, excesses appeared for

                               8-5

-------
cancer of the pancreas (8 obs / 4 exp / SMR 200) ,  respiratory
tract (30 obs / 26.3 exp / SMR 114), and lung cancer (30 obs /
25.1 exp /SMR 120 ); none were statistically significant (p >
0.15).  There were no notable deficits in observed deaths.
     Limitations of this study included:   the inexplicable
inclusion of 1,533 exposed workers in the "unexposed" group;
inadequate exposure data concerning acrylamide,  and no data
concerning other exposures; no specification of  the standard
population used to compute expected numbers of deaths in the
trend analysis, and no analysis of time from initial exposure to
death from cancer.
                     /
     It is difficult to accept this report as evidence of no
association between acrylamide exposure and cancer excess without
               »
further information clarifying the apparent inclusion of 1,533
exposed workers in both the unexposed and exposed  groups."*
Especially troubling is the idea that deaths among these exposed
workers may have been tabulated in the unexposed totals.
     The expected rates in the trend analysis for  the unexposed
are not the same as those the authors used earlier in the study,
apparently because a different, internal, standard was used.
Because the comparison population has not been specified, the
validity of this analysis cannot be ascertained.  If these
"unexposed" workers - who show a statistically significant excess
of respiratory cancer, compared with U.S. white  male rates  (169
obs / 128.6 exp / SMR 1.31 / p = 0.0004)  - have  been exposed to
                               8-6

-------
other potentially carcinogenic agents, this internal comparison
would not be a suitable one.
     Acrylamide exposures at the four plants began as early as
1954, but no monitoring was conducted until 1977.  The Warners
New Products department was never monitored.  Thus, the
approaches to estimating exposures do not constitute an exposure
assessment.
     Overall, the results of these three studies do not indicate
a statistically significant excess of cancers from exposure to
acrylamide, although all three studies showed apparent increases
in lung or respiratory cancers among acrylamide-exposed workers,
                      /
and lung cancer incidence was increased in the mouse oncogenicity
work (Bull et al., 1984a,b).  Small sample sizes in Collins
               »
(1984) and Sober et al. (1986), the ambiguous definitions of the
study groups in Collins et al. (1987) , and the incomplete^
exposure assessments in all three studies were insufficient",
however, to accurately assess any causal role for acrylamide
related to cancer incidence.  Consequently, all of the studies
discussed here are inadequate to judge the human carcinogenicity
of acrylamide.

8.3.  Two-Year Bioassav Data
     A chronic bioassay on acrylamide has been performed by Dow
Chemical Company (Johnson et al., 1984, 1986).
                               8-7

-------
8.3.1.  Summary of Protocol/Conduct of Study
     Several dose levels were administered via the drinking water
to both sexes of F344 rats:  0.0, 0.01, 0.1, 0.5, and 2.0 mg/kg
body weight/day for two years.  The purity-of the test material
ranged from 98.1% upwards, with soluble polymer (230 to 600 ppm)
being the major contaminant.   The concentration of acrylamide
after four days in the drinking water in the cage bottles was at
least 92% of the original concentration, therefore drinking water
solutions were prepared twice a week.  The amount of acrylamide
in the drinking water was within acceptable variance of the
targeted level of all dose levels as determined by high
                     /
performance liquid chromatography (HPLC) on a Ci8 column, and the
dosing method is considered to have been adequate.  Ninety
               *•
rats/sex/dose were started on the study, but only 60
rats/sex/dose were part of the group designated for the 2-year
terminal sacrifice.   The remainder of the rats were used fo'r
either interim sacrifices at 6, 12, or 18 months, or a separate
electron microscopy assessment of the nervous system. (See
neurotoxicity section.)
     Parameters evaluated were mortality, clinical signs of
toxicity, body weight, food consumption, water consumption,
clinical chemistry, hematology, urinalysis, organ weight, gross
pathology, and histopathology.
8.3.2. Limitations of Study
     No major design or conduct limitations are apparent: in the
Dow study." There were transient symptoms in some rats consistent

                               8-8

-------
with a viral infection  (sialodacryoadenitis virus) that occurred
beginning on day 210 of the study.  All animal groups (both males
and females, including controls) were equally affected.  The
symptoms are judged not to significantly cenfound the results of
the carcinogenicity study.  The viral infection has been shown
not to significantly affect body weight, survival, and incidences
of adrenal tumors, mammary tumors and leukemia of F344 rats (Rao
et al. 1988).

8.3.3.  Reported Results
     The MTD appears to have been achieved at the high-dose level
                     t
based on decreased survival, decreased body weight gain, and the
observance of several toxic effects at the high-dose.  Mean body
weights of high-dose rats were less than controls - male rats at
the high-dose reached a maximum difference of about 4% le«s than
controls after a year or more on test.  Males given 0.5 mg/kg/day
and females given 2.0 mg/kg/day had mean body weights about 2%
less than their respective control groups.  Increased mortality
was observed in high-dose rats, especially females.  The increase
occurred only beginning at about the 21st month when mortality in
males at the high-dose was 6/60 compared to 3/60 in controls, and
mortality in females was 11/60 compared to 4/60 in controls.  At
termination, both males and females in the high-dose group showed
statistically significant increases in mortality:  males 25/60
and females 32/60 compared to male and female control mortality
of 16/60 and 10/60, respectively.  Also histopathology revealed

                               8-9

-------
that female high-dose rats showed degeneration of the peripheral
nerve (tibial) and some degeneration of the spinal cord;
recognized neurotoxic effects of acrylamide.
     According to the study authors, increased, incidences of a
variety of tumors (benign and/or malignant) were statistically
significant (p<0.05) at the high-dose, including; females -
mammary gland (benign and malignant),  CNS (malignant),  thyroid
gland - follicular epithelium (combined benign and malignant) ,
mouth (benign), uterus (malignant),  clitoral gland (benign) and
pituitary gland (benign); males-scrotal mesothelioma (malignant),
thyroid gland follicular epithelium (benign), and adrenal gland-
                     /
pheochromocytoma (benign).  An EPA analysis of selected tumor
sites revealed that the incidence of benign pituitary gland
               »
tumors (females)' was not statistically significant.  Note that
both treated males and females showed a statistically significant
increased incidence of tumors derived from the follicular
epithelium of the thyroid gland.  Tumor incidence data for the
above-mentioned sites are summarized in Table 8.1.  Pooled tumor
incidence data are summarized in Table 8.2.

     a.  Testes Scrotal Cavity
     The only tumor type that was significantly increased  in
males at a dose below the high-dose was scrotal mesothelioma in
rats given 0.5 mg/kg/day.  The incidence of scrotal mesothelioma
does not appear to be significantly increased in rats given 0.01
or 0.1 mg/lcg/day, although the incidence of scrotal mesothelioma

                               8-10

-------
                         Table  8.1. Tumor  Incidence Data from the Lifetime Drinking Water Bioassay
                                                 Males
Tumor Site/Type
 Number with tumor/number at risk /
 dose6 (statistical  significance)

Controls   0.01  0.1    0.5        2.0
                                                              Females
1) Testes-Scrotal Cavity
    Mesothelioma

2) Thyroid Gland
      Adenoma  (AO
      Adenocarcinoma(AC)
      A,AC, or Follicular
      Tumor

3) Mammary Gland
      Adenocarcinoma
      Total Benign Adenoma
      Combined Benign
      and Malignant

A) CNS
      Combined Glial Tumors
      or "proliferations"*
      Glial tumors (without
      "proliferations")

5) Adrenal Gland Pheochromocytoma
      Benign
      Malignant
      Combined     .
  3/57
  1/57
  5/60
  5/60
  3/57
  2/57
  5/57
                                       Number with tumor/number
                                 at riskd/dosee (statistical ,significance)
                                                                               Controls  0.01   0.1
0/50  7/57  ll/53(a)  10/5A(a,b)
0/53  2/57   1/53
2/60  0/60   3/60
2/60  0/60   2/60
8/50  7/57   5/52
0/50  2/57   1/52
8/50  9/57  /6/52
 7/5A(a,b)
0/5A    0/55   1/50   1/5A
1/5A    0/55   0/50   0/5A
1/5A    0/55   1/50   1/5A
 8/60
 6/60
10/5A(a,b)
 0/5A
10/5A(NS)
3/50(b]
3/50
5/50(bi
                                              2/58     1/58   1/52   2/55       6/57(c;
                                              10/60    11/60   9/60   19/58      23/60(a
                                              10/60    12/60  10/60   20/58(b)   28/60(g
1/60    2/59   1/60  '1/60       9/60(a
1/60    2/59   1/60   0/60       7/60(g

-------
                                                    Table 8.1. Continued

                                                    Males	
Tumor Site/type
Number with tumor/number at risk /
dose6 (statistical significance)
                                                             Females
                                   Controls   0.01  Q.I
                       (Li
      Number with tumor/number
at riskd/dosee (statistical :significance)
Controls  0.01   0.1    Q..5
6)



7)
8)



9)



a
b
c
d
e
f
g
NS
Clitbral Gland
*,!-;_
Adenoma 	 	
Adenomacarcinoma 	 	 	
Combined
Uterus Adendcarcinoma 	 	 	
Oral Cavity
Squamous Papilloma 4/57 7/50 0/57
Squamous Carcinoma 2/59 0/53 1/58
Combined — — 	 	
Pituitary Gland 	 	 	
Adenoma 	 	 	
Adenocarcinoma 	 	 	
Combined

__ ____ n/i
""••"• —— — \J 1 £t
	 	 0/2
0/2
— — _*__«. 1 / 5o

5/53 ' 4/54(NS) 0/58
0/58 2/56(NS) 0/60
	 0/60
	 _____ 25/59
_. 	 ••..T— 7 / C Q
/ / J 7
	 	 32/59


1/3 3/4
0/3 0/4
1/3 3/4
2/56 1/51

3/58 2/52
0/60 0/60
3/60 2/60
30/60 32/60
9/60 1/60
39/60 33/60


2/4
2/4
4/4
0/55

1/56
2/60
3/60
.27/60
3/60
29/60


5/5(a,b)
0/5
5/5
5/49(a)

7/56(a,b
1/60
8/60(a,b
32/60(a)(!
5/60
37/60(NS)

» Dow designated as statistically significant, Fisher Exact probability test, =0.05
= Dow designated as statistically significant, mortality adjustment via Mantel-Haenszel procedure
- Dow designated as linear trend by Mantel-Haenszel
• number at risk is the number of rats assumed to be
» dose in mg/kg/day in drinking water ,
= includes glial proliferations that are suggestive
= PiO.05 by Fisher Exact probability test
= Not significant by Fisher Exact probability test
(Deto)
-0.05
extension of Cochran-Armitage test =0.05
alive at time of appearance of

of an early tumor


the first tumor

-------
               Table 8-2.   Pooled Tumor Incidence Data
                                          Dose (ma/ka/dav)
Males                   0.0

Number of Animals       7/57
with Tumors8
(testes, thyroid
adrenalb)

Number of Animals       3/57
with Malignant
Tumors {testes)

Females

Number of Animals    /  13/60
with Tumors3
(thyroid, mammary,
CNSC,  oral,  uterus)
               »
Number of Animals       4/60
with Malignant
Tumors (thyroid,
mammary, CNS,
uterus)
 0.01     0.1

 8/53    13/57    14/53     22/54
 0/53
7/57    11/53
                  10/54
18/60    14/60    21/60
 5/60
3/60
         4/60
                  46/60
20/60
     Benign or malignant.  Site must be statistically significant at
     the high-dose (treated vs. control) for tumors to be
     considered.
     includes only adrenal benign tumors
     tumors only (no "proliferations")
                               8-13

-------
exceeded concurrent and historical control mean values in male
rats given 0.1 mg/kg/day.  Note that although concurrent controls
in the Dow study had an incidence of scrotal mesothelioma of
5.3%, Dow states in its report that the historical incidence is
3.8% for this tumor type in control males at 24 months,  and the
NTP lists mesothelioma as 1.3%, and 1.0%, for the testes and
peritoneal cavity, respectively, for F344 rats.  The percentage
incidence of scrotal mesothelioma in the present Dow study was
(dose in mg/kg/day):  5.3% (0.0), 0% (0.01), 12.3% (0.1), 20.8%
(0.5) , and 18.5% (2.0).
     b.    Thyroid
     The only tumors for which both sexes showed a statistically
significantly increased incidence were those derived from the
follicular epithelium of the thyroid gland.   Both males  and
females showed a statistically significant increase in tfc.e
incidence of thyroid adenoma at the high-dose.   Further,
adenocarcinomas were found in three females in the high-dose
group.

     c.   Mammary Gland
     The incidence of benign adenoma is statistically significant
in females at the high-dose, and the increased incidence of
combined adenoma and adenocarcinoma is significant at both the
high-dose (P=0.00047) and the next lowest dose level (P=0.022).
There appears to be a time-to-tumor (latency) effect for both
adenocarcinoma and total benign tumors of the mammary gland; this
possible latency effect was assessed for adenocarcinoma from the
                               8-14

-------
individual animal data EPA received.  Although the occurrence of
adenocarcinoma was not (according to Dow) statistically
significant by either pairwise analyses used by Dow, 5 of the 6
adenocarcinomas seen in high-dose females were observed before
termination whereas the 2 adenocarcinomas in control females were
not observed until termination of the bioassay (day 743).  The
five high-dose females with adenocarcinoma were evaluated on day
577 (spontaneous death), 593 (moribund),  677 (moribund),  680
(spontaneous death), and 732 (moribund).   The cause of death of 2
of these 5 females was attributed by Dow to the mammary tumors, 1
to pituitary tumor compression of the brain, and undetermined
cause of death for the other 2 females.  This apparent decreased
tumor latency supports the statistically significant trend seen
for adenocarcinoma.
                                                        »v
d.   Clitoral Gland
     There appears to be a dose-dependent response in combined
benign and malignant tumors of the clitoral gland, and there is a
statistically significant increase in adenomas (benign) and in
combined benign and malignant tumors in females at the high-
dose.   Due to the low number of clitoral glands examined (see
Table 8.1), this tumor site was not included in the pooled tumor
incidence data in Table 8.2.  However, it is important to note
that even with such small numbers of tissues examined, a rather
high incidence of tumors was observed.  Historically, clitoral
adenoma occurs in F344 female rats of the NTP testing program at
an average rate of 1.2% (Haseman et al.,  1984), and Dow's
                               8-15

-------
historical control incidence of clitoral adenomas in F344 rats is
listed in their report as 0.5%.

     e.   Oral Cavity
     Tumors originating from the mucosa of the mouth are shown in
Table 8.3.  Females show a statistically significant increase in
papillomas, and combined papillomas and carcinomas,  at the high-
dose when tumors of the "tongue" are combined with those of "oral
tissues".  The Dow historical control incidence in female F344
rats for squamous cell papilloma of the oral cavity is 2.2%
(range 0-4%).  Additionally, according to the Dow report, an
increased incidence of focal cell hyperplasia of the hard palate
was observed in both males and females, and the incidence of this
               ^
hyperplasia was -statistically significantly increased in  males
at both 0.5 and 2.0 mg/kg/day.

     f.  Uterus
     Adenocarcinoma of the uterus was significantly increased in
high-dose females compared to controls.  Endometrial stromal
palyps were also reported in this study; however, their
occurrence may not be related to acrylamide exposure because they
are commonly found in 344 rats.

     g.   Nervous System
     Table 8.4 shows the combined incidence of glial tumors
(malignant} or "proliferations" in the CNS.  Glial proliferations
are not considered by Dow to be tumors, but rather to be
                               8-16

-------
                              Table 8-3. Tumors Originating from the
                                       Mucosa of the Moutha
                                                                Dose (mg/kg/davl
Female Rats

Squamous papillomas, benign, origin
  from tongue, hard palate or lip

Squamous cell carcinoma, malignant,
  origin from hard palate or gingiva

Combined papilloma and carcinoma
JLP. -  0.01

0/58   3/58    2/52   1/56   7/56(b,c)


0/60   1/60    0/60   2/60   1/60


0/60   3/60    2/60   3/60   8/60(b,c)
Male Rats

Squamous papillomas, benign, origin
  from tongue, hard palate or lip

Squamous cell carcinoma, malignant,
  origin from tongue, hard palate,
  gingiva or pharynx

Combined papilloma and carcinoma
4/57   7/50   0/57    5/53    4/54
2/59   0/53    1/58    0/58    2/56
6/59   7/53    1/58    5/58    6/56
a =  Papillomas  and  carcinomas  occurred in separate rats,  according to the Dow. authors

b =  Dow designated  as  statistically significant,  Fishers' exact probability test,      =0.05

c =  Dow designated  as  statistically significant,  mortality adjustment via
     Mantel - Haenszel  procedure  (Peto),      = 0.05
                                              8-17

-------
                Table 8-4.  Combined Incidence of Glial Tumors of the CNS.
HALES
Brain

Astrocytoma
Glial Proliferation
  (suggestive of early tumor)
Oligodendroglioma

            (Cervical. Thoracic,
          Dose (mg/kq/day)
Astrocytoma
Undifferentiated glial cell tumor
Glial proliferation

Total rats with a tumor of
  glial origin

Total rats with a tumor of glial
  origin or a glial proliferation
  suggestive of early tumor
0.0

3/60
0/57

0/60
1/59
1/44
0/44

5/60
5/60
0.01

0/60
0/50

2/60
         0/53
         0/47
         0/47

         2/60
         2/60
                   0/60
                   0/57
0^5.

2/60
1/53
                   0/60     0/60
2/60
1/54

1/60
          0/58    0/58     3/56
          0/46    0/44     0/35
          0/46    0/44     1/35

          0/60    2/60     6/60
          0/60    3/60     8/60
FEMALES
Brain

Astrocytoma
Glial proliferation
  (suggestive of tumor)
Oligodendroglioma
0/58
0/56

0/56
Spinal Cord (CervicalT Thoracic, and Lumbosacral)

Astrocytoma                                      1/60
  Total rats with a tumor of
    glial' origin

  Total rats with a tumor of glial
    origin or a glial proliferation
    suggestive of early tumor
1/60


1/60
          1/58
          0/56

          1/56
         0/59

         2/59


         2/59
          0/52
          0/51

          1/51
          0/60

          1/60


          1/60
0/56
1/55

0/55
0/60

0/60


1/60
3/itf
3,'

1/49



3/60

7/60c


9a/60b
        one female rat given  2.0 mg/kg/day had an astrocytoma  in the cervical section of
        the spinal cord and glial  proliferation  in the brain.

        Statistically significant  increase over  control according to Dow (no details on
        level of significance).

        P10.032.
                                           8-18

-------
suggestive of early astrocytomas.  However, one consulting
veterinary pathologist for Dow considered these lesions to be
astrocytomas.  When brain and spinal cord are considered
separately, these individual categories do-not appear to be
statistically significant for either sex.  However, when
combined, the high-dose females have an increased incidence of
total glial tumors and proliferations.  Note the glial
proliferations in females only occurred in the two highest dose
groups.  Combined CNS glial malignant tumors (omitting glial
proliferations) are also statistically significantly increased in
females at the high-dose (P=0.032).  Additionally, other
                     A
combinations of glial malignant tumors were cited by Dow to be
statistically significant (i.e., astrocytomas from brain and
               »
cord, or astrocytomas and glial proliferations from brain and
cord) for females.  However, neither the incidence data or the
results of the statistical analyses were presented.  In males,
the increased number of CNS tumors or proliferations at the high-
dose was not statistically significant.  However, it should be
noted that the concurrent control incidence (8.3%) was greater
than either Dow's historical controls  (1.0% or NTP's historical
controls (
-------
altered cells in the adrenal cortex was observed at the 0.1 and
2.0 dose levels, although an increasing dose-response trend was
not observed.  This foci increase may indicate a preneoplastic
condition and as such supports the contention that this tumor
type may be chemically induced.  Note that neither malignant, nor
combined (p=0.11) tumor incidence was statistically significantly
elevated in males.

     i.    Pituitary Gland
     According to the Dow report, a statistically significant
increase in the incidence of benign tumors of the pituitary gland
                     /
was observed in high-dose females, but this difference was not
significant according to a statistical analysis (Fisher Exact
               »
test) performed'by EED/EPA.
     The incidence of pituitary adenoma,  and adenocarcinOma, in
Dow concurrent controls was 42.4%, and 11.9%, respectively; in
Dow historical controls 35.7%, and 12.4%; and in NTP historical
controls 44.0%, and 3.5%, for these tumor types.  Thus, the
incidence of adenomas in Dow concurrent controls (42.4%) is
somewhat higher than in Dow historical controls (35.7%).  Also,
it should be noted that the NTP historical control incidence for
adenocarcinoma  (3.5%) is much less than the Dow concurrent
controls (11.9%) and Dow historical controls (12.4%).  The high
incidence of pituitary tumors in the Dow concurrent controls may
be due to the influence of the viral infection which occurred
beginning "on day 210 of the study.  Rao et al.  (1988) have shown

                               8-20
-------
that male F344 rats with sialodacryoadenitis virus infection have
a significantly higher incidence of pituitary tumors.  Additional
support for the biological significance of the pituitary adenomas
is the fact that adenomas in the high-dose- females were generally
noted earlier, and were larger, than those in the control group.
However, the treated animals do not show an increase in
adenocarcinoma, and this is of course a factor in the combined
tumor incidence, which is also not statistically significant.
     In summary, there is some weak support for the biological
significance of pituitary tumors in high-dose females.  However,
a statistical analysis of pituitary tumors by EED/EPA has
                     /
revealed a lack of statistical significance for pituitary tumor
incidence in females (P<0.16 for adenomas), thus pituitary tumors
               »
are not included in the pooled tumor incidence in Table 8.2.
                                                        ^
8.4.  Second Lifetime Oncogenicity Study in Rats with Acrylamide
     — (American Cyanamid Company, Final Report, June 27, 1989)
     A second lifetime oncogenicity study in rats with acrylamide
was designed to clarify some ambiguities (considered by American
Cyanamid Company) of the prior carcinogenicity study by Johnson
et al. , (Toxicol. Appl. Pharmacol. £11:154-168, 1984).

8.4.1.  Background and Summary of Results
     Statistically significantly increased incidences of mammary
gland tumors (fibroadenomas, or fibroadenomas and adenocarcinomas
combined), scrotal mesotheliomas, and thyroid neoplasms  (adenoma,
or adenoma and carcinoma combined, in both males and females)
                               8-21
-------
were observed in Fischer 344 rats administered acrylamide in the
drinking water.  Increased incidences of uterine adenocarcinomas,
clitoral gland adenomas, and papillomas of the oral cavity
observed in the prior study were not found- in the rats of this
study.  However, there was a slight increase in cutaneous
fibromas in the female high-dose group of this study.
Furthermore, there was a dose-related positive trend in the
incidence of malignant reticulosis of the brain in the dosed
female rats and the incidence of CNS glial tumors (astrocytomas)
                                                         5
was higher in the high-dosed male and female rats than in
controls.  The totality of findings in this study confirm the
multiple-sites carcinogenicity of acrylamide in the rat.

8.4.2.  Summary'of Protocol/Conduct of Study
     Groups of male and female Fischer 344 rats (6-7 weefcs of
age) were administered acrylamide (99.9% pure) in their drinking
water for 2 years (106-108 weeks) at doses of:  0.1, 0.5 or 2.0
mg/kg/day (male); 1.0 and 3.0 mg/kg/day (female).  The body
weight range of male rats at the start of the study was 72-144 g
while that of the female rats was 75-122 g.   In the male rat
study, two independent control groups of 102 male rats in each
group were used; the low-, mid-, and high-dose groups consisted
of 204, 102 and 75 animals, respectively.  In the female rat
study, two independent control groups were also used consisting
of 50 female rats in each group; the low- and high-dose groups
each consisted of 100 animals.

                               8-22
-------
     Mortality, clinical signs of toxicity, body weight, food
consumption, and water consumption were measured during the
study.  A complete necropsy was done on al-1 animals sacrificed at
the termination of the study and on those animals which were
sacrificed in a moribund condition or that died during the study.
At terminal sacrifice, weights of brain, liver, kidneys and
testes were obtained.  Microscopically, the thyroid and the
testes were examined in all male rats.  In addition, the brain,
spinal cord and gross lesions were examined in all control and
high-dose male rats and in all low- and mid-dose rats found dead
                     /
or sacrificed moribund.
               »
8.4.3.  Limitations of Study
     No major design or conduct limitations are apparent^in this
study.  The MTD appears to have been achieved at the high-dose
level based on decreased body weight gain, decreased survival,
and the observance of several toxic effects at the high-dose.

8.'4.4.  Reported Results
     The mean body weight of the high-dose (2 mg/kg/day) male
group was consistently lower than that of each individual control
group or the combined control groups with maximum decrease (8.6%)
at the end of the study.  There were no differences in the mean
body weight between the other treated male groups and the control
groups.  The mean body weight of the high-dose (3 mg/kg/day)

                               8-23
-------
female rats was significantly decreased compared to controls as
early as week 3 of the study and reached a maximum decrease of
9.2% at week 92.  The mean body weight of the low dose
(1 mg/kg/day) female rats was decreased from that of the controls
early in the study but from week 32 until the end of the study
differences from controls were insignificant.
     There was increased mortality in the high-dose male group
(2 mg/kg/day) beginning from month 17 and continuing until the
end of the study.   At the end of the study,  mortality was 75% in
this group compared to 58.5% in the control  groups.   In the
females, increased mortality was observed in the high-dose group
during the last month of the study.  Total mortality reached 49%
at the end of the study compared-to 34% mortality in the control
groups.  There was no increased mortality in the low- or mid-
dose male or the low-dose female groups.   At terminal sacrifice,
there were slight differences in organ weight which were
occasionally statistically significant.  There were increased
incidences of the palpable masses and mild degeneration of the
peripheral nerve in the high-dose group of both sexes.
     Under the conditions of this study,  there were increased
incidences of CNS glial tumors (astrocytoma) and of thyroid
follicular cell neoplasms of both males and females at the high-
dose of 2 and 3 mg/kg/day, respectively.   In addition, there was
a significant increased incidence of scrotal mesotheliomas at the
high-dose males and of mammary gland neoplasms in both low- and
high-dose females (1 and 3 mg/kg/day) when compared to that of

                               8-24
-------
the control groups.  Although not statistically significant,
there was a dose-related positive trend in the incidence of
malignant reticulosis of the brain in the dosed female rats and
there was a slight increase in cutaneous fibrqmas in the female
high-dose group.  Increased incidences of uterine
adenocarcinomas, clitoral gland adenomas, and papillomas of the
oral cavity observed in the prior study were not found in this
study.  The tumor incidence data for each organ site are
summarized in Tables 1-4.

     a.  Testes Scrotal Cavity
                     /
     The incidence (13/75, 17.3%) of mesotheliomas of the testes
was statistically significantly~(P<0.001) increased in the high-
               »
dose (2.0 mg/kg/day) male group in comparison to the control
group (4/102, 3.9%).  Furthermore, there was a slight, btft not
statistically significant, increased incidence (8/102, 7.8%) in
the mid-dose group (0.5 mg/kg/day) (Table 1).  Of the 13 high-
dose rats diagnosed with scrotal mesothelioma, only 2 were
identified at the terminal sacrifice.  The other 11 were
identified in animals found dead or sacrificed moribund, with the
earliest dead occurring in week 66 of the study, and a total of 6
identified prior to week 91.  The short tumor latency is
indicative of the carcinogenic potency of acrylamide.
                               8-25
-------
     b.  Thyroid
     There was a slight increase in the incidences of follicular
cell carcinomas in both the high-dose male and female groups,
which were not statistically significant.  -However, both male and
female rats in the high-dose groups (2 and 3 mg/kg/day,
respectively), as well as females of the low-dose group
(1 mg/kg/day), had statistically significantly increased
incidences of thyroid follicular cell adenoma, or adenoma and
carcinoma combined.  Thyroid follicular cell adenomas were seen
in 2/100 (2%) and 1/102 (1%) controls, 9/203 (4.4%) low-dose,
4/101 (4%) mid-dose, and 15/75 (20%) high-dose male rats; the
                     /
incidences in the controls (two groups), low- and high-dose
females were 0/50 (0%), 0/50 (0%'), 7/100 (7%), and 16/100 (16%),
               *
respectively.  Combined thyroid follicular cell adenomas and
carcinomas were seen in 3/100 (3%) and 3/102 (3%) control,
12/203 (5.9%) low-dose, 4/102 (4%) mid-dose, and 17/75 (22.7%)
high-dose males;  the incidences in female rats were:  1/50 (2%)
and 1/50 (2%) controls, 10/100 (10%) low-dose, and 23 /100 (23%)
high-dose (Table 2).

     c.  Mammary Gland
     The incidences of mammary fibroadenomas, or combined
fibroadenomas and fibrocarcinomas were statistically increased in
both acrylamide treated female groups (1 and 3 mg/kg/day) as
compared with those in the control groups.  Mammary fibroadenomas
were seen in 5/46 (10%) and 4/50 (8%) controls, 20/94 (20%)  low-

                               8-26
-------
dose, and 26/95 (26%) high-dose female rats; the incidences of
fibroadenomas and fibrocarcinomas (combined) in the female rats
were: 7/46r (15%) and 4/50 (8%) controls, 21/94 (22%) low dose,
and 30/95 (32%) high dose.  There was no significant difference
in the incidence of hyperplasia or adenocarcinoma between the
treated rats and the controls (Table 3).

     d.  Brain and Nervous System
     There were no significant differences in the incidence of
combined brain and spinal cord (CNS) total glial tumors
(astrocytomas and oligodendrogliomas) between control and treated
                     /
rats, male or female.  However, the incidence of astrocytoma, a
CNS tumor of glial origin, was statistically significantly
               »
increased (P<0.01) in the male high-dose groups (3/75, or 4%) as
compared with that in the controls (1/102, or 1% and 0/102, or
0%); the incidence of astrocytomas in high-dose females was also
higher (3/100, or 3% vs. 0/100, or 0% in controls).  Furthermore,
an infrequently occurring non-glial tumor of the brain, malignant
reticulosis, was diagnosed in 2 low-dose and 3 high-dose female
rats; 1 control male and 1 low-dose male but no female control
rats had this neoplasm  (Table 4).  The numbers of brain and
spinal cord (CNS)  sections examined histopathologically in this
study were not specified; in the prior study, greater numbers of
sections of these tissues were examined (brain — 6 sites vs. 3
routinely; spinal cord — 3 sites vs. 1).
                               8-27
-------
     e.  Skin
     There was a slight increase in cutaneous mesenchymal
neoplasms (fibromas) in the female high-dose group (3 mg/kg/day)
with a total of 5 fibromas in the high-dose group versus 1 in the
combined control groups.
     A variety of cutaneous epithelial neoplasms were observed in
all groups of animals in this study.  The incidence of these
neoplasms was similar in all groups.

     f.  Other tissues
     In this study, there were no significant increased
                      /
incidences of tumors of the oral cavity,  clitoral gland, or
uterus in the dose groups.
               »
                \
8.5.  Limited Bioassavs (One-Year studies)               ^.
     Bull et al. (1984, a,b) showed that  acrylamide acted as a
skin tumor initiator in both SENCAR mice  and Swiss-ICR mice.
Acrylamide also increased lung adenoma incidence in SENCAR,-
Swiss-ICR, and A/J strains of mice (Bull  et al.  1984a, b;
Robinson et al 1986).
     Female SENCAR mice (6-8 weeks old, 40 mice per dose level)
were administered acrylamide at 0,12.5, 25.0 or 50.0
mg/kg/application of acrylamide for a total of 6 applications
over a 2-week period by gavage, i.p., and dermal routes.  Two
weeks after the final dose of acrylamide  was applied
(representing the initiation phase of the experiment), 1.0 ug of

                               8-28
-------
                                           TABLE 1

                              Incidence of Scrotal Vlesotheliomas

Group	          1                2               3                 4                 5
                                                         N»	      	 	
Dose
  (mg/kg/day)      0                0             0.1               0.5               2.0

No. Examined      102              102            20A               102                75

Animals with
Mesothelioma,
Testes tunic       4                49                 8                13*

Per cent          3.9              3.9            4.4               7.8               17.3


*Significantly different from combined controls, p<0.001 (Peto, 1980,  and               '
 Tarone, 1975)
-------
                                                        TABLE 2

                                      Incidence of Thyroid Follicular Cell Lesions
Males
Groups
Dose
(mg/kg/day)
No. in Group
No . Examined
Adenoma
%

Carcinoma
%
Hyperplasia
%

Combined
Neoplasms%
1
0
102
100
2
2

1
1
1
1

3
3
2
0
102
102
1
1

2
2
2
2

3
3
3
0.1
204
203
9
4.4

3
1.5
7
3.4

12
5.9
4
0.5
102
101
4
4

0
0
3
3

4
4
Females
5
2.0
75
75
15*
20*
1
3
4
2
2.7

17**
22.7**
1
0
50
50
0
0
•
1
2
0
0

1
2
2
0
50
50
0
0

1
2
1
2

1
2
•• 1 •
3
1.0
100
100
7*
7*

3
3
5
5
i
10*
•10*

4
3.0
100
100
16*
16*

7
7
1
1

23**
23**
*Fisher exact, P<0.05
*Statistically significant, p<0.001 (Peto, et al., 1980)
-------
                                   TABLE 3

                 Incidence of  Primary Mammary  Gland Neoplasms
Females
Group
Dose
( mg/kg/day )
No. in Group
No. Examined
Fibroadenoma
%
Adenocarc inoma
%
Combined
% 	
1

0
50
46
5
10
2
4
7
15
2

0
50
50
4
8
0
0
4
8
3

1.0
100,
94
20*
20*
, 2
2
21*
22*
4

3.0
100
95
26*
26*
4
4
30*
32*
*Fisher»s Exact, p<0.05
**Statistically significant, p<0.001, (Peto,  et al.,  1980)
-------
1

Groups
Dose
(mg/kg/day)

No . Examined
Astrocytdma
%
Oligodendro-
glioma
%
Combined
%
*Statisticallv
TABLE A
Incidence of Primary CNS Glial Tumors

1
-
0

102
1
	 	 1
0

0
1
1
sienificant.

2

0

102
0
0
1

1
1
1
p<0.01.
Males
3

0.1

98
1
1
1

1
2
2
Peto. et al...

A

0.5

50
0
0
1

2
1
2
(1980)
, »
5
Nfc
2.0

75
, 3*
A*
0

0
3
A


1

0
50
50
0
0

0
0
0
0


2

0
50
50
0
0

0
0
0
0

r
Females
3

1.0
100
100
2
2

0
0
.'' 2
2



A

3.0
100
100
3
3

0
0
3
3

-------
promoter, 2-0-tetradecanoyl-phorbol-13-acetate  (TPA), was applied
topically to the shaved back of each mouse 3 times per week for 20
weeks.  Animals were observed weekly for tumors, and sacrificed 52
weeks after the initial acrylamide administration.  By all three
routes of administration, a dose-response increase of squamous cell
carcinomas and papillomas was observed in SENCAR mice treated with
acrylamide and TPA (Table 8.5).  The incidence of squamous cell
papillomas or squamous cell carcinomas as a function of the route of
exposure was:  oral > i.p. > topical.  There was a highly
significant dose-response relationship for time to first tumor as
well as the appearance of multiple tumors by all three routes of
                        /
administration (P«0.01).  Acrylamide ppeared to be roughly as
potent as ethyl carbamate (CH2-CH2-0-C-NH2) ,  a structural analogue of
                 «
acrylamide in .initiating skin tumors.  However, acryiamide treatment
alone (without TPA) did not result in an increased incidence of skin
tumors in the SENCAR mice under the conditions of the test (Bull et
al. 1984a).
    In a combined skin papillona/lung adenoma assay, significant
increases in skin and lung tumor incidences were noted in SENCAR
mice administered 50 mg/kg of acrylamide by a single i.p. injection
followed by triweekly application of 1.0 ug TPA for 20 weeks
(Robinson et al.  1986).
    The tumorigenic activity of acrylamide in the skin and lung
  has  also been studied  on Swiss-ICR mice (Bull et al.  1984b).
  Acrylamide dissolved in water was administered by gavage in doses
                                8-33
-------
       Table 8.5.   TUMOR-BEARING ANIMALS AND CLASSIFICATION OF
                        HISTOLOGICALLY EXAMINED
SKIN TUMORS IN'SENCAR MICE INITIATED WITH ACRYLAMIDE BY VARIOUS ROUTES3
Cumulative No. No. of Squamous
1 of Tumor Bearing Cell Papillomas/
Total Dose Route of Animals6/No. of No. of Animals
(mg/kg) Administration TPA Animals Initiated Examined
_ »
Ethyl
Carbamate
300
Acrylamide
Oc
75
150
300

300
Oc
75
150
300
300
p.o. +

P.O. +
p.o. +
p.o. +
P.O. +

p.o. -
i.p. +
i.p. +
i.p. +
i.p. +
i.p.
8/20

2/40
12/40
23/40
30/40

0/20
0/40
10/40
13/40
21/40
0/20
0/16e

0/34
I
3/35
8/33
11/38

0/17
0/35
2/38
3/36
- ' 6/35
4
0/17
No. of Squamous % of Animals
Cell Carcinoma/No. Bearing Squamous
of Animals Examined Cell Carcinomas
3/16e

0/34
2/35
7/33
6/38

0/17
0/35
2/38
4/36
4/35
0/17
18.8

0
5.7
21.2
15.8
i
0
0
5.2
11.1
11.4
0
-------
                              Table 8.5.   (Cont.) TUMOR-BEARING ANIMALS AND CLASSIFICATION OF
                                                   HISTOLOGICALLY EXAMINED
                           SKIN TUMORS IN SENCAR MICE INITIATED WITH ACRYLAMIDE BY  VARIOUS  ROUTES3
Cumulative No.
of Tumor Bearing
Total Dose Route of Animals6/No. of
(mg/kg) Administration TPA Animals Initiated
No. of Squamous
Cell Papillomas/
No. of Animals
Examined
No. of Squamous
Cell Carcinoma/No.
of Animals Examined
% of Animals
Bearing Squamous
Cell Carcinomas
Acrylamide ^
Od
75
150
300
300
topical +
topical +
topical +
topical, . +
topical -
7/40
A/40
11/40
18/40
0/20
5/36
3/38
3/35
2/34'
0/20
0/36
1/38
2/35
3/34
0/20
0
2.6
5.7
8.8
0
Source:   Bull et al.; 1984a
^to be included in the cumulative count, the animal must have had a lesion.   1  mm in diameter,  for 3  consecutive weekly
 observations.
cDistilled deionized water was administered in the same volume (0.2ml/mouse) and frequency as used in .experimental groups
"Ethanol  was administerd topically in the same volume (0.2ml/mouse) and frequency as used for experimental  groups.
lumber of animals available for histological examination following death,  or at termination of the experiment  at  1  year;
 number excludes heavily autolyzed and cannabalized animals.
-------
of 12.5, 25, or 50 mg/kg 6 times over a 2-week period.  The total
dose administered was 75, 150, or 300 mg/kg to 40 female mice per
dose level. - Two weeks after dosing with acrylamide, 2.5 ug of TPA
was applied 3 times a week for a period of 20 weeks to the shaved
back of the mice.  The mice were observed weekly for the development
of skin tumors and after one year for the development of lung
tumors.  Acrylamide was found to initiate squamous cell adenomas and
carcinomas in the skin and to increase the yield of adenomas and
carcinomas in the lung.  (See Tables 8.6 and 8.7.)  Skin tumor
development was dependent on TPA promotion whereas lung tumor
induction was not.  As shown in Table 8.6, treatment with acrylamide
                       /
followed by TPA promotion resulted in a dose-related increase in
squamous cell carcinoma; at the high-dose level this increase in
                 *
carcinomas, and al'so squamous cell papillomas, was statistically
significant (P<0.05).  Only one carcinoma was observed in t?he
acrylamide-treated group that did  not receive TPA.  Note that ethyl
carbamate followed by TPA promotion was somewhat less potent than
acrylamide plus TPA in increasing the number of skin tumor-bearing
animals as well as the number of skin tumors per animal.  The yield
of alveolar bronchiolar adenomas and carcinomas (combined) was
increased in a dose-related manner (P<0.03) by acrylamide treatment,
as shown in Table 8.7. , and this increase was observed in the
absence or the present of TPA promotion.  As also shown in Table
8.7, ethyl carbamate treatment yielded a slightly higher combined
yield of alveolar bronchiolar adenomas and carcinomas than did
acrylamide treatment.

                                8-36
-------
                             Table 8.6.   TUMOR INITIATING ACTIVITY OF ACRYLAMIDE IN THE SKIN OF
                                                    FEMALE SWISS-ICR MICE3



Chemical
Distilled H^O
Acrylamide



Ethyl Carbamate


Doseb
(mg/kg)
0.2 ml
75
150
300
300
300

Cumulative No.
of Tumor Bearing
TPAC Animalsd
+ 0/40 (35)
+ 4/40 (36)
+ 4/40 (32)
+ 13/40 (32)
10/40E (36)
+ 4/40 (33)


Cumulative No.
Tumors /Animals
^ »
0
0.10^
0.13^
0.43
0.03
0.18
Histological
of Squamous
at
Papilloma
0
1
0
6
0
2
Classification6
Cell Tumors
Autopsy
Carcinoma
0
1
3
4
1
3
Source:   Bull et al.,  1984b

      was administered orally in six divided portions over a 2-week period to the total  indicated.
cTPA was applied at a dose 2.5 ug/mouse in acetone 3 times/week for 20 weeks beginning 2 weeks after the last
 dose of  initiator.                                                                                    <

4ro be included in the cumulative count a 1 mm mass had to be present in the same location for 3  consecutive
 weekly observations.  Numbers of animals begun in each group is indicated as the denominator of the ratio.
 Animals surving to terminaton of the experiment at 1 year included in parentheses.
     difference between total tumors observed at autopsy versus the cumulative yield of tumors  was  related to
 the regression of papillomas.

%his value is probably 1/40 since it was stated in the text of the publication that only 1  tumor was  found in this  group.
-------
                         Table 8.7.   LUNG TUMOR INCIDENCE IN SWISS ICR-MICE TREATED WITH ACRYLAMIDE3
Chemical
Distilled HjO
Acrylamide



Ethyl Carbamate
Doseb
(mg/kg) TPAC
0.2 ml +
75 +
150 +
300 +
300
300 +
Alveolar Bronchiolar
N Adenoma Carcinoma
36
3A
36
3A
36
36
3
-» 6
5
10
>A
9
1
2
1
1
10
8
Total
A
8
6
11
1A
17
Source:  Bull et al., 1984b

      was administered orally in 6 divided portions over a 2-week period to the total dose indicated.
°TPA was applied at a dose of 2.5 ug/mouse in acetone 3 times /week for 20 weeks beginning 2 weeks after the
 last dose of initiator.

^N - Number of animals available for autopsy without unacceptable levels of autblysis.  Each group initially contained AO
animals .
-------
     Groups of 40 strain A/J mice (8 weeks old) per sex per dose
level were administered acrylamide po at 6.25, 12.5, or 25.0 mg
of acrylamide per kg of body weight 3 times per week for 8 weeks
(Bull et al. , 1984a).  Ethyl carbamate was administered by the
same regimen at doses of 10.5, 21.0, and 42 mg/kg.  Mice were
sacrificed at 9 months of age.  Both acrylamide and ethyl
carbamate increased the yield of lung adenomas in each sex in a
dose-related manner; a dose-response relationship (P<0.01) was
observed for both animals with tumors and the multiplicity of
                »
tumors for each chemical.  The potency of ethyl carbamate was
seven times higher than acrylamide per unit of dose in increasing
                      /
the number of tumors per animal.  These data were presented in a
figure in Bull et al. (1984a), but are not presented in this
                »
review since the'data were not tabulated.
     Groups of 16 strain A/J mice (8 weeks old) per sex per dose
were administered acrylamide by ip at 1, 3, 10, 30, or 60 mg/kg
of acrylamide, 3 times per week for 8 weeks.  Positive control
groups received a singe ip injection of ethyl carbamate at 500 or
1000 mg/kg.  Animals were sacrificed at 8 months of age.  A dose-
response relationship was observed for the number of lung
adenomas (P«0.01) produced by acrylamide treatment, as shown in
Table 8.8. A higher dose of 60 mg acrylamide/kg was also
attempted, but discontinued after the llth injection due to frank
peripheral neuropathy and decreased survival.  Based on some
additional testing. Bull et al. (1984a) concluded that acrylamide
is slightly more potent in lung tumor production by the ip route

                               8-39
-------
Table 8.8,
EFFECTS OF ACRYLAMIDE ADMINISTERED INTRAPERITONEALLY
  ON DEVELOPMENT OF LUNG ADENOMAS IN A/J MICE*
Treatment
None

Distilled H20

Acrylamide
1 mg/kg/dayc

3 mg/kg/dayc

10 mg/kg/dayc

30 mg/kd/dayc

No. of Survivors/
Initial No. of
Sex Mice
M
F
M, F
M
F
M, F
M
F
M, F
M
F
M, F
M'
F
M, F
M
F
M, F
16/16
14/16
30/32
16/16
15/16
31/32
16/16
17/17
33/33
16/16
17/17
33/33"
17/17
14/15
31/32
15/16
15/16
30/32
Percent of
Mice with
L~ung -Tumors
31
50
40
13
8
10
50
35
42
38
53
46
59
79
68
93
93
93
Average No.
of Lung
Tumors /Mouse
0.31+0.48b
0.50+0.52
0.40+0.50
0.06+0.25
0.13±0.35
0.10+0.30
0.75+0.93
0.35±0.93
0.55+0.78
0.69+1.03
0.88+1.11
0.79+1.05
0.88±0.99
1.57±1.79
1. 19+1.40
1.87+1.55
2.53+1.46
2.20+1.52
aSource:  Bull  et al.,  1984a

bMean ± S..D.

cGiven 3 days/week for 8 weeks
                                      8-40
-------
than by the po route,  since A/J mice usually develop from l to 2
lung tumors per animal at 1.5 years of age, acrylamide treatment
acceleratedvtumorigenesis since this yield was attained in
treated animals at a younger age.

8.6. other Data to Support the Carcinogenicity Conclusion

8.6.1.  Mutagenicity Data
     Acrylamide has been shown to be a clastogenic agent in both
in vivo and in vitro studies, and this effect is more pronounced
in germ cells than in somatic cells (for details see mutagenicity
section).  Evidence presently available has not shown acrylamide
to be an effective point (gene) mutagen.  Additional evidence
                i
suggesting that acrylamide can interact with DNA in such a way as
                                                         ^
to produce mutation and possible cancer include DNA damage, and
repair effects of acrylamide and a study that shows that
acrylamide binds to DNA.  A positive in vitro cell transformation
study (and possibly the amplification of SV40 DNA inserts in SV-
40-transformed Chinese hamster cells treated with acrylamide)
further supports the carcinogenic potential of acrylamide.

8.6.2. Absorption. Distribution anfl Metabolism Data
     Absorption of acrylamide has been shown to occur via the
oral and dermal routes of administration.  After absorption,
acrylamide rapidly equilibrates in the body, and does not
accumulate in tissues other than the red blood cells (presumably

                               8-41
-------
acrylamide binds to the sulfhydryl groups of hemoglobin).
Acrylamide is rapidly metabolized in the body, primarily by
conjugatioh?with glutathione, and excreted.  The mechanism of
carcinogenic action is most likely via direct alkylation by
Michael addition, or via metabolism to the epoxide, but the data
do not conclusively support one of these two possible mechanisms
for the observed carcinogenic or mutagenic effects of acrylamide.
(For details see metabolism section.)
     As mentioned above in the discussion of Bull et al. (1984a)
acrylamide (followed by topically applied TPA) was shown to be a
more potent skin tumor initiator when administered orally than
when applied topically to SENCAR mice. (See also Table 8.5.)
Carlson and Weaker (1985) compared the I4c-acrylamide
                i
distribution, and binding to macromolecules, in SENCAR mice
                                                        "A
following oral and topical administration.  Comparing the .two
routes, comparable concentrations were observed in all tissues
except the skin where both the total amount of 14C-acrylamide,
and the amount bound to DNA, was much greater after topical
administration than that observed after oral administration.
Therefore, the effect of route on skin tumor formation cannot be
explained on the basis of either a difference in distribution to,
or binding to DNA in, the target organ.

8.6.3.  Structure-Activity Relationship (SAR) Data              :
     The structure-activity relationship of acrylamide and vinyl
carbamate (the presumed proximate carcinogen form of ethyl

                               8-42
-------
carbamate) has been suggested by Bull et al., (1984a) who showed


that ethyl carbamate and acrylamide behaved similarly in


tumorigeniclty studies. (See 8.2.3 for discussion.)  The authors


argued that ethyl carbamate acted through vinyl carbamate as an


intermediate, and suggested the structural similarity of vinyl


carbamate an acrylamide (note that both are alpha, beta-


unsaturated nitrogen-containing molecules).


     The sites of tumors observed in test animals administered


ethyl carbamate, acrylamide and a third chemical, acrylonitrile


show some similarities, and some dissimilarities (IARC, 1979).


The amount of tumor site concordance observed suggests some
                      /

similarity in the carcinogenic action of these three chemicals.


            O                   "                         O
            II   »                                          M

     H2C=CH-G-NH2'                                  H3C=CH2-OC-NH2

                                                         •s

     acrylamide                                   ethyl carbamate





              O
              n

     H2C=CH-O-C-NH2                                H2C=CH-C=N



     vinyl carbamate                              acrylonitrile





8.7.  Data for Risk Assessment


     Table 8.1 summarizes the tumor incidence data to be


considered for preliminary risk assessment purposes.  Of all the


sites considered, testes, thyroid, and mammary gland were


identified as those to be used for preliminary risk assessment of



                               8-43
-------
individual sites should one be deemed necessary.   Testes was
chosen because the tumors are all malignant and are significantly
increased at a dose level below the highest dose.   Thyroid gland
was chosen because the tumors were significantly increased for
both sexes indicating the sensitivity of this  site to chemical
action across sexes.   Mammary gland was chosen because benign,
and combined benign and malignant tumors, were statistically
significantly elevated at both the high dose,  and the next lowest
dose.
     The EPA guidelines for cancer risk assessment recommend
pooling tumor incidence data for purposes of risk assessment,
since risk numbers derived from site-specific  tumor incidence
data may not be predictive of (and may in fact underestimate)
                 *
"whole-body" risks that are determined using the pooled
individual animal data.  The dose-response curves for each sex
based on the pooled tumor incidence (benign and malignant) data
is the data set of choice for risk assessment.  Further, a second
pool was created for each sex to allow consideration of the
contribution of malignancies alone to the overall risk estimates.
Table 8.2. summarizes the pooled tumor incidence data.
     In order for tumors at a particular site  to be added into
the pool, the tumor site must have been statistically significant
at least at the high-dose level (treated vs. control).  The first
pool contains males having tumors of the testes, thyroid, or
adrenal gland.  For the adrenal gland, only the (benign) adenomas
were considered since neither malignant alone, nor adenoma and

                               8-44
-------
adenocarcinoma combined, were statistically significant.  The
second male pool contains testes mesothelioma only.  The third
pool contains females having tumors of the thyroid (combined),
mammary gland (combined), CNS (tumors only, ho -"proliferations"),
uterus (adenocarcinoma) and oral (combined).  The fourth pool
contains females having malignant tumors of the thyroid, mammary
gland, CNS, and uterus.  Tumors of the clitoral gland in females
are not included in the pooled tumor incidence data because only
a very low number of tissues were examined in each dose group.
                               8-45
-------
                           9.   EXPOSURE
9.1.  Manufacture
9.1.1.  Production Sites. Quantities, and Trends
     Acrylamide is produced in the United States by three
chemical manufacturers (Table 9-1; Figure 9-1).  Total domestic
production capacity as of January 1, 1985 was 225 X 106 pounds
(Ibs.).  In 1985, 140 X 106 Ib.  of acrylamide were produced in
the United States, approximately 62% of production capacity.  The
production volume increased at an average annual rate of 15.3%
between 1978 and 1984; however, future increase in demand is
estimated to average 4% a year through 1989. (Table 9-2).
     American Cyanamid Co. has the_ largest capacity for
acrylamide production (110 X 106 Ibs.),  accounting for 49%  of
total domestic capacity (Table 9-1) .                       •*.
     Less than 2 X 106 pounds  of acrylamide  are imported, mostly
for grouting purposes.

9.1.2.  Production Methods and Processes
     The major commercial production methods that have been used
to manufacture acrylamide are:
               1.   Sulfate process and
               2.   Catalytic hydration process
     Industrial production of acrylamide was initiated by
American Cyanamid Co. in 1954, using the sulfate process
(Matsuda, 1977).  The sulfate process was the major production
                               9-1
-------
        Table 9-1.  Acrylamide Manufacturers, Locations, and Annual

                         Production Capacities
Manufacturer
American Cyanamid Company
Dow Chemical Company
Nalco Chemical Company
TOTAL
Location
Linden, NJ*
Avonale, LA
Midland, MI
Garyville, LA

Annual
Capacity
(X 106 Ibs.)
80
30
100
15
225
Percent
of Total
36
13
44
7
100
 U.S. EPA, 1986.      *
                       I



*This facility was closed in 1986 (American Cyanamid Company, 1936).
                                    9-2
-------
vo
to

           ri..|tm-  '• - 1 .   l*i oilin.-t ion- :;i tr:i lor  ucr y I amide.   From Knviron.  Sc i .  \\\\^ .  (1081)
-------
     Table 9-2.  Annual Acrylamide  Production Rates


                                    Acrylamide Production
           Year                          (X  106 Ibs.)
           __                             _

           1980                             77
           1981                             82
           1982                             86
           1983                             95
           1984                            130
           1985                            140
           1989                            164 (estimate)


U.S. EPA, 1986.

method until the early 1970s  (Davis et al., 1976).   In this

process, approximately equimolar quantities of acrylonitrile,

sulfuric acid, and water are  reacted to produce acrylamide

sulfate  (Figure 9*2) .  The acrylamide sulfate is then neutralized

by reaction with bases (ammonia, calcium hydroxide,  or sodi\im

carbonate) and separation by  filtration.  This procedure is

expressed by the following reaction formulae:
     CH2=CHCN + H20 + H2S04  - > CH2=CHCONH2 .

    Acrylonitrile   Sulfuric Acid      Acrylamide  Sulfate
  CH2=CHCONH2   H2S04   +  2NH3  - > CH2=CHCONH2  +

Acrylamide Sulfate  Ammonia      Acrylamide   Ammonium Sulfate



     Finally, the acrylamide is recovered by crystallization  from

the filtrate.  Dow Chemical Co. (1980) , however, separated

acrylamide from the sulfate salt by passing the  acrylamide

sulfate through a bed of cation-exchange resin  (Figure 9-2) .


                               9-4
-------
10
I
                                                    ACRYIONITRILE
                                                     CHj= CHCN
                                                  IOO*C
                                                  1 Hour
                                                  CH, -
                                                    Acrylamlda Sullata
Lima rrocasa
|ca
-------
     During the early 1970s, the sulfate process for acrylamide



manufacture was replaced by the catalytic hydration process



(Davis et al., 1976; Matsuda, 1977).  In this process  (Figure 9-



3), an aqueous solution of acrylonitrile is passed through a bed



of copper-based catalyst.  This method is expressed by the



following reaction formula:







               CH2*=CHCN  +  H2O  	>    CH2=CHCONH2



          Acrylonitrile       Cu Catalyst        Acrylamide
                               9-6
-------
WATER STORAGE
ACftVLOMITRILE
   STORAGE
     I
REOENERATIpWCVCLE

   I          1
                            CATAIVST
                        ACRYLOMITRIIE
                       AND M,O STRIFTER
 AQUEOUS
ACRYL AMIDE
 STORAGE
   Catalytic fiydratioh  [)roccss  for acrylamide.
   (1981) .
                              From Knvi ron.  JIci
-------
     The catalyst is washed and regenerated in a regeneration
cycle.  The acrylamide solution is concentrated and stripped of
unreacted acrylonitrile by distillation.  The distillate from this
process is recycled to the reactor.  Water is added to the
acrylamide solution to reach the desired concentration, usually 30
to 50 percent acrylamide by weight (Davis et al., 1976).

9.2. Uses
9.2.1.  Acrvlamide Monomer
     Acrylamide consumption in 1984 was estimated to be about 130
million pounds.   Approximately 95% of all acrylamide produced was
used to manufacture polymers of acrylamide.  These polymers,
frequently called^polyacrylamides.  are widely used in water
treatment applications; in the petroleum industry, primarily in
enhanced oil recovery; in pulp and paper production; and in mineral
processing.  The remaining 5% of acrylamide production was consumed
in several small applications.  Table 9-3 lists the 1984 end use
pattern for acrylamide.

            Table  9-3.  Acrylamide  End Use  Pattern,  1984

     End Use             Form         Percent of Total Market
Water Treatment          Polymer               45
Petroleum                Polymer               20
Pulp and Paper           Polymer               20
Mineral Proce'ssing       Polymer               10
Miscellaneous            Polymer/Monomer        5
Source:  CMR, 1985
                               9-8
-------
     Overall, acrylamide demand is forecasted to grow 4% per year
through 1989 (CMR, 1985).

9.2.2.  Polvacrylaroide Polymers
     Acrylamide undergoes polymerization to produce a group of
versatile, synthetic polymers called polyacrylamides.  Polymers of
commercial significance include those derived from polymerization
of acrylamide with itself (homopolymer) and those derived from
polymerization with other comonomers (copolymers) (Schwayer, 1981).
     While a large number of commercial polyacrylamides exist, they
are usually classified according to their molecular structure,
molecular weight (size), and the electrical charge they exhibit in
water.  Polyacrylamides are either low (< 100,000 g/mole) or high
(> 1,000,000 g/mole)  in molecular weight.  The polymers are*, also
categorized as either anionic, catibnic,  or nonionic, depending on
whether they exhibit a negative, positive or no electric charge.
The polymer's electric charge and molecular structure are
determined by the particular method of preparation.
     Polyacrylamides are readily soluble in water over a broad
range of conditions.   They can be engineered to fit a large number
of uses (Davidson, 1980).  Polyacrylamides have several properties
which have led to their use in a wide variety of industries.  Table
9-4 lists the key functions of polyacrylamides, sample commercial
applications, and the industries currently using them.  The
specific polymer used in each application is determined by the
performance requirements of the specific use and the
characteristics of the polymer itself.
                               9-9
-------
              Table 9-4.   Functions of Polyacrylamides
Function
Application
Industry
Flocculation
Rheology Control
Adhesion
Solids recovery
Waste removal
Water clarification
Retention aid
Drainage aid
Waterflooding
Viscous drag reduction
Dry strength
Wallboard cementing
Mining
Sewage
General
Paper
Paper
Petroleum
Petroleum
Fire fighting
Irrigation
  pumping
Paper
Construction
Source:  Davidson? 1980
     The general properties of polyacrylamides are derived •tfrom
the chemical nature and structure of the acrylamide monomer. "
Commercially available polyacrylamides are subdivided into two
major classes based upon their electrical charge.  Anionic or
negatively charged polyacrylamides are obtained from the
hydrolysis of the acrylamide's amide group or through
copolymerization of acrylamide with acrylic acid.  Cationic or
positively charged polyacrylamides are produced via
copolymerization of acrylamide with other chemicals, such as
formaldehyde and amines.
9.2,2.1.  Polyacrvlamide Manufacturing Process
     Anionic polyacrylamides are produced by one of two
processes.  The first involves simultaneous polymerization and
                               9-10
-------
hydrolysis of acrylamide.  Sodium hydroxide or sodium carbonate
is added to an aqueous solution of acrylamide to give the desired
degree of hydrolysis.  The acrylamide is then polymerized.  The
resulting polymer contains carboxyl groups which ionize in water
to give the polymer its anionic character.  Alternatively,
anionic polyacrylamides are produced by co-polymerization of
acrylamide with acrylic acid or acrylate salts.   Polymers of
comparable molecular weights appear to function similarly
regardless of the method of preparation (Schwayer,  1981).
     Cationic polyacrylamides are prepared by one of two methods.
In the first, acrylamide is first polymerized to produce a
homopolymer which is then reacted with formaldehyde and
dimethylamide to form a cationic aminomethylated polyacrylamide.
Alternatively, acrylamide can be co-polymerized with a variety of
monomers which impart cationic functionality.  Some of the more
important products are diallyldimethyl ammonium chloride-
acrylamide copolymer, aminomethacrylate-acrylamide copolymer,  and
dimethyl-aminoethyl methacrylate-acrylamide copolymer (Kirk-
Othmer, 1982).
     Polymerization of acrylamide occurs under a number of
conditions.  The most commonly used process is aqueous solution
polymerization.  Figure 9-4 depicts the important steps in the
manufacture of liquid polyacrylamide by this method.  The
polymerization reaction can be initiated by peroxides,
persulfates, azo compounds, or redox reagents.  The molecular
weight of the final polymer can be varied by several techniques,

                               9-11
-------
VO
I
H
(O
Acrylpn.*. MB
                           -fi
                              SH
                           fetrwUMtan
                             •MiUMt
                                                 0*0
                                       u
fi
      fteHSO,
     W*M>|
      rL

o
o
i'
o
o
                                                              •um»
                                •v   min unk
        Figure 9-4.  General diagram of solution polymerization,  liquid products proces:
                    (Kirk-Othmer, 1978).
-------
including acrylamide concentration, initiator concentration,
polymerization temperature, polymer chain-transfer agents, and
electrolyte concentration (Kirk-Othmer, 1978).  The main drying
techniques used to manufacture solid high-molecular weight
polyacrylamides from aqueous solutions are thermal sheet drying
and solvent precipitation (Kirk-Othmer, 1978).

9.2.3  Polvacrylamide End Uses
9.2.3.1  Water Treatment Applications
     Water treatment applications represent the largest market
for polyacrylamide.  It is estimated that these uses currently
consume about 45% of the acrylamide produced annually in the
United States (CMR, 1985).  Applications include the use of
polyacrylamides in municipal water and wastewater treatment*
plants, in sludge conditioning operations, and in industrial raw
and wastewater treatment plants.  In addition to being the
largest use, these applications probably represent the most wide-
spread use of polyacrylamides.  In general, the use of
polyacrylamides in these uses is based upon the polymers' ability
to facilitate the removal of solids from water.

9.2.3.2.  Pulp and Paper Applications
     Polyacrylamide use in pulp and paper applications accounted
for approximately 20% of acrylamide consumption in 1984 (CMR,
1985).  These uses are based on polyacrylamide's excellent
flocculation and chemical bonding properties (Bikales, 1973;

                               9-13
-------
Kirk-Othmer, 1978).  Water-soluble polyacrylamides are used for
the following purposes.
     o    To improve the dry-strength of paper sheet;
     o    To increase the retention of pigments, inorganic
          fillers, and other small particulate matter that are
          added to paper sheet; and
     o    To improve the drainage of water during paper
          formation.

(American Cyanamid, 1969; Bikales, 1967; Kirk-Othmer, 1978).

9.2.3.3.  Petroleum Applications  .
     The petroleum, industry uses polyacrylamides for several
applications.  The largest use is in enhanced oil recovery "*•
processes.  They are also used in drilling-mud additives and in
lubricant additives (Chem Purch, 1983; Kirk-Othmer, 1982a and b;
Nyguist and Yocum, 1973).

9.2.3.4.  Mining and Mineral Processing Applications
     Polyacrylamide applications in mining and mineral processing
account for 10% of the consumption of aery1amide.  High-molecular
weight polyacrylamides are used as flocculants for the recovery
of valuable metals from ores, the recovery of tailings and
slimes, and the clarification of wastewater.  Low-molecular
weight polyacrylamides function as sequestering and
antiprecipitating agents, which reduce scale deposits in

                               9-14
-------
equipment and keep solids suspended in solution for subsequent
removal (American Cyanamid, 1969; Bikales, 1968; Nyquist and
Yocum, 1973).

9.2.4.  Other Uses of Acrylamide
9.2.4.1.  Acrvlamide Grouts
     Acrylamide has been used in the formulation of chemical
grouting agents.  Chemical grouting is the practice of injecting
chemicals into soil, rock or concrete formations to aggregate
them to form a water barrier.  Grouting is normally employed when
it is desirable to restrict or redirect the flow of water through
a formation or to improve formatioji strength.  Grouting has been
performed during the construction or rehabilitation of dams,
buildings, tunnels, mine shafts, and other structures (KirJo-
Othmer, 1979).
     The predominant use of acrylamide grouts is in stopping
water infiltration into sewers.   It is estimated that about one-
half million pounds of acrylamide grout is used annually for this
purpose (Geochem, 1985).  Specially-designed injection equipment
is moved through underground sewer lines,  and positioned near
leaking sewer joints through the use of video cameras.  After
treatment, the cameras provide an immediate determination as to
the effectiveness of the seal (Kirk-Othmer, 1979)  (some grout is
injected through hand held injector guns,  especially in manhole
sealing operations).  Smaller quantities of acrylamide grout are
                              9-15
-------
believed to be used in repairing cracks in tunnels, mine shafts,
and ground application (Celtite, 1985).
     American Cyanamid had been the major domestic producer of
acrylamide grout until the company terminated production in 1978.
Imported grouts based on acrylamide are available from:  Avanti
International, Inc. (Webster, TX), Cues, Inc. (Orlando, FL), and
Polymer Chemicals  (Atlanta, GA).  Avanti and Cues import and
market an acrylamide grout produced in Japan by Nitto Chemical
Company.  Polymer Chemicals imports from a French company.
Mitsubishi and Sumitomo Trading Companies act as agents for Nitto
in these sales.  Avanti markets the grout under the tradename AV-
3.00 and Cues markets it under the trade name Q-Seal (Cues, 1985;
Geochem, 1985).  $oth products are sold as dry powders (Karol,
1983) .                                                      *
     Acrylamide-based grouts generally consist of a mixture of
two monomers:  acrylamide, which represents 95% of the mixture,
and a cross-linking agent, such as methylene-bis-acrylamide,
which represents the remaining 5%.  In commercial use, an aqueous
solution of the powdered acrylamide grout is prepared at the
field location.  The solids content of the solution can vary, but
the solution is prepared so as to contain 10% solids when
injected into the sewer line crack.  An activator chemical,
typically triethanolamine, is then mixed into the solution.  A
second solution is prepared consisting of an initiator or
catalyst, typically ammonium persulfate, in water.  Once the
injection equipment is positioned at the area to be repaired, the

                               9-16
-------
 solutions  are mixed  and  injected  into  the crack.   Upon injection,
 the acrylamide  polymerizes  and  the  cross-linking  agent binds the
 polymer  chains  together,  converting the mixture into  the gel,
 thus closing the hole  in the  sewer  line  (Geochem,  1985;  Karol,
 1983) .   The gels are considered permanent and are unaffected by
 exposure to chemicals, except for very strong acids and  bases.
 The gels are, however, subject  to mechanical deterioration when
 exposed  to alternating drying and/or freezing cycles  (Karol,
 1983).
     In  gel form, the  grout contains very little  free acrylamide.
 When properly prepared,  the gel produced from a 10% solution of
.grout contains  less  than 0.03%  acrylamide.  It is known  that many
 microorganisms  in*the  soil  and  natural water assimilate  ungelled
 acrylamide (Karol, 1983).                                  ^

 9.2.4.2.   Monomer Derivatives
     The total  number  of applications  of acrylamide and  its
 derivatives are unknown.  One important and growing use  of
 acrylamide is in the textile  industry.
     Acrylamide is used  to  produce  n-methylol acrylamide, which
 is used  as a textile anti-creasing  agent.  In one patented
 process  the n-methylol acrylamide is introduced into  cellulosic
 fibers,  such as cotton,  and is  cured with radiation.   The
 irradiation step causes  the double  bond of the acrylamide        ;
 derivative to cross-link with the cellulose.  The use of this
 derivative in fabric treatment  results in good resistance to

                               9-17
-------
creasing and generally good abrasion resistance  (Textile Res J,



1960).



     A small amount of acrylamide is used in laboratories for the



preparation of gels for electrophoresis separations.







9.3.  Acrylaroide Exposure



9.3.1.  Populations Exposed



     Plant workers at both acrylamide and polyacrylamide



production facilities and those involved in applying acrylamide



grout are exposed to the monomer.  Exposure can result from



bagging operations at solid acrylamide production plants (the one



U.S. site producing solid acrylamide ceased the production in



1985), disposal of; used containers, washing down or otherwise



treating spilled monomer, reactor or transport vessel cleaning,



or mixing dry grouting formulations in the field.  Direct worker



contact (skin and eyes) with either the liquid or solid chemical



is the most likely (and perhaps more important) route of



exposure,  although the workplace air can be a source if the



monomer has accumulated in any areas, particularly in concrete



walls or floors.  Workers are routinely advised to wear



protective clothing,  rubber gloves, goggles, and respirators,



where necessary.



     In addition to the persons described above, some laboratory



personnel are exposed to acrylamide during the preparation of



polyacrylamide gels for gel electrophoresis.  Such exposure could



mainly occur during weighing,  mixing, and during the subsequent





                               9-18
-------
use of the gel solution.  The route of exposure could be either
inhalation or dermal, although dermal exposure is more probable
because crystalline acrylamide is rather hygroscopic.
     There are no individual consumers of acrylamide or related
products.  Users consist of other companies purchasing monomer
for use as a feedstock, municipalities or industries purchasing
polyacrylamide products for any one of their multiple uses (see
section 9.2.3).  Thus, product exposure would be limited to
industrial and municipal workers.  The exposure would be through
residual acrylamide in the polyacrylamide product.   In the
general population, persons in cities served by drinking water
supplies treated with polyacrylamide flocculating agents and
consumers of sugar, may be exposed to acrylamide monomer, again
from its occurrence as a residue in the polymer product.   *
     Other portions of the general population that might be
affected would be those people living in the vicinity of grouting
applications for soil stabilization and dam repair and
populations living near acrylamide production plants.  Soil
stabilization and dam repair are minor application areas compared
to sewer rehabilitation and, therefore, population exposure would
be expected to be minimal.  Environmental monitoring at
acrylamide production sites indicates little potential for
exposure (GCA, 1980).
                              9-19
-------
9.3.2.  Population Estimates

     Table 9-5 presents estimates of the number of persons

exposed during the manufacturing, processing, and use of

acrylamide or acrylamide-based products.

     NIOSH estimates that between 700 and 1,000 workers are

potentially exposed to acrylamide in the manufacture of

acrylamide monomer and polymers, based on reporting from the

acrylamide manufacturers (NIOSH, 1985).  The NIOSH, National

Occupational Hazard Survey lists the estimated total number of

U.S. workers who are potentially exposed to acrylamide to be

approximately 10,000 employed in 27 occupations (NIOSH, 1980).

Workers performing acrylamide grouting are estimated to number

1,800 (Abt Associates, 1990).  In addition, at least 100,000 to

200,000 persons are potentially exposed to acrylamide in   ^

laboratories from the preparation and use of  polyacrylamide gels

for electrophoresis (Versar, 1986).

     Indirect exposure from the use of polyacrylamide flocculants

to treat drinking water ranges from 4.5 million to 30 million

persons (Delpire, 1985).   The former number is the most likely

estimate given flocculant usage rates, polymer sales and
               •
household water consumption.

     The number of persons exposed to acrylamide from the use of

other polyacrylamide products is unknown.
                              9-20
-------
             Table 9-5.  Estimates of the Number of
                         Persons Exposed to Acrylamide
Exposure Category	Estimated Number Exposed

Manufacture at total of                        < 100
 Four Facilities

Manufacture/Processing of                     500-1,000
 Acrylamide monomer                    as per Public Comments and
                                       EPA Review

Soil Grouting                                   1,800

National Occupational Hazard Survey —         10,000
 Occupational Exposure —
 27 Occupations

Laboratory Personnel—                      100,000-200,000
 Gel Electrophoresis

Drinking Water                             4,500,000-30,000%000

Sugar Consumption                            Up to 230,000,000
                              9-21
-------
9.4.  Exposure Estimates
     The exposure information presented below was obtained from
recent EPA sponsored field studies and industry supplied data.
When the first section 8(e) notice reported findings of
statistically significant numbers of tumors in female rats, it
was determined that the existing exposure information would be
inadequate for risk assessment.  Consequently, after reviewing
the populations potentially exposed, it was decided that two
categories should be monitored, manufacture and processing of
aery1amide monomer and soil grouting to repair sewers.  The two
categories were selected because they have the greatest potential
for significant dermal and inhalation exposure as opposed to
persons potentially exposed to acrylamide from its presence as a
residue in polyacrylamide products.  Such exposures would be very
low in many instances even assuming worst case scenarios (e.g*.,
potable water treatment).
     The first study by EPA examined acrylamide manufacturing and
processing sites and one soil grouting site.  This study was
conducted by NIOSH under an interagency agreement.  Subsequent
monitoring by EPA at additional soil grouting site was prompted
by the high dermal exposure potential reported by NIOSH.  Other
exposure estimates presented are based on usage rates of
polyacrylamide products (drinking water treatment and sugar
refining).  Other users of polyacrylamide products were not      :
extensively studied because the focus of the risk assessment is
on persons exposed to acrylamide monomer.  Drinking water

                               9-22
-------
treatment and sugar refining were examined in limited detail
because of the very large number of persons potentially exposed.

9.4.1.  Acrylamide Manufacture and Processing
     NIOSH conducted industrial hygiene surveys at all four
domestic acrylamide manufacturing plants represented by three
companies (American Cyanamid, Dow and NALCO).   These plants also
manufacture polyacrylamides.   Area and personal air monitoring
was performed as well as observations of work practices.
     Six job classifications cover the workers in the acrylamide
monomer and polymer production areas.  These are monomer
operators, polymer operators, monomer material handlers, polymer
material handlers? maintenance workers, and utility operators.
The monomer and polymer operators control the manufacturing*
process.  A portion of their time is spent in the control room
and the remainder is spent in the process area.   The material
handlers load bags, drums, trucks, and rail cars with the final
product.  Other workers in the acrylamide areas are the
maintenance workers and the utility operators who supply water,
electric power, and steam to the plant (Hill  and Greife, 1986).
     NIOSH collected a single, full period eight-hour personal
sample for all workers who were potentially exposed to
acrylamide.  Eight-hour time-weighted averages (TWA) were
calculated in order to compare exposures between manufacturing
plants.  Wipe samples were collected as an indication of possible
dermal exposure.

                               9-23
-------
      The sampling and analysis was accomplished by collecting
 acrylamide vapor and particulate on a 37 mm mixed cellulose ester
 filter-pore size 0.8 urn,  Millipore type AA, followed by a silica
 gel  tube SKC no.  226-10,  100/50 at a flow rate of 1 liter per
 minute.   The collected material was then desorbed from the
 collection media with water and analyzed by high performance
 liquid  chromatography with an ultraviolet detector.  The method
 was  validated at air concentrations of 0.03 to 0.6 mg acrylamide
 per  cubic meter of air based on a sample size of 480 liters of
 air  (American Cyanamid Company,  1981)  (Hill and Greife,  1986).
      Wipe samples were collected as an indication of non-
,:respiratory exposures.  The surfaces that were wiped included the
 exterior of gloves^ and hard hats,  inside the face piece of
 respirators,  desk tops in the control rooms,  on cafeteria table
 tops, on laboratory bench tops,  and on reactor vessels.
      The air monitoring data for the 4 manufacturing sites are
 summarized in Tables 9-6,  9-7,  and 9-8.   Two personal air samples
 were taken at the grouting site which were 0.002 and 0.007 mg/m3
 as 9 hour TWA's.   The mean is 0.005 mg/m3.
      In addition to the NIOSH data,  Sobel et al.  (1986)  reported
 (Dow Chemical)  8 hour TWA levels in the monomer production area
 ranging from 0.1 to 1.0 mg/m3 for pre-1957, 0.1 to  0.6 mg/m3 from
 1957 to 1970,  and <0.1 mg/m3  for post-1970.  In the polymer
 production area,  most jobs had polymer dust level of <2  mg/m3
 except  for packagers and  dryer operators who had levels  >2 mg/m3.
 If a maximum 1% residual  acrylamide is assumed (Sobel et al.,

                               9-24
-------
 1986),  acrylamide  exposure for the  polymer workers described
 above would be  <0.02 mg/m3 and  >0.02 mg/m3, respectively.  These
 data  agree well with the  NIOSH results.  Koppers  (1979)  reported
 acrylamide levels  in an amino  resin production  area ranging from
 <0.1  to <0.6 mg/m3 for an amino resin operator and  <0.09 to <0.05
 mg/m3 for the assistant amino resin operator, and <1 mg/m3 for
 the filter press operator.   The < notation denotes  analytical
 limit of detection (monitoring was  performed to duplicate  an OSHA
 compliance officer monitoring  technique).  Koppers  reported that
 other monitoring indicated exposure potential to be less than  the
 OSHA  standard of 0.3 mg/m3.  Finally,  Collins (1984) reported
.acrylamide exposure levels ranging, from  0.08 to 0.70 mg/m3 in the
 monomer production, area,  0.03  to 0.2 mg/m3 in the polymer
 production area, and 0.15  mg/m3 for maintenance workers.   "*
      Based on the  NIOSH survey, American Cyanamid's Linden,  New
 Jersey  site had the greatest range  of acrylamide air levels.
 This  was due mainly to the production of dry acrylamide.   The
 highest levels  were found  in enclosed areas of  the  production
 building, which workers would  only  occasionally enter with a
 full-facepiece  canister-type respirator  for acrylamide dust  and
 vapors  (this site  is no longer producing acrylamide).  Unlike
 this  site, the  other sites had all  or a  portion of  their
 production process equipment open to the ambient air.  The
 acrylamide production equipment at  American Cyanamid's other site
 is entirely outdoors.  The production equipment at  Dow and NALCO
 are within a building, but natural  ventilation  is provided.

                               9-25
-------
     Table 9-6.  NIOSH Monitoring
  American Cyanamid-Linden, NJ Site
               PERSONAL AIR SAMPLING DATA
ACTUAL CONG.
MONOMER PRODUCTION
   8 HOUR-TWA
NO.
JOB TITLE SAMPLED
OPERATORS
UTILITY OPERATORS
MATERIAL HANDLERS
MAINTENANCE
TOTAL FOR
MONOMER AREA

6
2
3
4
15

NO.
JOB TITLE SAMPLED
OPERATORS
MATERIAL HANDLERS
TOTAL FOR
POLYMER AREA
9
3
12
RANGE
.087-
.309-
.026-
.001-
.001-

IN ma/m3
.238
.410
.298
.080
.410

ACTUAL CONC.
RANGE IN ma/m3
.012-
.022-
.012-
.223
.043
.223
MEAN
.185
.360 . »
.156
.048
.166
PERSONAL
MEAN
.082
.031
.069
SD RANGE
.055
.071
.136
.034
.118
.079-.
.306-.
.017-.
.001-.
.001-.

227
392
260
073
392
I
AIR SAMPLING DATA
SD RANGE
.075
.011
.068
.011-.
.018-.
.011-.
MEAN
.170
.349
.138
.035
.152
POLYMER
SD
.053
.061
.122
.029
.115
PRODUCTION
8 HOUR-TWA
MEAN SD
181
035
181
.069
.026
.059
.061
.009
.056
-------
 Table 9-6 Cont.  NIOSH Monitoring
American Cyanamid-Avondale, LA Site
           MONOMER PRODUCTION
ACTUAL CONC. 8 HOUR-TWA
NO.
JOB TITLE SAMPLED RANGE IN mcr/m3 MEAN SD RANGE MEAN SD
OPERATORS
MATERIAL HANDLERS
MAINTENANCE
TOTAL FOR MONOMER

OPERATORS
MATERIAL HANDLERS
TOTAL FOR POLYMER
5 .034-. 050 .041 .006
1 .517
3 .016-. 157 .064 .081
9 .016-. 157 .102 .161
POLYMER PRODUCTION
8 .055-. 105 .074 .016
1 .018
9 .018-. 105 .068 .024
.032-. 046 .039 .005
.051
.015-. 132 .054 .067
.015-. 132 .045 .035

.044-. 098 .065 .019
.017
.017-. 098 .060 .024
-------
  Table  9-7.   NIOSH Monitoring
Dow Chemical Co.-Midland, MI Site

       PERSONAL AIR MONITORING  DATA
         MONOMER  PRODUCTION

           ACTUAL  CONC.
8 HOUR-TWA
NO.
JOB TITLE SAMPLED
Operators
Lab Tech
Total for Monomer


8
1
9


NO.
JOB TITLE SAMPLED
Operators
Maintenance
Lab Tech
Shipping Clerk
Total for Polymer
10
5
1
1
17
RANGE
.001-. 119
.053
.001-. 119


RANGE
.001-. 024
.001-. 012
.015
.004
.001-. 024
mg/m3
MEAN SD RANGE
.062 .033
.053
.061 .033
.001-. 113
.049
.001-. 113
PERSONAL AIR MONITORING DATA
POLYMER PRODUCTION
ACTUAL CONC. .
mg/m3
MEAN SD RANGE
.008 .007
.005 .004
.015
.004
.007 .006
A
.001-. 021
.001-. Oil
.014
.004
.001-. 021
MEAN
.049
.049
.0487

8 HOUR-TWA
MEAN
.007
.005
.014
.004
.007
SD
032
—
030


SD
006
004
—
—
005
-------
                               Table 9-8.  NIOSH Monitoring
                           Nalco Chemical Co.-Garyville,  LA Site

                                         PERSONAL AIR MONITORING DATA
                                         MONOMER AND POLYMER PRODUCTION

                                       ACTUAL CONC.
8 HOUR-TWA
NO.
JOB TITLE SAMPLED
Operators
Utility Opera tors
Lab Tech
Material Handlers
Maintenance
Total
9
2
2
4
2
19
RANGE
.003-. 012
.004
.003-. 004
.003-. 005
.011-. 014
.003-. 014
mg/m
MEAN
.006
.004
.004
.004
.013
.006
SD RANGE
.003
0
.001
.001
.002
1
.003
.003-. 012
.004
.003-. 004
.003-. 005
.005-. 006
.003-. 012
MEAN
.006
.004
.004
.004
.006
.005
SD
.003
0
.001
.001
.001
.002
ND below .0035 mg/m
-------
     Monomer operators at American Cyanamid's sites and at Dow
had a mean exposure level twice that of the polymer operators.
Two utility operators at the Linden, New Jersey site had eight-
hour TWA exposures above the OSHA standard.  The company
performed side-by-side sampling during the NIOSH survey, and
confirmed the utility operators' exposure levels.  The higher
levels could not be explained, because their job duties do not
involve direct exposure to acrylamide.  Subsequent monitoring of
these workers by the company revealed much lower exposure levels.
Further monitoring of utility operators is necessary to determine
sources of exposure.  Monomer material handlers can have brief,
high exposures when loading trucks- or rail cars with acrylamide
solution if they £ail to wear personal protective equipment.  The
workers that were observed wore full-facepiece respirators,^
neoprene gloves, and an apron during this operation.
     During the study, only minor maintenance occurred at each
facility.  However, these maintenance workers may potentially be
exposed to airborne levels of acrylamide above the OSHA
permissible exposure limit or they may be exposed by dermal
contact.  At the NALCO site, the major cleaning of the reactor
equipment was performed by workers from a small contracting firm
rather than in-house personnel.
     The wipe samples were collected from personal protective
equipment and workplace surfaces where skin contact was most
likely.  Nearly all of the wipe samples had no detectable
acrylamide, except for a wipe sample collected on the exterior of

                               9-30
-------
a polyacrylamide reactor vessel which had 30 ug/sample, a
laboratory counter top which had 3.2 ug/sample, and a door handle
which had 0.9 ug/sample (Hill and Greife, 1986).

9.4.2.  Soil Grouting
     Although soil grouting is a minor use of acrylamide, it
presents the potential for high dermal exposure.  As part of the
NIOSH field investigation, a sewer grouting repair site was
investigated.  Subsequently, EPA carried out a field monitoring
survey at four sewer grouting sites (MRI, 1987).  At the sewer
line repair site surveyed by NIOSH, two employees performed the
repairs with the use of two service trucks.   One service truck
held the grouting*equipment (hoses, pumps, mixing tanks, and
video monitoring equipment), while the other truck contained
hoses and a water supply.   The workmen first assembled the
packer, which was lowered into the manhole and then passed
through the sewer.  (See Figures 9-5 and 9-6.)   It was positioned
by the use of cables and a video camera viewed from the truck's
control panel.  At the site of a leak, both ends of the packer
were inflated, isolating the leaking joint.   In the service
truck, an employee poured the grouting material, which was 95%
acrylamide and 5% methylene-bis-acrylamide,  into a mixing tank
containing water.  The acrylamide was bagged with an inner liner
bag that could be placed below the water level of the mixing tank
to prevent dust particles from escaping into the air.  Acrylamide
solution, along with a catalyst, was injected under pressure from

                               9-31
-------
        lo.lin. and < Uaucol
        MOMi la S«|.U« V.MU
VO
 I
                                     PocUc. fotlllofwd
                                     al Uali Point
IV C
                               9^5,    Sleeve  packer  and camera assembly  inside mainline sewer.
-------
lO
 I
w
                                                                                                                                                Conliol Panel

                                                                                                                                   (?) Quo«J Lin* Chemical Hot* uiid
                                                                                                                                        felevitlon fianimiiiioit Cable.
                                                                                                                                        Powci Reel
 (j)  Mulli - Cioul Chemical
      Pump Attend)y

 (4J  Slolnleu Steel Clwinical lunkt

 (T) Walci Sloiaoe (unit

 (A) Alt Conditioned anJ lleuted
     C'onltul  Monilorinu Room

(^)  llcctiic Start Generator

(ft)  tlecliii  Ail
                                                                           4
                                 9-6,    Typicdl   line mdinU>iidnce vehicle  shuwinq  yroutififj  e(|ui|)inerit,
-------
the service truck via hoses to the center of the packer and
forced into the surrounding soil.  The monomer polymerized to
form a water-impermeable gel, sealing the leak.  Once the leak
was sealed, the packer was deflated and moved to the next joint.
Two short-term exposure samples (18 minutes) were collected in a
personal monitor on the individual who poured and mixed the dry
acrylamide grout into the mixing tank.  Nine-hour personal air
samples levels were 0.002 and 0.007 mg/m3.   Two short term
personal air samples, of 18 minute duration, collected during the
pouring and mixing of the grout, did not show any detectable
acrylamide.  This may be due to the high moisture content of the
acrylamide and the method of pouring the grout.  Area air samples
collected inside the service truck and the company garage
revealed 0.009 and 0.001 mg/m3,  respectively.               *
     Surface wipe samples collected from the control panel in" the
service van, on a safety cone placed on the road, and on the
exterior of two rubber work gloves were free of acrylamide
contamination.  However, a wipe sample from the side of the grout
mixing tank detected 44 ug/cm2.  The  source  of  this  contamination
was apparent when a small amount of solid acrylamide was spilled
on the tank and floor during the mixing.  The spill was not
cleaned up, and when the moisture in the grout evaporated,
acrylamide could have become airborne as a dust or vapor.   Other
poor handling procedures observed, such as storing empty grouting
bags and mixing cups on the floor, or only occasionally wearing
                              9-34
-------
rubber gloves when handling the packer or washing the equipment,
could lead to dermal contact with the acrylamide.
     Another source of dermal acrylamide exposure is the inside
of the work gloves.  The gloves were intended to prevent skin
contact with the grout.  When distilled water was rinsed inside
one of the gloves, the rinse water was found to contain 65 ug of
acrylamide.
     In the EPA study, four sewer grouting sites were
investigated.  However, in contrast to the NIOSH study, the
primary purpose of the study was to measure dermal rather than
respiratory exposure.  Air monitoring was also done.  At the four
sites, three grouting procedures were surveyed, manhole repair
(two sites), mainline repair and lateral line repair.  The latter
two are remote controlled operations.                      *
     Personal and area air samples were collected on a 0.8 urn*
mixed cellulose ester filter and silica gel sampling train, using
calibrated battery-operated sampling pumps at a nominal flow rate
of 1.0 L/min.  One area sample,  except at the two manhole
grouting sites, was collected inside the Mobile Reveal and Seal
Unit, where the mixing tanks are located.  Personal samples were
collected in the breathing zone of workers during the chemical
grout mixing operation and during the grout injection operation
whenever manholes were entered.
     Air sampling data were recorded on the Air Sampling Data
Sheets (Figure 9-7).  The information collected for each personal
or area air sample included:  employee and work site data,

                              9-35
-------
                                                            Contract Number:
                         .Date (Ho-^i/V
  Site ID
 Work Assignnt No.
•            47
Substance  Ncn.
1  ACRYLAMIDE
                                           EMPLOYEE AND UORK AREA DATA
  Employee Nwe
iUork Location  Description
  Job TitleAtork Duties
                                                           jUc«th«r  Conditions

SftflMB EOUIPWfT
(
Iistruacnt ! Model No. 5cri<| No.
i
U Personal (I Are* U Bulk

+
\
Field Smplt ID Njibtr
Start T,t»
Siflplt Collection Media: Lot No. j
B.8u MIXED CELLULOSE ESTER FILTER AM) SILICA G£L TUBE ]
FIELD SAMPLING- INFORH4TKM
-

| Stoo Tine
: Sirpie Duration (mms.)
P«rp Flow Rat* (
Sirple Air Voluw (
; Laboratory Analysis
Analrte
. l. flCRYLAMIDE
= 2.
3. I
. Signature:
Calculations Checked bx:

)
)

Uiit»~)








Laboratory
Results









Air
Concentration









Laboratory
Results









Air
Concentration







i

Laboratory < Air
Results Concentration
;

1 ;
1 :
J
Date:
Date:
                           Figure 9-7.   Air SanpLing Data Sheet

                                                9-36
-------
Calibration Method
                                            CALIBRATIth DATA
Yoltne
Prt Oit*
1.
2.
3.
Signaturt

Pott Ditt
1.
2.
3.
Signjturt
Rtsistan
Cilculit



Avtrtge T
Addition*! Cow*nts, ObsiPwttions, Oitgrwis,  0»ta
                                                       , etc.
                               Figure  9-7 cont.
                                                    9-37
-------
sampling equipment data, sampling parameters, exposure data,



calibration data and general observations  (work practices, safety



equipment, etc.).



     Dermal contact sampling was performed using the dermal pad



and hand rinse methods as described by Durham and Wolfe  (1962).



Observations made during a preliminary site visit to a chemical



grouting operation indicated that significant dermal contact



could occur on the face, neck, and forearms of workers even



though they wore impervious clothing.  The worker's torso, upper



arms, and legs are protected by the impervious suit.  When full



protective clothing was used by a worker, dermal pads were placed



at six body locations to assess dermal contact to the face, neck



and forearms.  Wheji protective clothing was not utilized, dermal



pads were placed at ten body locations to assess dermal contact



to the entire body.  Hand rinses were conducted using the bag*



techniques as first described by Durham and Wolfe (1962).  Dermal



contact assessments were conducted during equipment assembly



operations, grout mixing operations, grout injection operations,



and equipment disassembly operations.



     Air monitoring results expressed as 8-hour TWA's appear in



Table 9-9.  The 8-hour TWA's ranged from ND to 0.12 mg/m3,  with a



mean of 0.045 mg/m3.   Except for the maintenance  supervisor at



site number one, the values are similar to the NIOSH data for



manufacturing/processing workers.  However, the dermal monitoring



indicates significant dermal exposure, even among workers using



safety equipment.  (See Tables 9-10, 9-11, 9-12,  and 9-13.)





                               9-38
-------
             Table  9-9   Acrylamide Inhalation Exposures
Sample description
                                   Air
                                  cone.
8-h
TWAb
                 OSHA
                 PELC
                 ACGIH
                  TLV*1
(mg/m )  (mg/m )  (mg/m )*  (mg/m )
Site no. 1
 Breathing zone of Maintenance
 Supr.
 Breathing zone of Util. Worker
 no. 1

Site no. 2
 Breathing zone of Util. Worker
 no. 2
0.360

0.100
0.120

0.003
0.040   0.010
         0.03

         0.03



         0.03
0.03

0.03




0.03
Site no. 3
Breathing zone of Grout Foreman
Breathing zone of Laborer
Area sample near mixing tanks
Site no. 4
Breathing zone of Utility Worker
Area sample near mixing tanks
Area sample approx. 20 ft from
service van

0.060
0.040
0.050

0.008
0.080
NDe


0.060
0.040
0.050

0.008
0.070
ND


0.03
0.03
0.03

0.03
0.03
0.03


0.03
0.03
0.03

0.03
0.03
0.03
_,
8Air concentration of acrylamide during sampling period in milligrams
 per meter of air.
bEight-hour time weighted average exposure (TWA)  assuming no exposure
 to acrylamide during periods not sampled.
Occupational Safety and Health Administration  Permissible Exposure
 Limit.
threshold limit value (TLV)  published by the American Conference of
 Governmental Industrial Hygienist (1987-1988  edition).
eND = none detected, the detection limit was 0.004  mg/m3  for a 480-L
 air sample.
* "skin" notation
                                 9-39
-------
                                Table 9-10  Dermal Contact Estimate
 p.te no.  1
 Manhole Sealing Operation
 Maintenance  Supervisor
Surface
area of
region8
Body
region
(cm')
Exposure pads
used to
represent
body
region
Surface
area of
pads
(cm2)
Total Time
acrylamide of
found
(mg)
Dermal
contact
exposure estimate1*
(h)
(mg/h)
Head             1,090
Back of neck       115
Front of neck      115
Forearms         1,290

Hands             N/A
shoulder pads  (4)   400
back pads  (2)       200
chest pads  (2)      200
forearm pads  (4)    400
                                                       1.76
                                                       0.109
                                                       6.75
                                                       0.386
2.9
2.9
2.9
2.9
                          hand rinses (6)      N/A     12.1      7.4

                                          Total Dermal Contact Estimate
  1.6
  0.02
  1.3
  0.43

  1.7

= 5.0
iiourface area of body regions based on  the  anatomic dimensions of the 50th
  Ifercentile man  from Popendorf  (1976,  1982).  Diffrient et al. (1974)', and NASA
  (1962) .                          A    x S,
bDermal contact estimate  (mg/h) =  TOT	'
                                   T x S2

where ATOT = total  acrylamide found (mg)

     S, = surface area of body region  (cm2)
     S2 = surface area of pads (cm2)
     T  = time of exposure  (h).
                                       9-40
-------
                       Table 9-11  Dermal Contact Estimate
Site no. 2
Manhole Sealing Operation
Utility Worker no. 2
Body region
Head
Back of neck
Front of neck
Forearms
Hands
Surface
area of
region*
(cm2)
1,090
115
115
1,290
N/A
Exposure pads Surface Total
used to area of acrylamide
represent pads found
body region (cm2) (rog)
shoulder pads (4)
back pads (2)
chest pads (2)
forearm pads (4)
400
200
200
400
hand rinses (8) N/A
Total Dermal
0.267
0.050
0.020
2.18
0.445
Contact
Time
of
exposure
(h)
3.0
3.0
3.0
3.0
7.7
Estimate
Dermal
contact
estimate6
(mg/h)
0.24
0.01
0.004
2.3
0.058
= 2.6
"Surface area of body regions based on the anatomic dimensions of the  50th
 percentile man from Popendorf  (1976, 1982), Diffrient  et  al.  (1974),  and NAi
 (1962).                         A    x S,
bDermal contact estimate (mg/h)  = TOT	
                                  T x S2

where ATOT = total acrylamide found  (mg)

     S1  = surface area of body region (cm2)
     S2  = surface area of pads (cm2)
     T  = time of exposure (h).
                                      9-41
-------
                       Table 9-12  Dermal Contact Estimate
Site no. 3
Mainline Operation
Body region
Surface
area of
region8
 (cm2)
Exposure pads
  used to
  represent
 body region
Surface
area of
 pads
 (cm2)
  Total     Time     Dermal
acrylamide   of      contact
  found    exposure estimate6
   (mg)       (h)      (mg/h)
Grout Foreman

Head             1,090
Back of neck       115
Front of neck      115
Forearms         1,290

Hands             N/A
          shoulder pads  (2)   200      0.241      9.3
          back pad (1)        100      0.081      9.3
          chest pad  (1)       100      0.123      9.3
          forearm pads (2)    200      1.26       9.3

          hand rinses  (6)  .   N/A      7.40      10.2
                                                  0.14
                                                  0.01
                                                  0.02
                                                  0.87

                                                  0.73
                                         Total  Dermal  Contact Estimate = 1.8
Head             1,090
Back of neck       115
Front of neck      115
Forearms         1,290

Hands             N/A
 shoulder pads (2)   200     0.158     9.3
 back pad (1)        100     0.067     9.3
 chest pad (1)       100     0.099     9.3
 forearm pads (2)    200     0.086     9.3

 hand rinses  (6)      N/A     4.54     10.2
                                                           0.09
                                                           0.008
                                                           0.01
                                                           0.06

                                                           0.44
                                         Total Dermal  Contact  Estimate = 0.61
8Surface area of body regions based on anatomic dimensions of the 50th
 percentile man from Popendorf  (1976, 1982), Diffrient et al.  (1974),  and NASA
 (1962).                         A    x S,
bDermal contact estimate (mg/h)  = TOT	
                                  T x S2

where ATOT = total acrylamide found (mg)

     S, = surface are'a of body region (cm2)
     S2 = surface area of pads (cm2)
     T  = time of exposure (h).
                                      9-42
-------
                      Table  9-13   Dermal  Contact Estimate
Site no. 4
Lateral Line Operation
Utility Worker
Body region
Surface
area of
region*
(cm2)
Exposure pads
used to
represent
body region
Surface
area of
pads
(cm2)
Total
acrylamide
found
(mg)
Time
of
exposure
(h)
Dermal
contact
estimate6
(mg/h)
Head             1,090
Back of neck       115
Front of neck      115
Back             1,540
Chest            1,540
Upper arms       3,170
 & shoulders
Forearms         1,290
Thighs & hips    5,210
Lower legs &     3,820
 feet

Hands             N/A
shoulder pads  (2)    200      0.018      8.2
back pad  (1)         100      0.008      8.2
chest pad  (1)        100      0.006      8.2
back pad  (1)         100      0.008      8.2
Chest pad  (1)        100      0.006      8.2
shld. & forearm      400      0.062      8.2
 pads (4)
forearm pads  (2)"    200      0.044      8.2
•thigh pads  (2)       200      0.112      8.2
shin pads  (2)        200      0.131      8.2
hand rinses  (6)     N/A      0.494      8.2
0.01
0.001
0.001
0.02
0.01
0.06

0.03
0.36
0.30
0.06
                                         Total Dermal Contact Estimate = 0.85
"Surface area of body regions based on the anatomic dimensions of the 50th
 percentile man from Popendorf (1976, 1982), Diffrient et al.  (1974), and NASA
 (1962).                         A    X S1
bDermal contact estimate (mg/h)  = TOT	
                                  T x S,
where ATOT = total acrylamide found (mg)

     S,  = surface area of body region (cm2)
     S2  = surface area of pads (cm2)
     T  = time of exposure (h).
                                      9-43
-------
Based on previous observational site visits  (which included still


and video recording), EPA expected that manhole repair had


potentially the highest dermal exposure.  The results of the


monitoring study confirm this hypothesis.  The two manhole sites


had total dermal contact estimates of 5.0 and 2.6 mg/hr, whereas


the mainline and lateral line repair sites were 1.8, 0.61 and


0.85 mg/hr, respectively.  In addition, one worker had the


classic symptoms of aerylamide-induced peripheral neurotoxicity


such as skin peeling.  Two other workers indicated that they had


experienced peeling skin on the hands.





9.4.3.  Potable Water Consumption


     Polyacrylamide products are used as flocculants to clarify
                      \

raw water in potable water plants.  Because a low level of


residual acrylamide monomer is present in the polymer, treated  " .


water may contain a certain amount of acrylamide.  How much


depends on polymer use levels, residual monomer levels, and the


removal, if any, of acrylamide in drinking water treatment.


     Based on data supplied by the Synthetic Organic Chemical


Manufacturers Association (SOCMA)  (1986), typical monomer


concentrations in drinking water are about 0.01 ppb.  This is


derived from 0.2 ppm (typical polymer use level)  X 50 ppm


(typical residual monomer in polymer)  = 0.01 ppb.  Assuming the


consumption of two liters of water per day,  monomer consumption


would be 2.8 X 10'7 mg/kg/day for a 70 kg person.  The maximum


concentration is 0.5 ppb, assuming the maximum permitted




                              9-44
-------
         Table 9-14.  Acrylamide Concentrations in Sugar


Polymer Cone.
Type of Sugar
Granulated
Soft sugars
Molasses
Cane/beet juice
Cane/beet liquor
Corn sweeteners
ppm
4
4
4
3-4
10
8

Monomer Cone.
ppm
50
50
50
50
50
50
Resultant
Monomer Cone.*
ppb
0.012**
0.24-0.48**
42.4**
0.15-0.2***
0.5***
0.4***
*   Assumes no degradation during sugar processing or losses
    during clarification/filtration.
**  Assumes monomer partitions with water.
*** Assumes monomer partitions with polymer.
Table from SOCMA (1986).
9.4.5.  Other                                              i
     No data are available on the extent and magnitude of
acrylamide exposure for persons using the various polyacrylamide
products discussed in section 9.2.2.   However,  based on the
nature of the uses of these products, the fact  that acrylamide is
present as an impurity, they are added to process streams either
remotely or infrequently by bucket,  and that the polymers are
added in aqueous solutions indicates that the exposure potential
would be low.  The laboratory use of acrylamide for gel
electrophoresis presents the potential for significant dermal
exposure whenever gloves are not worn in handling the monomer or
gel.
                              9-46
-------
                       10.   RISK ESTIMATION







     The determination of individual risk can be made through the



use of epidemiologic or animal studies.  Epidemiologic studies



suitable for low-dose quantitative risk extrapolations are rarely



available, and are not available in the case of acrylamide.  For



acrylamide, cancer, neurotoxic, and reproductive risks were



estimated through the use of animal data.  This necessitates



extrapolation from high to low doses because, typically, test



animals are exposed to concentrations higher than those expected



to be experienced by humans.  These extrapolations are carried



out by fitting mathematical models to the observed animal data



for cancer risk estimation and by calculating margins of exposure



from NOEL and LOEL data for noncancer health hazards.      i







10.1.  Noncancer Health Risks



     The hazard section of the risk assessment identified four



noncancer effects of acrylamide:  neurotoxic, reproductive,



developmental, and genotoxic effects.  The data bases for these



effects are comprised primarily of animal data; only in the case



of neurotoxic effects are human data available.  The strength of



the associations vary considerably, with neurotoxicity being the



strongest and developmental effects the weakest.  To evaluate the



risk potential for acrylamide's noncancer effects, each effect is



examined here in turn and where possible, risk estimates are



generated.  The traditional method of assessing the risk





                               10-1
-------
presented by noncarcinogens is to calculate an "acceptable daily
intake" or ADI.  The ADI is calculated by applying "safety
factors" in multiples of ten to either NOEL'S or LOEL's derived
mainly from animal studies.  For instance, if the NOEL in a 90-
day feeding study were 5 mg/kg/d, the ADI would be 0.05 mg/kg/d.
This is derived by dividing the NOEL by a safety factor of 100
(10 for less than a chronic study and an additional 10 to account
for animal to human extrapolation).   This would be considered a
"safe" level of exposure.  This approach is an assumption that
there is a threshold dose below which no toxic effects occur.  If
there are any residual risks at the  ADI, they are at most trivial
and of no public health significance (Rodricks and Taylor, 1983).
     Unfortunately, the term "safety factor" suggests, perhaps
inadvertently, the notion of absolute safety,  i.e., absence^, of
risk.  While there is a conceptual basis for believing in the-
existence of a threshold and "absolute safety" associated with
certain chemicals, in the majority of cases a firm experimental
basis for this notion does not exist.
     As a consequence of the use of  the term "safety factor", the
ADI is viewed by many as an "acceptable" level of exposure,  and,
by inference, any exposure greater than the ADI is seen as
"unacceptable".  This strict demarcation between what is
"acceptable" and what is "unacceptable" is contrary to the views
of most toxicologists, who typically interpret the ADI as a
relatively crude estimate of a level of chronic exposure which is
not likely to result in adverse effects to humans.  The ADI is

                               10-2
-------
generally viewed as a "soft" estimate, whose bounds of
uncertainty can span an order of magnitude.  That is, within
reasonable limits, while exposures somewhat higher than the ADI
are associated with increased probability of adverse effects,
that probability is not a certainty.  Similarly, while the ADI is
seen as a level at which the probability of adverse effects is
low, the absence of risk to all people cannot be assured at this
level.
     In response to the conceptual problems associated with ADIs
and safety factors, the concept of the "reference doses (RfD)"
was developed (EPA, 1986).  The RfD is a benchmark dose
operationally derived by consistent application of generally
order of magnitude uncertainty factors (UFs) that reflect various
                  \
types of data used to estimate RfDs (for example, a valid chronic
human NOAEL normally is divided by an UF of 10-fold) and an  -
additional modifying factor (MF),  which is based on a
professional judgement of the entire data base of the chemical.
The RfD is determined by use of the following equation:
                        RfD = NOAEL/(UF X MF)
In general, the RfD is an estimate (with uncertainty spanning
perhaps an order of magnitude or greater)  of a daily exposure to
the human population (including sensitive subpopulations)  that is
likely to be without an appreciable risk of deleterious effects
during a lifetime.  The RfD is appropriately expressed in units
of mg/kg/day.


     The RfD is useful as a reference point for gauging the
                              10-3
-------
potential effects of other doses.  Usually, doses which are less
than the RfD are not likely to be of regulatory concern.
However, as the frequency of exposures exceeding the RfD
increases, and as the size of the excess increases, the
probability increases that adverse effects may be observed in a
human population.
     In the risk characterization process, a comparison can be
made between the RfD if one has been established and the
estimated (calculated or measured)  exposure dose, which should
consider exposure by all sources and routes of exposure.
     An alternative measure that is useful and which will be used
here in addition to an RfD comparison is the "margin of exposure
(MOE)", which is the magnitude by which the NOAEL of the critical
                 •
toxic effect exceeds the estimated exposure dose where both are
                                                           •\
expressed in the same units:
     MOE = NOAEL (experimental dose) / (human dose)
     For instance, if a NOEL of 0.3 mg/kg/day were used and the
human exposure was predicted or measured at 0.1 mg/kg/day, then
the MOE would be 3.  The lower the MOE the higher the concern.
However, in each case, consideration of the MOE derived must be
tempered by consideration of the quality of the data base used.

10.1.1.  Neurotoxicity
     While there are a variety of effects on the nervous system
from chronic acrylamide exposures,  those that are most serious,
widely studied, and appear most prominent are the effects on
                               10-4
-------
motor systems in peripheral nerves and the spinal cord.  Because
the human data are primarily limited to case reports of
unquantified exposure levels, a quantitative risk assessment for
neurotoxicity must rely on the best available animal data.  In
general, the available studies reveal no peculiarities in the
sensitivity of exposed animal groups, although most studies are
limited by design and group size in their ability to elucidate
such issues.  Most of the animal species tested for these effects
showed a sensitivity in the same general exposure range, which
provides some confidence in making interspecies'extrapolations.
     The best animal studies for evaluating chronic acrylamide
exposures are summarized in Table 5-2 of section 5.3.3.  The
lowest observed effect levels (LOEL's) for three species (rats,
cats, and monkeys) are reported as 1 mg/kg/day.  For cats and
rats, effects at this level were seen after only 3 to 4 months of
exposure.  Thus, 90 days or more of exposure to 1 mg/kg/day is
associated with neuropathological effects.  This is generally
consistent with the one quantified human exposure report (Igisu
et al., 1975), where dramatic effects were observed at about 10
mg/kg/day after less than 1 month of exposure.
     The no observed effect levels (NOEL's) range from 0.2 to 2.0
mg/kg/day from the best available data.  For each species tested,
however, the NOEL's are slightly different.  For rats, 0.2
mg/kg/day for 3 months is the best NOEL (from electron microscopy
study).  Using light microscopy, however, a 2-year NOEL of 0.5
mg/kg/day was observed.  Although a 2-year study, it was not

                              10-5
-------
judged the best because the NOEL was found with light microscopy,



a fairly insensitive measure of structural integrity.  For cats



the NOEL is 0.3 mg/kg/day for 1 year and for monkeys the NOEL is



less than 1.0 mg/kg/day for 1.5 years.



     Margins of exposure (MOE) were calculated based on the



lowest, best NOEL of 0.2 mg/kg/day from the Burek et al. (1980)



study  (see Table 10-1).  A similar NOEL of 0.5 mg/kg/day was



observed in the Johnson et al. (1986)  rat study.  The MOE's are



for chronic workplace or dietary exposure.



     As is evident from Table 10-1, soil grouting workers have



very low MOE's.  This may account for the observation of overt



signs of acrylamide neurotoxicity in one of the nine workers



interviewed during the EPA-sponsored field study (See section



9.5.2.).  Two other workers indicated that they had experienced



peeling skin on the hands.   Even more important is the fact that



the MOE's using the lowest LOEL of 1.0 mg/kg/day, which has been



reported for rats, cats, and monkeys,  range from 6 to 43.  For



comparison, an RfD based on a LOEL would be derived by dividing



the LOEL by an uncertainty factor of 1,000.  The oral RfD for
           «


chronic acrylamide exposure is 0.0002 mg/kg/day which is



significantly below the exposures reported for soil grouting



workers.



     The MOE's for the manufacturing/processing groups range from



29 to 182.  Drinking water MOE's are greater than 1,000.;
                               10-6
-------
TABLE 10-1.  MARGINS OF EXPOSURE  (MOE)  FOR CHRONIC ACRYLAMIDE
             EXPOSURE — NEUROTOXICITY
Nalco8
                       Air Cone,  in
Exposure
Category/
Company
OSHA Std.
Cyanamid8
mg/m — 8 hr TWA
Number Dermal Exp. Exposure in MOE
Exposed in mcr/hr* mcr/ka/d of
N/A 0.03
57 monomer prod
polymer prod
ma/m3rskin) 4.3 X 10-2
. 0.045 mg/m3 6.4 X 10"3
. 0.060 mg/m3 8.6 X 10"3
for NOEL
0.2 ma/kg
5
31
23
2 6  monomer/
    polymer prod.  0.005  mg/m
7.1 X 10"
                                              282
Dow8
55  monomer prod. 0.049 mg/m3   7.0 X 10"3       29

    polymer prod. 0.007 mg/m3   1.0 X 10~3     200
Soil       1800
 Grouting8
          0.005 mg/m
7.1 X 10"
                                              282-
Soil                  0.12 g/nr
 Grouting8  site 1    	
                      1.25 mg/hr
                                1.6 X 10
                                                    -1
Soil                  0.01 mg/m
 Grouting   site 2    	
                      0.65 mg/hr
                                7.6  X 10
                                        -2
                               10-7
-------
Exposure
Category/
Company

Soil

 Grouting
            Number
            Exposed
             site 3a
TABLE 10-1 — CONTINUED

  Air Cone, in
 mg/m—8 hr TWA
   Dermal Exp.     Exposure in  MOE for NOEL
   in mg/hr*

   0.06 mg/m3
                        0.45 mg/hr
Bia/ka/d
of 0.2 ma/kg
                     6.0 X 10
                                                  -2
Soil
                                 »•
 Grouting    site 3b    0.04 mg/m
                        0.15 mg/hr
                                          2.3 X 10
                                                  -2
Soil
                 »
 Grouting    site '4
                        0.008 mg/m
                        0.21 mg/hr
                                          2.5 X 10
                                                  -2
                                      8
                                      ^
Drinking

 Water0
             3 X 10

             max.  possible
                  0.5 ppb maximum**  >1,000

                  1.4 X  10~	   	
                                       0.01 ppb typical  »1,000
                                       2.8 X 10"7
*  For inhalation and ingestion 100% absorption is assumed.  For
   dermal absorption, the mg/hr estimates in Tables 9-10 to 9-
   13 were multiplied by 0.25 to account for the difference in
   absorption by this route.
** Concentration in water treatment plant.

8 NIOSH (1985).
b EPA (1987).                                             J
c SOCMA (1986).
                               10-8
-------
10.1.2.  Reproductive Effects
     The reproductive effects of acrylamide have been observed in
several animal studies  (Section 6).  A dominant lethal effect and
decreased copulatory performance have been observed in rats.  In
one assay, rat testosterone levels were depressed, but the
functional significance of this finding is uncertain.  In the
mouse, decreased fertility was observed following exposure of
male mice to acrylamide and an increased resorption rate resulted
from the exposure of male or female mice.  Degeneration of
testicular epithelial tissue and a dominant lethal effect have
also been observed in other mouse studies.
     The available data indicate that acrylamide can act directly
on the reproductive system, rather than indirectly through stress
or other systemic effects such as neurotoxicity.  Therefore"1,
concern for the reproductive toxicity potential is heightened'.
In other data reviewed in the rat and the mouse, neurotoxicity
occurred in each case, while reproductive toxicity was sometimes
absent.  It appears then, that while acrylamide can exert a
direct effect on the testes, the weight of evidence for effect
levels and probability of occurrence for acrylamide induced
neurotoxicity is much higher.
     The LOEL and NOEL for reproductive effects, as indicated by
the available data (Table 6-2 in section 6.1.), are 8.8 and 4.2
nig/kg/day, respectively.  As Table 10-2 indicates, all groups
have MOE's an order of magnitude or more greater than those for
neurotoxicity (chronic exposure).

                              10-9
-------
TABLE 10-2.  MARGINS OF EXPOSURE  (MOE) FOR ACRYLAMIDE  BASED
             REPRODUCTIVE EFFECTS
Exposure
Category/   Number
Company     Exposed
Soil
 Grouting8
              Air Cone,  in
             mg/m—8 hr TWA
               Dermal Exp.     Exposure in  MOE for NOEL
               in mcr/hr*         mg/ka/d    of 4.2 ma/kg
OSHA Std.
Cyanamid8

Nalco8
Dow8

N/A 0.03
57 monomer prod.
polymer prod.
2 6 monomer/
polymer prod.
55 monomer prod
polymer prod.
«
ma/m3fskin)
0.045 mg/m3
0.060 mg/m3
. 0.005 mg/m3
. 0.049 mg/m3
0.007. mg/m3
4.3 X 10*2
6.4 X 10~3
8.6 X 10"3
7.1 X 10"4
7.0 X 10"3
1.0 X 10"3
98
656
488
595
600
4,200
3
18
13
160
16
112
1800
0.005 mg/m
7.1 X 10"
5,915  160
Soil
Grouting6 site 1
Soil
Grouting site 2
0.12 g/m3
1.25 mg/hr
0.01 mg/m3
0.65 mg/hr '
1.6 X 10-1 26 0.7
7.6 X 10-2 55 2
                                   10-10
-------
Exposure
Category/
Company

Soil
 Grouting
Number
Exposed
 site 3a
                          TABLE 10-2 — CONTINUED
 Air Cone. in
mg/m—8 hr TWA
  Dermal Exp.
  in mcr/hr*

  0.06 mg/m3

  0.45 mg/hr
Exposure in  MOE for NOEL
  mg/kg/d    of 4.2 mg/kg
 6.0 X 10
                                          -2
70
Soil
 Grouting
 site 3b
  0.04 mg/m3

  0.15 mg/hr
                                              2.3 X 10
                                                      -2
                                                 183
Soil 0.008 mg/m3
0.21 mg/hr.
Drinking 3 X 107
Water6 max. possible
2C Y 1 n "^ 1 fn
• D A J.U J.OO
0.5 ppb maximum** >lf
1.4 X 10-5 >100.


000
000
                                             0.01 ppb typical  »1,000
                                             2.8 X 10"7        >100,000
*  For inhalation and ingestion 100% absorption is assumed.  For
   dermal absorption, the mg/hr estimates in Tables 9-10 to 9-13
   were multiplied by 0.25 to account for the difference in
   absorption by this route.
** Concentration in water treatment plant.
8 NIOSH (1985)
b EPA (1987)
e SOCMA (1986)
                                   10-11
-------
10.1.3.  Developmental Effects
     Acrylamide has been shown to produce developmental and
postnatal effects in mouse and rat offspring following
administration to pregnant dams.  On a physiological level, there
is evidence that acrylamide produces neurotoxic effects (tibial
and optic nerve degeneration) in the neonates at levels that are
not toxic to the dam.  In addition, on a biochemical level,
acrylamide causes changes in dopamine levels and intestinal
enzyme levels in the fetus at dose levels where no maternal
toxicity is apparent, but the toxicological significance of these
biochemical changes is not clear from the available data.   There
is a basis for concern for the conceptus following maternal
exposure during gestation since these data show that acrylamide
can have a direct effect on the conceptus.  However, the specific
doses at which neurotoxic effects were observed in rat offspring
was not reported in the study.  Consequently,  numerical risk
estimates cannot be calculated for these neurotoxic effects.  For
the biochemical effects, the data indicate a lowest observed
effects level (LOEL) of 20 mg/kg/day.  Since this was the  lowest
dose level tested in the study, a no observed effect level (NOEL)
was not established (Table 6-1 in section 6.1).  However,  because
the toxicological significance of these biochemical changes is
not apparent, risk of functional deficits cannot be estimated.
                              10-12
-------
 10.1.4.  Genotoxic  Effects
     The genotoxicity  data on acrylamide provide  a  basis  to
 support the carcinogenicity hazard identification and to  support
 heritable mutation  as  a  separate  endpoint of concern for
 acrylamide exposures.
     The guidelines for  mutagenicity risk assessment address
 evaluation of the potential genetic risk associated with  human
 exposure to chemicals  with effects such  as those  of acrylamide.
 The body of evidence suggests that acrylamide may induce
 alterations in the  genome  of germinal cells.  These data  provide
 a strong weight of  evidence bearing on the potential for  human
 germ-cell mutagenicity and its heritability.  The strength of a
 human germ-cell mutagenicity concern would only be  further
 strengthened by direct human evidence (the highest  level  of*
 evidence for human  mutagenicity).
     The existing mutagenicity testing results for  acrylamide
 demonstrate that it induces chromosome breaks but not point
 mutations.  In addition, germ cell cytogenetics and in vivo germ
 cell test results in males of two  species  adequately demonstrate
 that acrylamide can reach  cellular targets in mammalian germ
 cells and produce chromosomal aberrations  that can  be transmitted
 to the next reproductive generation.   However, the  data do not
 allow the development  of reliable  quantitative estimates  of
jheritable risk.  Therefore,  the remainder  of  this section will
 simply illustrate ways to  estimate the potential  magnitude of
 such risks.  Undue  emphasis should not be  placed  on the

                               10-13
-------
quantitative risk estimates of that risk.  The ensuing discussion
will examine the effects of acrylamide in inducing dominant
lethal effects which are noted in the offspring of exposed
parents and chromosomal translocations, a proportion which have
deleterious heritable consequences.
     A quantitative estimate of risk for a biological effect is
dependent upon an estimate of the shape of the dose-response
curve for the effect under review and the magnitude and nature of
anticipated exposure.  In combination these two considerations
give a measure of the incidence or frequency of the given
endpoint.  In quantitating risk from some heritable effects a
second factor must be considered; that is, the proportion of
mutations that may, be manifest as adverse health effects in the
recipients of the mutations.                               •*
     From a theoretical standpoint dominant lethal mutations may
be conceived as the result of a single break in a chromosome.
This leaves an acentric fragment which is not able to move
properly in cell division.  The resulting conceptus is unable to
survive embryonic and fetal life and dies in utero, thus the
term, dominant lethal.  If each chemical-chromosome interaction
could result in a break, then the expected dose-response curve
would be linear, at least at low doses.  In practice, however, we
do not know how many interactions are required to induce a
chromosomal break, but it seems reasonable that a linear dose-
response relationship may apply under certain circumstances and,
at least, could represent a reasonable upper bound on the risk.

                              10-14
-------
     For translocations that are passed from treated parent to
offspring, two different chromosomes would need to be broken and
their corresponding arms recombined.  Thus, theoretically at
least two independent chemical-chromosome interactions would be
necessary.  Such a requirement would suggest that dose-response
curves would be curvilinear upward.
     Evaluation of the existing data on acrylamide does not allow
one to make definitive judgements about the shape of the dose-
response curve.  First, the molecular mechanisms by which
acrylamide induces chromosome breaks are essentially unknown.
Second, because the doses in the mouse heritable translocation
test are so close together (40 & 50 mg/kg/d)  it does not allow
one to characterize the underlying dose-response relationship.
Third, there are gaps in the data base for all the germ cell
tests conducted on acrylamide.  These include a lack of low and
multiple dose testing, inability to discern dose-rate
interrelationships, lack of reproductive performance data on
individual treated males,  and lack of standardized protocols.
These shortcomings limit the ability to characterize the
heritable risks for this agent.   Thus, instead of an in-depth
analysis of the risk, only rough estimates will be developed as
an illustration of how one may proceed to conduct these analyses.
     It is evident from the mouse germ cell data that the
mutagenic response to acrylamide exposure differs significantly
as a function of the germ cell stage.  Like a number of other
chemical substances that produce chromosomal  breakage in male

                              10-15
-------
germ cells (e.g., ethylene oxide, ethylmethanesulfonate), it
appears that acrylamide exerts its major effect upon post-meiotic
stages of sperm development (i.e., late spermatid and sperm).
Thus, it is the germ cells in about the last two weeks of their
development that are at risk for mutation.  This implies that
although there may be some risk of mutation shortly after
exposure to acrylamide, that risk may decrease greatly following
longer intervals of time as the cells at risk fully mature and
disappear from the male reproductive tract.
     In an attempt to make a rough estimate of potential risk
from exposure to acrylamide, a number of assumptions were made
which will be applied to both dominant lethal effects and
heritable translocations.
a)   Mouse germ cells are at risk during the last two weeks*of
     development.
b)   The post-meiotic sensitive period is the same length in
     humans as in the mouse.
c)   The mutagenic response of humans and animals is comparable
     on a body weight basis.
d)   The relevant exposure is the accumulated total received over
     a 2-week period.
Due to the shortcomings in the design and reporting of the data
sets on acrylamide, only point estimates of potential mutagenic
risk are presented (to the nearest order of magnitude);
confidence intervals on the models are not included.
                              10-16
-------
     Four dominant lethal studies were examined:  two in mice and
two in rats.  Each of the mouse studies  (Shelby et al., 1986;
1987) involved a single dose level of acrylamide and a control.
The two rat studies (Smith et al., 1986; Nalco, 1987) used
several dosages.  A simple linear extrapolation was performed
from the lowest dose showing an effect to the origin.  When
dominant lethals were recorded as  function of post-treatment
time (Shelby et al., 1986), the average proportion of dominant
lethals over a two-week period (approximate) was used for the
linear extrapolation.  Exposure to treated animals was assumed to
be the total received over a two-week period (e.g., 5 mg/kg/d x
14d = 70 mg/kg in drinking water (Nalco, 1987)).  Slope factors
(risk per unit exposure) were calculated for each study; a
similar estimate was found for all studies (0.001 (mg/kg) ~1J*>
except the Smith et al. study which was about an order of
magnitude lower.
     Estimates of dominant lethal risk were made, assuming that
exposures were equivalent to the soil grouting workers (table 10-
5) that had been monitored.  If daily dermal and inhalation
exposures range from about 0.02 to 0.16 mg/kg,  then two work
weeks of exposure would vary between about 0.20 and 1.6 mg/kg.
Thus, accumulated excess risk that might accrue over a 2 week
period following exposure might then range from 10"4 to 10~3,
respectively.   :
     An analogous calculation can be made using the mouse
heritable translocation data, assuming a linear dose response as
                              10-17
-------
suggested by Bishop and Kodell (1980).  A slope was calculated
using the 200 mg/kg total exposure (40 mg/kg/d for 5d); it was
the same as that calculated for the dominant lethal studies
(0.001 (mg/kg)"1).  Thus, extrapolation to the level of exposure
reported for the human grouting workers leads to risks in the
range of 10~4 to 10~3.  Recognizing that heritable translocations
may not behave in a linear fashion with dose, other models were
applied to all dose groups in the mouse study to extrapolate to
the human exposure situation.  Using a multistage model with two
stages, point estimates of the risks were in the range of 10"7 to
10"5  .  A third mathematical construct, the quadratic, assumes
two chemical-cell interactions for. an effect; this model
predicted point estimates of the risks as 10~7 to 10"5  .  Given
the theoretical requirement of at least two breaks to form 
-------
translocations show about 25 percent changes relative to non-
carriers:  decreases in live births, increases in fetal deaths,
and reductions in reproductive fitness  (Morton et al., 1975;
Jacobs et al., 1975).  These effects in humans are like those
noted in mice with translocations (Generoso et al., 1980).
     The present estimates of heritable chromosomal risk indicate
that acrylamide exposures associated with a certain occupational
setting (i.e., soil grouting) may be associated with dominant
lethal effects at risk levels about two orders of magnitude below
estimated lifetime excess cancer risks.  (See section 10.2. Table
10-4.)  The corresponding risk for heritable translocations may
be as high as the dominant lethal risks but may be orders of
magnitude lower tft,an that.  Even though these risk estimates for
heritable effects are only preliminary, they should not be ^
totally ignored.

10.2.  Quantitative Cancer Risk Assessment
     Since risks at low exposure levels cannot be measured
directly either by experiments in animals or by epidemiologic
studies, a number of mathematical models have been developed to
extrapolate from high to low doses.   The Office of Science and
Technology Policy (OSTP) published principles on model selection
which states that "No single mathematical procedure is recognized
as the most appropriate for low-dose extrapolation in
carcinogenesis.  When relevant biological evidence on mechanism
of action exists, the models or procedures employed should be

                              10-19
-------
consistent with the evidence.  When data and information are
limited, however, and when much uncertainty exists regarding the
mechanism of carcinogenic action, models or procedures which
incorporate low-dose linearity are preferred when compatible with
the limited information."  Data relevant to selecting a model for
cancer risk extrapolation associated with exposure to acrylamide
were reviewed; much of the biological information supports a
direct relationship between exposure and carcinogenicity.  (See
section 10.2.2.)  However, other mechanistic data are lacking.
Thus little information was available to propose a non-linear
extrapolation model to estimate the risks of acrylamide.
Therefore, in keeping with the OSTP guidance and EPA's Guidelines
for Carcinogen Risk Assessment, the assessment employed a linear
model (i.e., linearized multistage procedure).             •*•
     Data from the Johnson et al (1986)  study were used to
estimate risk from acrylamide exposure.   The EPA guidelines for
cancer risk assessment recommend pooling tumor incidence data for
purposes of risk assessment, since risk numbers derived from
site-specific tumor incidence data may not be predictive of (and
may in fact underestimate) "whole-body" risks that are determined
using the pooled individual animal data.  The dose-response
curves for each sex based on the pooled tumor incidence (benign
and malignant) data compromise the data sets of choice for risk
assessment.  The most sensitive sex will be chosen to represent
possible human risk.  Further, a second pool was created for each
sex to allow consideration of the contribution of malignancies

                              10-20
-------
alone to the overall risk estimates.  Table 10-3 summarizes the



pooled tumor incidence data.  The dose levels used in the



extrapolation; 0, 0.01, 0.1, 0.5, and 2 mg/kg/day, had been



previously derived by adjustment for varying water consumptions



across time in the study and across dose levels.



     Tumors at a particular site were added into the pool only



when the tumor site had statistically significantly increased



incidence at least at the high dose level (treated vs. control).
                                       13


The first pool contains males having tumors of the testes,



thyroid, or adrenal gland.  For the adrenal gland, only the



(benign) adenomas were considered since neither malignant alone,



nor adenoma and adenocarcinoma combined, were significantly



elevated.  The second male pool contains testes mesothelioma



only.  The third pool contains females having tumors of the*.



thyroid (combined), mammary gland (combined),  CNS (tumors only,



no "proliferations"), uterus (adenocarcinoma), and oral



(combined).  The fourth pool contains females having malignant



tumors of the thyroid, mammary gland, CNS,  and uterus.  Tumors of



the clitoral gland in females are not included in the pooled



tumor incidence data because only a very low number of tissues



were examined in each dose group.



     A trans-species conversion for risk was carried out.  For



example, if male humans weigh approximately 70,000g and female



rats weigh approximately 200g then,  by the method of Mantel and



Schneiderman, whereby body surface area is a surrogate for
                              10-21
-------
*Changes to be made in final R.A. and brought out in public
comments (See Response 4.)

Table 10-3.  Animal Test Data Sets from Dow Acrylamide Study Used
            for Extrapolation.

      Male Rats;   Number  of  Animals with Tumors  —  Testes.
                      Thyroid,  and Adrenal

Dose fmg/kg/day)          Number Responding        Number at Risk

      0.0                       7                        59
      0.01                      8                        52
      0.1                       13                       57
      0.5              ,14                       57
      2.0                       22                       54

  Male Rats:  Number of Animals with Malignant Tumors — Testes

Dose fmq/kg/day)          Number Responding        Number at Risk

      0.0                        3                       57
      0.01                       0                       49
      0.1                        7                       57
      0.5        •               11                      53
      2.0         '               10                      52
                                                           •*.
     Female Rats;   Number of Animals with Tumors — Thyroid-
                  Mammary. CNS.  Oral, and Uterus

Dose fmg/kg/day}          Number Responding        Number at Risk

      0.0                        13                      60
      0.01                       18                      60
      0.1                        14                      60
      0.5                        21                      60
      2.0                        44                      60

     Female Rats:   Number  of  Animals with Malignant  Tumors  —
                Mammary.  Thyroid. CNS.  and Uterus

Dose (mg/kg/day)          Number Responding        Number at Risk

      0.0                        5                       60
      0.01                       5                       60
      0.1                        3            :           60      -
      0.5                        2                       60
      2.0                        20                      60
                              10-22
-------
                 TABLE 10-4.  SUMMARY OF ACRYLAMIDE  CANCER RISK ESTIMATES
                          Air Cone,  in
Exposure
Category/
Company
OSHA
Std.
Cyanamid"
Nalco"
Dow*
rag/nr — 8 hr TWA
Number Dermal Exp.
Exposed in ma/hr*
N/A 0.03 mg/m3 (skin)
57 monomer prod. 0.045
polymer prod. 0.060
26 monomer/
polymer prod. 0.005
55 monomer prod. 0.049
polymer prod. 0.007
Cone. In
PPB
,. t
mg/m3
mg/m3
mg/m3
i
mg/m3
mg/m3
LADE**
mq/ka/day
1.7 X 10"2
2°. 4
3.2
2.7
2.6
3.8
X
X
X
X
X
lO'3
lO'3
10"4
lO'3
10"4
Cancer Risk
MLE ( upper-bound)
2 X 10"
3X10"3
4X10"3
3X10"4
3X10"3
4X10"4
2 (8 X 10"2 )
(1X10"2 )
(1X10"2 )
(1X10"3 )
(1X10"2 )
(2X10"3 )
Soil         1800     0.005 mg/m3
  Grouting8
2.7 X 10"4     3X10"4 (1X10"3  )
-------
                                  TABLE  10-4  — CONTINUED
Exposure
  Air Cone, in
mg/m—8  hr TWA
Category/ Number
Company Exposed
Soil
s**»s%«t+- 4 ***•»" M 4 ^A 1
G routing site i
Soil
Grouting . ...site 2
Soil
Grouting site ja
Soil
\j routing site JD
Soil
fvrtii^ ^ w»*r e^^o A
Dermal Exp.
in ma/hr*
0.12 mg/m3
1.25 mg/hr
0.01 mg/m3
0.65 mg/hr
0.06 mg/m3
0.45 mg/hr
0.04 mg/m3
0.15 mg/hr
0.008 mg/m3
Cone. In LADE** Cancer Risk
PPB ma/kcr/dav MLE (uDDer-bound)
6.3 X 10"2 8X10"2 (3X10"1 )
t
3.0 X 10'2 4X10"2 (1X10"1 )

2.4 X 10"2 3X10'2 (1X10"1 )
f
9.0 X 10"3 1X10"2 (4X10"2 )

9.8 X 10'3 1X10"2 (4X10*2 )
                          0.21 mg/hr
-------
Unit Risk
                                  TABLE  10-4  —  CONTINUED
Exposure
Category/
Company
Drinking
Waterf
Number
Exposed
Air Cone, in
mg/m3 — 8 hr TWA
Dermal Exp.
in mq/hr*
3 X 107
max. possible
Cone. In
PPB
0.5
O.Qlr
LADE**
mq/kg/day
1.4 X 10'5
2.8 X 10"7
Cancer Risk
MLE ( upper-bound)
4X10"5 (6X10'5 )
9X10"7 (1X10*6 )
1.0 X 10
                                                              -3
(4.5X10"3 )
* For inhalation and ingestion 100% absorption is assumed.  For dermal absorption the
  mg/hr contact estimates in Tables 9-10 to 9-13 were multiplied by 0.25 to account for
  the difference in absorption by this route.
**LADE = Lifetime Average Daily Exposure = based on 40 years occupational exposure and 70
  years consumer exposure.
*** Maximum Likelihood and upper bound estimates from linearized multistage model.
    Concentration in water treatment plant.

8  NIOSH (1985).
b  EPA (1987).
c  SOCMA (1986).
-------
Resultant risk estimates were multiplied by 7.05 to account for
metabolic differences between rats and man as discussed in
section 10.2.  No adjustment was made to the risk estimates due
to different routes of exposure because the available studies
show that the pharmacokinetics and tissue distribution of
acrylamide were not significantly affected by the dose
administered, by the route of administration, or by giving the
chemical over consecutive days.  (See section 10.2.2. below for
further discussion.)  Dermal exposure estimates were adjusted
before risk estimation to reflect lower absorption (approximately
25 percent) by the dermal route.  One hundred percent absorption
was assumed for oral and inhalatio.n exposures which is supported
by the laboratory*data.  For example, the lifetime average daily
exposure (LADE) of 6.3 X 10"2 mg/kg/day for soil grouting, site
1, maintenance supervisor was calculated as follows.   The
inhalation contribution is 0.12 mg/m3.  This  level  was multiplied
by (1.25 m3/hr x 8 hours)  then  divided by 70  kg giving 0.017
mg/kg.  The contribution by dermal exposure was 5.0 mg/hour.
This is multiplied by 8 hours - 70 kg and by 0.25 (the absorption
factor) to get 0.143 mg/kg.  Combining by addition to get a total
exposure of 0.16 mg/kg, this is further adjusted for lifetime
exposure for use as a LADE in risk calculation by substitution
into the estimation procedure.   This adjustment calls for
multiplication by 5/7 days/week, 50/52 weeks/year and 40/70
working years/lifetime or 0.392.  Thus the LADE is 0.161 mg/kg x
0.392 = 6.3 x 10"2 mg/kg/day.

                              10-27
-------
     The highest estimated upper-bound excess risks are to



persons engaged in sewer repair or grouting operations where



these extra risks range from 10"3   to 10"1 .  Adjusting for less



than 40 hours per week and exposures other than 50 weeks per year



leaves the risk in the 10"3 to  10"2  range.  For instance,  if one



assumes that the maintenance supervisor from site 1 only grouts



20 hours per week for 4 months per year, the estimated risk is



still in the 10"2  range  (0.5 X 0.33  X 3 X 10'1  = 5 X 10"2 ).



Such assumptions reflect that municipal grouting operations do



not occur year-round and that grouting is not performed daily.



     The manufacturing/processing group has upper-bound risks in



the 10"3  to 10"2  range.   Persons  exposed to acrylamide via



drinking water have upper-bound risks of 10"6  to  10"5 .



     As discussed in section 10.2. above, two pooled tumor -data



sets were examined with the female rats all tumors being selected



to represent potential human risk.  The data set of male rats all



tumors produced upper-bound risk estimates about an order of



magnitude lower for all exposure groups.  However, two other data



sets were defined, males and females with malignant tumors, to



allow consideration of the contribution of malignancies to the



risk estimates.  For male rats, all tumors, the malignant



mesothelioma represents about 75 percent of the risk, whereas for



female rats, all tumors,  the malignancies represent about 30



percent.I                   ;
                              10-28
-------
10.2.2.  Pharroacokinetics and Acrvlamide Cancer Risk Assessment
     The information on pharmacokinetics of acrylamide indicates
that the applied dose is probably directly proportional to
internal effective dose over a range of tested dosing conditions.
Thus, human risk estimates calculated on the basis of applied
dose should stand without modification at this time.
     Acrylamide has none of the properties that suggest major
non-linearities between applied dose and internal dose.  It is
rapidly and completely absorbed by the gut.  It is quite
uniformly distributed among tissues and, being quite polar, does
not tend to accumulate in adipose tissue.  It is rapidly cleared
by metabolism.  The proportionality of tissue accumulation of
radioactive label*to applied dose holds over a range of tested
dosing rates.                                              •*.
     Although metabolism is primarily by conjugation with
glutathione, a number of metabolites are produced, and there is
some indication that mixed-function oxidase metabolism may be
involved.  One cannot rule out the possibility that, although
total metabolism is proportional to applied dose, the mix of
various pathways may change from low to high dose, which may
affect the consequent toxicity of a given amount of compound.
Current data do not suggest this problem, however.  The foregoing
points argue that the method we used for high-to-low dose
extrapolation is probably a reasonable means of analysis.
     Consequently, currently available information provides no
indication of nonlinearities between applied and internal dose

                              10-29
-------
that would substantially alter estimates of human risk.  On the
contrary, it seems internal dose ought to vary in direct
proportion to the rate of external application.

10.2.3.  Uncertainty in Cancer Risk Estimates
     In the absence of significant human data, animal data are
relied upon to estimate potential human cancer risks.  This
introduces a number of areas of uncertainty.
     First, does the finding of cancer in laboratory animals
signify a similar potential in humans?  And if so, are the risks
predicted from animal data realistic in light of the human
experience with the chemical?  Does the nature of the animal data
and its treatment*in the risk assessment introduce additional
uncertainty?  Finally, do the exposure data fairly represent the
magnitude and duration of exposure?

10.2.3.1.  Cancer Potential in Humans
     In the absence of information to the contrary, it is assumed
that the appearance of cancer in laboratory animals indicates
potential carcinogenicity in humans exposed to the substance.
Except for two epidemiologic studies, only animal data on
acrylamide carcinogenicity are available.  Although one human
study does show an excess of lung cancer in the exposed group,
this evidence was judged to be inadequate under EPA's Cancer Risk
Assessment Guidelines.  The animal data do not indicate any
unique susceptibility or mode of action in the species tested,

                              10-30
-------
consequently it must be assumed that humans can be equally
susceptible to the carcinogenic effects of acrylamide.

10.2.3.2.  Uncertainty Due to Risk Estimation
     Two of many factors can influence risk estimation; 1) the
nature of the dose-response relationship and 2)  the choice of
model for risk estimation.  Often the shape of the dose-response
relationship, such as whether it is concave or convex, can affect
the risk estimation process and increase the level of
uncertainty.  Also, the type of model chosen is important.  For
instance, models that assume a linear response at low doses are
often more conservative than tolerance distribution models.
     Because the two tumor data sets, males and females with
tumors, are only slightly concave, the difference between the
MLE's and upper-bound estimates is small.   For both the data sets
the difference is about a factor of two.  While such a small
difference does not in itself indicate that upper-bound estimates
are close to or represent the true risk, it does indicate that
the data are consistent with a linear response.
     To test the sensitivity of the linearized multistage model,
other models were used to estimate risks from acrylamide
exposure.  For this analysis a trans-species conversion factor of
5.85 was used.  The models used were the independent and additive
background probit, logit, Weibull, and gamma multi-hit.  (See
Appendix A for a discussion of these models.)   Tables 10-5 to 10-
10 present the results of this exercise.  The four data sets used

                              10-31
-------
           Table  10-5.  Males with Malignant Tumors —
                    Additive Background Models
Exposure LADE  fma/kq/dav)   Model*
                                          MLE
8.2 x 10
        -3
1.4 x 10
        -5
                            L.M.**
                            A.P.
                            A.L.
                            A.W.
                            A.G.

                            L.M.**
                            A.P.
                            A.L.
                            A.W.
                            A.G.

*  Linearized Multistage, and Additive Background Probit,  Logit,
   Weibull, Gamma Multi-hit Models.
** For comparison.
5 X
7 X
8 X
8 X
8 X
8 X
2 X
2 X
2 X
2 X
lO'3
lO'2
lO'2
lO'2
lO'2
lO'6
lO'4
ID'*
lO'4
ID'4
          UB or UCL
9
3
3
3
3
2
8
1
1
1
X
X
X
X
X
X
X
X
X
X
10
10
10
10
10
10
10
10
10
10
-3
-1
-1
-1
-1
-5
-4
-3
-3
-3
           Table 10-6.  Males with Malignant Tumors —
                  Independent Background Models
Exposure LADE mcr/kg/day

8.2 x 10"3
1.4 x 10
        -5
                          Model*

                          L.M.**
                          I.P.
                          I.L.
                          I.W.
                          I.G.

                          L.M.**
                          I.P.
                          I.L.
                          I.W.
                          I.G.
MLE
5
9
1
1
1
2
6
7
8
1
X
X
X
X
X
X
X
X
X
X
10
10
10
10
10
10
10
10
10
10
-3
-2
-1
-1
-1
-6
-4
-3
-3
-2
UB or UCL
            9 X 10
            3 X 10
            4 X 10
            4 X 10
            4 X 10

            2 X 10
            6 X 10
            4 X 10
            5 X 10
            5 X 10
       -3
       -1
       -1
       -1
       -1
       -5
       -3
       -2
       -2
       -2
*  Linearized Multistage, and Independent Background Probit,
   Logit, Weibull, Gamma Multi-hit models.
** For comparison.
                              10-32
-------
   Table 10-7.  Males with Tumors — Additive Background Models
Exposure LADE ma/kcr/dav

8.2 x 10~3
1.4 X 10
        -5
Model*

L.M.**
A.P.
A.L.
A.W.
A.G.

L.M.**
A.P.
A.L.
A.W.
A.G.
               MLE
9
9
1
1
9
2
2
2
2
2
X
X
X
X
X
X
X
X
X
X
10
10
10
10
10
10
10
10
10
10
-3
-2
-1
-1
-2
-5
-4
-4
-4
-4
           UB or UCL
                            1  X 10
                            3  X 10
                            4  X 10
                                                      4
                                                      3
                              X 10
                              X 10
                   -2
                   -1
                   -1
                   -1
                   -1
                            2  X  10
                            8  X  10
                            9  X  10
                            1  X  10
                            8  X  10
                   -5
                   -4
                   -4
                   -3
                   -4
*  Linearized Multistage,  and Additive Background Probit, Logit,
   Weibull, Gamma Multi-hit  Models.
** For comparison.
  Table  10-8.   Females with Tumors —Additive Background Model
Exposure LADE ma/ka/day

8.2 x 10"3
Model*
MLE
     **
1.4 x 10
        -5
L.M.
A.P.
A.L.
A.W.
A.G.
L.M.**
A.P.
A.L.
A.W.
A.G.
3
6
6
5
7
6
1
1
9
1
X
X
X
X
X
X
X
X
X
X
10
10
10
10
10
10
10
10
10
10
-2
-2
-2
-2
-2
-5
-4
-4
-5
-4
                         UB or UCL'
            3 X 10
                                                      1
                                                      1
                                                      1
              X 10
              X 10
              X 10
                                                      2  X 10
                                -2
                                -1
                                -1
                                -1
                                -1
                                                      4  X 10
                                                      2
                                                      2
                                                      2
                                                      3
              X 10
              X 10
              X 10
              X 10
                                -5
                                -4
                                -4
                                -4
                                -4
*  Linearized Multistage, and Additive  Background Probit, Logit,
   Weibull, Gamma Multi-hit Models.
** For comparison.
                              10-33
-------
Table 10-9.  Females with Tumors — Independent Background Models
Exposure LADE mq/kq/dav
8.2 x 10
        -3
1.4 X 10"
Model*

L.M.**
I. P.
I.L.
I.W.
I.G.

L.M.**
I. P.
I.L.
I.W.
I.G.
MLE
3
3
7
2
1
6

3
9
6
X
X
X
X
X
X
_
X
X
X
lO"2
10
10
10
10
10
0
10
10
10
-4
-3
-2
-2
-5
_
-8
-7
-8
                                                    UBorUCL
                                                      3
                                                      2
                                                      3
                                                      7
              X 10
              X 10
              X 10
              X 10
-2
-3
-2
-2
-2
                                                      5  X 10

                                                      4  X 10"5
                                                        - 0 -
                                                      3  X 10"7
                                                           -6
                                                           -7
                                                      9 X  10
                                                      5 X  10
*  Linearized Multistage, and  Independent Background Probit,
   Logit, Weibull, Gamma Multi-hit  Models.
** For comparison.
         Table 10-10.   Females with Malignant Tumors —
                    Additive Background Models
Exposure LADE mq/kq/dav   Model*
                                         MLE
8.2 x 10
        -3
1.4 x 10
        -5
                           L.M.**
                           A.P.
                           A.L.
                           A.W.
                           A.G.

                           L.M.**
                           A.P.
                           A.L.
                           A.W.
                           A.G.

*  Linearized Multistage, and Additive Background Probit,  Logit,
   Weibull, Gamma Multi-hit Models.
** For comparison.;
8
2
2
2
2
9
3
3
3
3
X
X
X
X
X
X
X
X
X
X
10
10
10
10
10
10
10
10
10
10
-3
-2
-2
-2
-2
-6
-5
-5
-5
-5
           UB or UCL
1
6
5
5
6
2
9
9
9
1
X
X
X
X
X
X
X
X
X
X
10
10
10
10
10
10
10
10
10
10
-z
-2
-2
-2
-2
-5
-5
-5
-5
-4
                              10-34
-------
with the linearized multistage procedure (section 10.2.) were
used here.  Of the eight models (independent and additive
background for each model), male rats with tumors-independent
background and female rats with malignant tumors - independent
background, failed to produce estimates because of the fit of the
data (basically how the model incorporated the background
response observed in the study).  Generally, the risk estimates
are about as high or higher than those from the linearized
multistage procedure.  Only the probit-independent background,
females with tumors, produced very low risks.

10.3.  Uncertainty Due to Exposure Estimation
     Often the exposure analysis of a risk assessment is weak and
it is,  therefore, capable of introducing a high level of   -1
uncertainty into the risk estimation process.  In the case of
acrylamide, however, the exposure assessment contains recent,
validated data.  The NIOSH study produced data for the three
domestic manufacturers of acrylamide as well as for a significant
number of the processing sites.  The confidence in these data are
increased because of the nearly identical manufacturing and
processing procedures currently in use.  In addition, the number
of workers at these facilities is reliably known.
     The grouting segment, while not as widely surveyed as the
manufacturing and processing segments, has been well
characterized.  A combination of surveillance and monitoring site
visits has provided a clear understanding of work practices and

                              10-35
-------
quantitative estimates of exposure.  As with the manufacturing/
processing segments, the exposed population is reasonably well-
known .
     The laboratory population using polyacrylamide gels is
relatively large, 100,000 to 200,000 persons, but exposure may be
minimal because of the use and handling characteristics of this
segment.  Only the downstream populations are not as well-
characterized.  However, in these instances, exposure to
acrylamide results from residual levels in polyacrylamide
products such as water treatment chemicals.  For instance, while
industry-supplied data have bounded acrylamide concentrations in
drinking water (at least in the treatment plant), the duration of
exposure and population size are not as well-known.  A similar
situation exists for sugar consumption.        .            *
     In summary, the manufacturing/processing and grouting
segments are well-characterized regarding numbers of persons
exposed, duration of exposure, and exposure levels.  These groups
also experience the highest cancer and noncancer risks.  In
contrast, those groups exposed to contaminant/residual levels of
acrylamide are not as well-characterized, but risks should
logically be much lower.

10.4.  Summary
     As Tables 10-1, 10-2, and 10-4 indicate,  persons involved in
sewer grouting face potential cancer risks of 10"3  to  10'1  as
well as margins of exposure (MOE)  for neurotoxic  and.reproductive

                              10-36
-------
     Finally, although the risk of heritable mutations is
difficult to quantify, the qualitative weight of evidence is
high.  Only human evidence could make the case stronger.
Consequently, in the absence of information to the contrary, all
exposed groups must be considered to be at some level of risk,
with grouting workers at the highest level followed by
manufacturing/processing employees.
                              10-38
-------
                    11.   RISK CHARACTERIZATION

     This risk characterization presents the major conclusions of
EPA's risk assessment of aery1amide.  Aery1amide is a known
neurotpxin.  Animal studies have established effect levels and
have begun to elucidate its mode of action in causing
neurotoxicity.  Other studies show that acrylamide can cause
reproductive and developmental effects in laboratory animals.
Subchronic and chronic exposure studies, corroborated by other
scientific evidence/ support a conclusion that exposure to
acrylamide poses a risk of cancer for humans.  Also/ the data
indicate a strong potential for heritable mutations in the human
population.      *
     The risk characterization reviews the underlying scientific
foundation for these findings and describes the strengths and"
weaknesses of supporting data.  It is divided into three sections
which discuss the qualitative aspects of the risk assessment/ the
exposure, and the quantitative risk estimations at current
exposure levels.

ll.l.  Hazard
     The available data provide a basis for identifying several
human health hazards related to acrylamide exposure/ but with
varying degrees of confidence,  in order of decreasing
confidence, the identified hazards include:  neurotoxicity,
carcinogenicity, genotoxicity (heritable mutations) reproductive

                              11-1
-------
effects, and developmental effects.  There is a strong case for
identifying neurotoxicity, carcinogenicity, and heritable
mutations/ a moderate case for identifying reproductive effects,
and a weak case for identifying developmental effects of
aery1amide exposure.  The available evidence of these hazards is
sufficient to conduct quantitative risk assessments for
neurotoxicity, carcinogenicity, and reproductive effects, to a
limited extent for heritable mutations, but is not sufficient for
such purposes for developmental effects.
     The strongest data set is that which identifies
neurotoxicity from acrylamide exposure.  This effect has been
observed in both humans and laboratory animals.  The evidence
identifying this Affect is incontrovertible based on observations
of neurotoxic effects in both the peripheral and central ne'rvous
systems of humans, by the irreversibility of effects in some
human case reports, by the cumulative effects of chronic
exposures, and by the presence of a dose-response (from animal
studies).  The best animal data indicate NOELs and LOELs of 15
and 25 mg/kg, respectively, following a single dose, and 0.2 and
1.0 mg/kg/day, respectively, following chronic exposures.  Also,
in the animals tested there does not seem to be any appreciable
difference in effect levels.  In rats, cats, and monkeys the LOEL
is 1.0 mg/kg/day.
     Exposure to acrylamide by I a variety of routes may produce
serious neurological effects in humans.  The human reports of
neurotoxicity also indicate that although most of the individuals

                              11-2
-------
recovered completely after the exposure was stopped/ some
severely affected persons did not appear to recover completely
following cessation of the exposure, indicating an irreversible
effect.
     Aery1amide destroys the most distal axons of both central
and peripheral neurons and interferes with retrograde axonal
transport.  It also produces a number of biochemical effects that
may or may not be relevant to its neurotoxicity.  The precise
mechanism of action by which acrylamide produces neurotoxicity is
unknown.
     There are 53 human cases of acrylamide toxicity reported in
the literature, all but 5 related .to occupational exposure.  In
these occupational cases dermal exposure predominates.  As is
typical in such cases, the exposure levels are only qualitatively
known.  However, the severity of effects in these reports does
seem to be a function of exposure duration.
     A recent field monitoring study sponsored by EPA at sewer
grouting sites found one case of neurotoxicity and documented two
reports of previous neurotoxic symptoms.  As opposed to the cases
mentioned above, inhalation and dermal exposures were measured at
the grouting sites.  The monitoring indicated that significant
dermal exposure and risks can be expected.  (See section 11.3.)
     The early signs and symptoms of exposure noted in these
cases include: skin peeling and other skin changes, numbness and
tingling of the extremities, coldness of skin, excessive
sweating, bluish-red skin color, and muscle weakness.  These are

                              11-3
-------
generally followed by fatigue, confusion and other psychological
effects, gastrointestinal problems, and weight loss.  These signs
typically precede the dramatic findings of overt peripheral nerve
dysfunction (i.e., ataxia and weak or absent tendon reflexes).
These may be followed by an inability to stand, body tremors,
slurred speech, difficulty in swallowing and other signs.  In
general, victims showed complete recovery, although not always.
Individuals with relatively mild symptoms recovered completely
following cessation of exposure.  In summary, the human case
reports present evidence of both central and peripheral nervous
system effects following short- and long-term exposures to
aery1amide.  In principle, central effects are much more likely
to be irreversible, if neural damage takes place and are,
therefore, of greater concern.                             •*•
     In only one human case study can exposure be estimated.  ' A
Japanese family of five individuals was exposed through ingestion
and external use of well water contaminated with acrylamide,  and
the exposure was estimated to be about 11.4 mg/kg as a daily dose
and lasted about one month when symptoms developed.  The symptoms
and doses correspond fairly well to data from prolonged exposure
in cats and monkeys.  Even so, the human data are quantitatively
inadequate for risk assessment because few data on human exposure
levels exist.  Based on recent and older studies of prolonged
exposure in rats, cats, and monkeys it appears that prolonged
exposure to 1 mg/kg/day or more causes neurotoxic effects.
                               11-4
-------
     Based on the available data there is evidence according to
EPA's Cancer Risk Assessment Guidelines to identify acrylamide as
a "B2-probable human carcinogen."  The evidence for this effect
of acrylamide is relatively strong and is strengthened by the
following:  occurrence of benign and malignant tumors/ tumors
observed in males and females, tumors observed at multiple sites,
tumors observed in 2 animal species (one tested in a lifetime
bioassay at very low doses and one tested in limited bioassays),
and a dose-response observed in a lifetime bioassay.  A final
report on the 2nd lifetime Oncogenicity Study in Rats with
Acrylamide was completed on June 27, 1989 (American Cyanamid
Company, Final Report, June 27, 1959).  This study confirmed the
multiple-sites carcinogenicity of acrylamide in the rat.  The
prior carcinogenicity study conducted by Johnson et al. (Toicicol.
Appl. Pharmacol. JL5:154-168, 1984) showed similar indications',
although results of the two studies differed somewhat regarding
specific incidences.  This constitutes "sufficient" evidence of
carcinogenicity from animal studies.  Also,  the genotoxicity data
indicate that acrylamide can interact with and damage DNA
material.  Three epidemiologic studies have been reviewed.  One
study showed a slight increase in lung and central nervous system
cancers.  However, the study cannot be characterized as an
adequate appraisal of potential cancer risk because of the need
for more complete information concerning workers' exposure and
mortality experience.  The other study, while not showing any
increase in malignancies in a workplace cohort when workers with

                               11-5
-------
previous dye exposure were excluded/ suffered from small cohort
size, limited follow-up and low exposure (in years).  The third
study was ambiguous in its definitions of the study groups as
well as having inadequate exposure data concerning aery1amide and
no data concerning other exposures.  Consequently, the available
human data are "inadequate" under the Guidelines to judge a human
cancer hazard from acrylamide.
     Although the lifetime bioassay in rats was a drinking water
study and some populations are exposed to acrylamide by the
dermal and inhalation routes, the available information on the
pharmacokinetics of acrylamide indicates that the applied dose
whether by the oral, dermal, or inhalation routes is probably
directly proportional to internal effective dose over a range of
dosing conditions.                                         •»•
     Acrylamide has none of the properties that suggest major*
non-linearities between applied dose and internal dose.  It is
rapidly and completely absorbed by the gut.  It is quite
uniformly distributed among tissues and, being quite polar, does
not tend to accumulate in adipose tissue.  It is rapidly cleared
by metabolism.  The proportionality of tissue accumulation of
radioactive labeled acrylamide to applied dose holds over a range
of tested dosing rates and multiple dosing patterns do not
introduce any non-proportionalities.  Thus there does not appear
to be the potential for major differences in the carcinogenic
potential of acrylamide by route of exposure.
                              11-6
-------
     The genotoxicity data on acrylamide provide a basis to
support the carcinogenicity hazard identification and to support
mutagenicity as a separate endpoint of concern for acrylamide
exposures.  The major concern for the genotoxicity of acrylamide
is its clastogenic activity, which appears more pronounced in the
germ cells than in somatic cells.  The interaction with the
germinal tissue suggests the possible heritability of acrylamide
induced mutations.
     The evidence identifying this effect is strengthened by the
consistent pattern of the laboratory results and by the strong
results in a mouse heritable translocation assay.
     The EPA guidelines for mutagenicity risk assessment allow
for an evaluation*of the potential genetic risk associated with
human exposure to acrylamide.  The genotoxicity data for   •*
acrylamide suggest an intrinsic mutagenic (clastogenic) activity
and provide sufficient evidence for chemical interaction in the
mammalian gonad.  The body of evidence suggests that acrylamide
may induce alterations in the genome the germinal cells.  These
data provide a fairly strong weight of evidence bearing on the
potential for human germ-cell mutagenicity and its heritability.
The results of a heritable translocation assay increase the
strength of a human germ-cell mutagenicity concern to a level
that would only be further strengthened by direct human evidence
(the highest level of evidence for human mutagenicity).
                               11-7
-------
     Additional data (at multiple doses) are necessary from the
mouse heritable translocation assay to establish a dose-response
and to strengthen quantitative risk estimation of this endpoint.
     A reproductive hazard from aery1amide exposure has been
identified based on the results of several animal studies.
Reproductive effects of aery1amide were observed both
physiologically and biochemically.  The doses at which these
reproductive effects were observed also produced other toxic
effects in all the animals tested, except in a few instances.
Consequently, the available data indicate that aery1amide can act
directly on the reproductive system, rather than indirectly
through stress or other systemic effects such as neurotoxicity.
Thus the concern for the reproductive toxicity potential is
heightened.                                                i
     The LOEL and NOEL for reproductive effects, as indicated'by
the available data (Table 6-2 in section 6-1.), are 8.8 and 4.2
mg/lcg/day, respectively.
     Although there is some evidence that acrylamide may cause
developmental effects, many of the available reports indicate
that developmental effects occur only in the presence of maternal
toxicity.  However, two reports, one involving biochemical
changes of unknown significance and another involving
neurotoxicity in the pups do provide some limited basis for
identifying a developmental hazard from acrylamide exposure.
Unless further study elucidates the developmental hazard, concern
                               11-8
-------
for other effects should be emphasized over these developmental
effects.
     While the available data provide a good general
understanding of the metabolism of aery1amide (including:
absorption/ distribution, biotransformation, and elimination),
there is still a lack of detailed pharmacokinetics for certain
aspects of its metabolism.  For example, there is currently a
lack of specific dermal absorption rates under various
conditions, although the data are adequate to place bounds on
these rates.  However, recent studies of the mechanism of action
of aery1amide's toxicity, especially its neurotoxicity, have
begun to yield an understanding of. some the processes or
mechanisms through which acrylamide produces its various toxic
effects.                                                   ^
11.2. Exposure
     The body of exposure data is relatively extensive for the
manufacturing/processing and sewer grouting sectors, less so for
the others.  Monitoring was performed by NIOSH at the three
domestic acrylamide manufacturing sites.  These sites also
include facilities to convert acrylamide to various
polyacrylamide products.  The NIOSH survey produced validated
exposure data that is in good agreement with concurrently
produced data by one of the companies monitored.  The one
drawback of the NIOSH survey is that dermal exposure was only
identified as a potentiality based on equipment wipe samples.
Even so, an analysis of the production process and a review of

                               11-9
-------
the safety equipment routinely used indicates a low potential for
dermal exposure.
     While not as extensively monitored as manufacture/
processing, the sever grouting sector is well-characterized
regarding both inhalation and dermal exposure.  Because the
grouting procedure and equipment used are relatively uniform
throughout the country (there are only two major suppliers of
grouting equipment), the data developed from the geographically
limited site surveys could be applied country-wide.  The NIOSH
data for a sewer grouting site agree well with the sites
monitored by EPA.  An uncertainty regarding the dermal exposure
estimates is the rate of absorption.  It was assumed that up to
25 percent of the*recovered aery1amide would be absorbed.  This
is not an overly conservative estimate and it agrees well with
the animal data.
     The two major consumer exposures, drinking water and sugar
consumption, can only be estimated because of a lack of
monitoring data.  (Polyacrylamide products, which contain
residual acrylamide monomer, are used as flocculants to clarify
raw water in potable water plants and as clarifying agents in
sugar refining.)  However, based on maximum allowable and typical
usage rates for the polyacrylamide products used, a range of
exposures can be presented with some confidence.
     No: estimates can be provided for the other exposure
categories, these mostly include users of polyacrylamide products
which contain residual acrylamide and were not a focus of the

                              11-10
-------
assessment.  Direct users of aery1amide such as by laboratory
personnel also have some potential for acrylamide exposure, but
this could vary greatly from laboratory to laboratory.

11.3.  Risk Estimates
     Three endpoints were quantified, cancer/ neurotozicity and
reproductive effects.  The risk of heritable mutations was
examined, but specific estimates of risk are not recommended.
Since adequate human data were not available, estimates of cancer
risk were generated using animal data from a long-term bioassay.
Statistically significant tumor incidence data in animals were
pooled for purposes of quantitative risk estimation, since risk
numbers derived ftom site-specific tumor incidence data may not
be predictive of (and may in fact underestimate) "whole-body"
risks that are determined using the pooled individual animal '
data.  The level of risk for neurotoxicity and reproductive
effects was determined by calculating margins of exposure.
     Since risks at low exposure levels cannot generally be
measured directly either by experiments in animals or by
epidemiologic studies, a number of mathematical models have been
developed to extrapolate from high to low doses.  The Office of
Science and Technology Policy published principles on model
selection which states that "No single mathematical procedure is
recognized as the most appropriate for low-dose extrapolation in
carcinogenesis.  When relevant biological evidence on mechanism
of action exists, the models or procedures employed should be

                              11-11
-------
consistent with the evidence.  When data and information are
limited, however, and when much uncertainty exists regarding the
mechanism of carcinogenic action, models or procedures which
incorporate low-dose linearity are preferred when compatible with
the limited information."  Data relevant to selecting a model for
cancer risk extrapolation associated with exposure to aery1amide
were reviewed.
     As a result of that review, the linearized multistage model
was used to estimate cancer risks.  An analysis of the data base
did not indicate that another model would be more appropriate or
that the doses should be adjusted to reflect different patterns
of distribution or metabolism.  The distribution of aery1amide
appears to be quantitatively the same regardless of route of
exposure.  Consequently, as indicated by EPA's Cancer Assessment
Guidelines and in the absence of information to the contrary,* the
linearized multistage model was used.
     Estimated excess lifetime individual cancer risks for
various exposure categories are presented in Table 11-1.  The
highest upper-bound risks are for sewer grouting workers, with
risks ranging from 10*2 to 10"1.  Risks for manufacturing
personnel range from 10'3 to 10"2 based on recent NIOSH monitoring
data.  Consumers of drinking water face risks of about 10'6.
Other exposed groups face unknown levels of risk due to a lack .of
exposure information.  However, except for lab workers, the other
groups would be exposed to acrylamide as a contaminant
                              11-12
-------
(or residue) in polyacrylamide products where the exposure
potential is expected to be very low.

Table 11-1.  summary of Estimates of Upper-Bound Individual
             Excess Lifetime Cancer Risks Associated with
             Exposure to Acrylamide

                                          Upper-Bound
Exposure Category                  Individual Risk Estimates
soil Grouting                           10"2  -  10"1
Manufacturing/Processing                10"3  -  10*2
Drinking Water                          10"6  -  10"5*
* Worst-case to typical exposure based on residual aery1amide
allowed.

     Other models*were also used to estimate risk, such as the
                  \                               w
independent and additive background logic, probit, Weibell,**. and
gamma multi-hit models.  In general, similar or higher risks were
obtained with these models as compared to the linearized
multistage model.
     The available hazard data for both reproductive effects and
neurotoxicity are adequate for risk estimation because well-
conducted studies are available to identify NOEL'S and LOEL's.
The data are not sufficient to do this for developmental hazards.
     As a way to measure this risk, margins of exposure (MOE)
were calculated using the NIOSH, EPA, and industry supplied data.
For neurotoxicity a NOEL of 0.2 mg/kg/day was used for risk
estimation.  A reference dose (RfD) of 0.0002 mg/kg/day was
developed using this NOEL and an uncertainty factor of 1000 (UP).

                              11-13
-------
The NOEL and RfD will be used to evaluate neurotoxic risks.  A
NOEL of 4.2 mg/kg/day was used for reproductive risks.  An MOB is
simply a measure of the proximity of an environmental exposure to
a NOEL or LOEL determined in a laboratory study (NOEL divided by
human exposure level).  For instance, if a NOEL of 0.3 mg/kg/day
were used and the exposure was 0.1 mg/kg/day, then the HOE would
be 3.  The lower the MOE the higher the concern.  Alternatively,
estimated human exposure can be compared to the RfD either
directly or whether the MOE is less or greater than the DP (X any
modifying factor if appropriate)  used to determine the RfD.  (See
section 10.1. for a discussion of RfD's.)
     The NIO8H and EPA monitoring data indicate a high potential
for neurotoxicity*in some groups.  As Table 11-2 and Figure 11-1
indicate, MOE's for sewer grouting are less than 10 and are*.
significantly lower than the UF of 1000 used to set the oral RfD
of 0.0002 mg/kg/day.  When viewed on a dose basis (Figure 11-
Kb)) two sites exceeded the OSHA standard.  This supports the
finding that one of the nine employees observed during EPA's
field monitoring survey exhibited signs of aery1amide induced
neurotoxicity and two others reported past episodes of
neurotoxicity, which indicates that many grouting workers may be
exposed to levels of acrylamide close to the estimated human
effect level (about 1 mg/kg).  In the manufacturing/processing
sector, MOE'S ranged from 23 to 282.  No sighs of neurotoxicity
were observed at any of the sites surveyed.
                              11-14
-------
     In studies where both reproductive and neurotoxic effects
have been seen, the neurotoxic effects were seen at lower dose
levels.  The evidence of neurotoxic effects is unquestionable;
both animal effects have been observed.  Adequate data are
available to show that there are no major differences in effect
levels in animals and limited evidence that similar levels cause
neurotoxic effects in humans.  In contrast, the weight of
evidence for reproductive effects in animals is not as high as
for neurotoxicity.  The MOE's for reproductive effects are higher
than those for neurotoxicity.  Reproductive effects have been
seen in some studies in the absence of neurotoxic effects but not
in others.  The risk presented for. reproductive effects is not
considered as high as for neurotoxicity at a given exposure
level.                                                     *
     Consumers of water have MOE's greater than 1000.  Laboratory
personnel may be at risk, but the level of risk is difficult to
gauge without better exposure data.  Persons who use
polyacrylamide products are expected to have high MOE's based on
the low exposure potential.
     Finally, acrylamide appears to be a potent clastogenic agent
in germ cells.  This presents the possibility for the
heritability of acrylamide induced mutations.  Under EPA's
mutagenicity guidelines only human evidence could make the case
stronger.  Consequently, in the absence of information to the
contrary, all exposed groups must be considered to be at some
                              11-15
-------
a)
MOEs

NOtL OSIIA
T'
i
1 .
I T
G C.C

b)
DOSE-
•g/kg/day
LOEL
J
I
I.'O


1
10
I
T
c

NOEL
\L
1
0.1


1
20
1



T
i
0.01


1
30
1
T t
H H

ACCIH
Jr
1
0.001

ACCJH
1 1 1 1
40 60 100 1000
1 1 II
TT
H M
f

fill
0.0001 . 0.00001 <0. 00001


1
>1000
1
t
DW


1


       C    CCC    MM      MM
DW
Figure Il-l.  Summary  of  neurotoxtc1ty risks, exposures, and exposure standards.
              C'Croutlng  site.   H**Hanufacturlng/Hru
-------
     Table 11-2.  Summary of Noncancer Risks Expressed as
                  Margins of Exposure Associated with Aery1amide
                  Exposure

A.  Neurotoxicity

Range of MOE's                    Exposure Groups

    <10                  OSHA standard, Grouting sites (EPA)
   11-50                 Cyanamid, Dow (M)
   51-100
 101-1000                NALCO, Dow (P), Soil Grouting (NIOSH)
  > 1000                 Drinking Water

B.  Reproductive Effects

Range of HOE'3                    Exposure Groups

   20-200                OSHA Standard, Grouting sites (EPA)
  210-700                Cyanamid, Dow (M)
  700-4000
   >4000                 Dow (P), NALCO, Soil Grouting (NIOSH)
    >105                  Drinking- Water
                 
-------
lethal and heritable translocation test results would be
required, hopefully using the same multiple-dose levels.
                              11-18
-------
12.  SUMMARY OF PUBLIC COMMENTS AND EPA RESPONSES ON PRELIMINARY



     ASSESSMENT OF HEALTH RISKS FROM EXPOSURE TO ACRYLAMIDE



Comments were submitted by the four following concerned public



groups:  synthetic organic Chemical Manufacturers Association/



Inc., Nalco Chemical Company/ the Dow Chemical Company/ and the



American Cyanamid Company.  The corresponding responses were



submitted by the U.S. Environmental Protection Agency.



The comments (C) and responses (R) are categorized into six major



areas of interest:  carcinogenesis/ genotoxicity, neurotoxicity,



reproductive effects, exposure/ and risk assessment.







     12.1.  Carcinogenesis;  (C)   The 2-year chronic bioassay



performed by the Aery1amide Producers Association does not



provide adequate support to classify aery1amide as a B2    ^



carcinogen.  Results were distorted by including data on benign



tumors in the total pool of tumors.  A corrected Table 10.3 has



been submitted.  A second carcinogenicity study completed by the



American Cyanamid Company is being analyzed.



     While an increased cancer incidence occurred when a promoter



treatment followed aery1amide, the promoter also showed



tumorigenic activity.



     Two human epidemiology studies showed aery1amide is not



carcinogenic to humans at levels of worker exposure.



     (R)  The B2 classification is supported by several studies.



Using some of.the same data/ aery1amide has been classified as a
                               12-1
-------
Group 2B carcinogen by the International Agency for Research of
Cancer.
     The results from the second carcinogenic study will not
negate the positive responses found in the first study.
     Risk assessment should not be based upon initiation activity
if no promoter is known to be present.  The risk of acrylamide
may be greater in mixtures than alone.
     The human studies performed are inadequate to correlate
exposure and incidence.

     12.2.  Genotoxicity;  (C)  Genotoxic effects occur when
acrylamide interacts with cytoskeletal proteins such as
microtubules and fticrofilaments.  This protein interaction may
explain why chromosomal effects can be produced without point
mutations.
     (R)  Acrylamide can alkylate DNA in vitro, induce
aberrations, and bind to testicular DNA in vivo.

     12.3.  Neurotoxicityi  (C)  Using electron microscopy,
chronic tests determined the lifetime neurotoxic NOEL to be
approximately 0.5 mg/kg/day.   The 90-day NOEL of 0.2 mg/kg/day is
derived only from a range-finding study.  The better data using a
0.5 mg/kg/day has been ignored.
     Neurotoxicity occurs at the same dose in the reproductive
studies as in the neurotoxicity studies.  The measure of
neurotoxicity many times is the gross observation of weakness of

                               12-2
-------
the bind limbs.  No histology was performed nor were quantitative
evaluations of hind limb alterations conducted.  The cat study of
acrylamide-induced hind limb weakness is not acceptable as all
control animals in the study died.
     The study on acrylamide-induced alterations of retrograde
transport to determine the single-dose NOEL and LOEL is
questionable since the functional deficit was not evaluated.
     No evidence of neurotoxicity was identified in a screening
test of workers associated with the current standard
concentration of 0.3 mg/m3.  No trend toward elevated mean scores
from the Optacon Tactile Tester have been observed either.
     Data exists stating that acrylamide and its neurotoxic
analogs reduced colchicine-binding in the sciatic nerve, but not
in the brain and cerebellum.                               t
     (R) The NOEL OF 0.2 mg/kg/day was accepted because electron
microscopy, the more sensitive method for endpoint measurement,
was used.  Light microscopy was used to determine the 0.5
mg/kg/day dosage.  EPA is required to be conservative in setting
reference doses and making risk estimates.

     12.4. Reproductive and Developmental Toxicity;  (C)
Calculations for the NOEL and LOEL were not based upon a
mg/kg/day dosage, but rather upon a mg/animal/day dosage.
Corrected dose levels are -submitted.  The method used for        :
detection was not sensitive enough to determine neurotoxicity at
lower concentrations.

                               12-3
-------
     A NALCO study showed no tibial nerve changes when pups were
exposed to aery1amide throughout gestation and lactation, but
instead showed tibial nerve lesions, weight gain, and possible
clinical effects at different dosages.

     (R)  The following revised NOELs AND LOELs are essentially
the same as those offered in the comments and should be included
in the corrected document:
     female reproductive effects:  LOEL = 17.4 mg/kg/day noted,
                                   NOEL = 13 mg/kg/day
     male reproductive effects:    LOEL =8.8 mg/kg/day
                                   NOEL = 4 mg/kg/day
The study was done in mice, not rats.
     The findings of developmental and reproductive toxicity
                 *
would not be altered by coincident neurotoxicity.
                                                           ^
     The results of the Nalco study do not eliminate the concerns
for developmental toxicity.

     12.5.  Exposure;  (C)  The total domestic involvement in
acrylamide manufacture is less than 100.  It is estimated that
there are 40 polymer plants in this country with fewer than 5
people unloading acrylamide.  Protective clothing and equipment
is widely used by workers.
     The average years-of-exposure for acrylamide manufacturing/
processing was determined to be 6.5 years.  No adjustment for
less-than-lifetime exposure was made.   -
                               12-4
-------
     (R)  The total number of workers exposed at the four
manufacturing sites and the 40 polymer processing sites should
range from 500 to 1,000.
     Uncertainties in the manner and frequency of using
protective equipment prevents us from considering their use in
exposure assessments.  The estimate of 10,000 workers potentially
exposed to aery1amide in 27 occupations was based on the National
Occupational Hazard Survey conducted by NIOSH.
     Potential exposure to aery1amide from polyacrylamide gel
electrophoresis has not yet been determined.

     12.6.  Risk Assessment;  (C) The linearized multistage model
does not fit the purported cytoskeletal protein binding mechanism
of action of aery1amide for data in the observable range.  *
     Regarding the factor of 5.85 for interspecies extrapolation/
toxicological effects from aery1amide occur in the same dose
range regardless of species.  A metabolic rate extrapolation
between species based on a surface area correction is
inappropriate.  Dose extrapolation is better done on a mg/kg
basis which would reduce the risk by a factor of 5 to 10.
     The 2-year study provides a NOEL estimate of 0.5 mg/kg/day
derived from exposure as long as 12 months.  Maximum Observed
Effects (MOE) should be included in the document without
including company names.
                               12-5
-------
     The requirement for multiple chromosomal modifications to
effect DNA damage indicates that a threshold model would better
represent the data.  A multiple hit model should be invoiced.
     Since the human epidemiology study showed no increased
incidence of cancer, the risk calculated by the EPA is
inaccurate.
     Calculations for exposure and the margin of exposure for
grouting sites in Table 10.1 should be corrected.
     (R) The use of nonthreshold models is appropriate when low-
dose linear behavior cannot be ruled out.  The multistage model
does fit the data in the observable range.
     An interspecies extrapolation factor of 7.05 was used for
risk assessment. *The aery1amide risk estimation procedure
follows that of the EPA Guidelines for Mutagenicity Risk   "*•
Assessment.
     NOELs across species range from 0.2 to 2.0 mg/kg.  LOELs are
very similar across species at 0.1 mg/kg.  After calculations are
made, the range of reference dose for preventing chronic
neurotoxic effects is 1-9 ug/kg/dy.
     All that is required for linearized multistage models such
as those used here are multiple biological events.  The number of
chromosome modifications needed to induce enough strain to result
in DNA breakage could depend on the location affected.
                              12-6
-------
                         13.  REFERENCES

1.    Abt Associates, Inc.  1990.  Economic Analysis of a Proposed
     Ban on Chemcial Grouts Containing Acrvlamide and N-Methvlol-
     acrvlamide. June 13, 1990.  Prepared under Contract for
     Economics and Technology Division (ETD, Office of Pesticides
     and Toxic Substances (OPTS), U.S. Environmental Protection
     Agency (USEPA).

1.1  Agrawal,  A., Seth, P.,  Squibb, R., Tilson, H., Uphouse, L.,
     and Bondy, S. 1981.  Neurotransmitter receptors in brain
     regions of acrylamide-treated rats.   I:  Effects of a single
     exposure to acrylamide.  Pharmacol.  Biochem. Behav. 14:527-
     531.

2.    Ali., S.F., Hong, J-S., Wilson, W.,  Uphouse, L., and Bondy,
     S.  1983.  Effect of acrylamide on neurotransmitter
     metabolism and neuroleptide levels in several brain regions
     and upon circulating hormones.  Arch. Toxicol. 52:35-43.

3.    American Cyanamid Company.  1969.  Chemistry of Acrylamide.
     70pp.  Process Chemicals Department, American Cyanamide
     Company,  Wayne, NJ.
                  i
4    Ibid.  1980.  Wayne, New Jersey.  TSCA FYI submission, cover
     letter dated May 28, 1980.  A fetal toxicity study of •«•
     acrvlamide in rats.  December 13, 1979. 49pp. Washington,
     DC:  Office of Toxic Substances (OTS),  USEPA.  OTS Doc. "ID
     No. FYI-OTS-0680-0076.

5.    Ibid.  1981.  Stanford Laboratory:  Validation of an
     Analytical and Air Sampling Method for Acrvlamide in Air.

6.    Ibid. 1983c.  Wayne, New Jersey.  TSCA 8(d) submission Fiche
     #0206055-4(5).  Genetic toxicology micronucleus test (MNT).
     Acrylamide monomer.  May 28, 1982. 33pp. Washington, DC:
     OTS, USEPA.  OTS Doc. Control No. 87-8211681.

7.    Ibid. 1983d.  Wayne, New Jersey.  TSCA 8(d) submission Fiche
     #0206055-5(1).  CHO/SCE.  In vitro sister chromatid exchange
     in Chinese hamster ovary cells  (CHO).  Acrylamide monomer.
     July 18,  1982.25pp.  Washington, DC:  OTS, USEPA.  OTS Doc.
     Control No. 87-8211682.

8.    Ibid.  1983a.  Wayne, New Jersey.  TSCA 8(d) submission
     Fiche #0206055-5(4).  Ames Salmonella/microsome plate test.
     Acrylamide monomer and acrylonitrile.  July 22, 1982.  21pp.
     Washington, DC: OTS, USEPA. OTS Doc. Control No. 87-8211684.
                               13-1
-------
9.   American Cyanamid Company.  1983e.  Wayne, New Jersey.  TSCA
     8(d)  submission Fiche #0206055-6(1).  Rat hepatocvte primary
     culture/DNA repair test.  Acrylamide.  September 21, 1982.
     35pp.  Washington, DC:  OTS, USEPA.   OTS Doc. Control No.
     87-8211285.

10.  Ibid.  1983b.  Wayne, New Jersey.  TSCA 8(d) submission
     Fiche #0206055-6(4).  CHO/HGPRT mammalian cell forward gene
     mutation assay.  Acrvlamide.  September 23, 1983.
     Washington, DC:  OTS, USEPA.  OTS Doc. Control No. 87-
     8211686.

11.  Ibid.  1985a.  Wayne, New Jersey.  TSCA 8(d) submission.
     Fiche #0206897.  Drosophila sex-linked recessive assay of
     acrylamide.  1984.Washington DC:  OTS, USEPA.  OTS Doc.
     Control No. 87-8216232.

12.  Ibid.  1985b.  Wayne, New Jersey.  TSCA 8(d) submission.
     Fiche #0206898-(7).   Rat hepatocvte primary culture/DNA
     repair test.  Acrvlaroide. April 2, 1985.  144pp.
     Washington, DC:  OTS, USEPA.  OTS Doc. Control No. 87-
     9216235.

13.  Ibid.  1985c»  Wayne, New Jersey.  TSCA 8(d) submission.
     Fiche #0206898-3(4).  Activity of T1717 in the in vitro
     mammalian cell transformation assay in the presence of^
     exogenous metabolic activation.  Acrylamide.  April 2, 1982.
     20pp.  Washington, DC:  OTS, USEPA.  OTS Doc. Control No,
     87-8216236.

14.  Ibid.  1985d.  Wayne, New Jersey.  TSCA 8(d) submission.
     Fiche #0206898-3(6).  Activity of acrvlamide in the
     morphological transformation of BALB/3T3 mouse embryo cells
     in the presence of exogenous metabolic activation.  July 10,
     1984. 15pp.  Washington, DC:  OTS, USEPA.  OTS Doc. Control
     No. 87-8216237.

15.  Ibid.  1989.  A Lifetime Oncoqenicity Study in Rats with
     Acrylamide.  A Final Report.June 27,1989.^

16.  Banerjee, S. and Segal,  A.  1986.  In vitro transformation
     of C3H/10T1/2 and NIH/3T3 cells by acrylonitrile and
     acrylamide.  Cancer Lett. 32:293-304.
   15.  Pgs 5-32  missing from original  text  received  from HERD,
     OPTS,  USEPA.
                               13-2
-------
17.   Bikales, N.M.  1967.  Flocculation.  In:  Encyclopedia of
     Polymer Science and Technology, vol. 7. New York:  Wiley and
     Sons.   pp 64-76.

18.   Bikales, N.M. 1968.  Mining and Metallurgical Application.
     In: Encyclopedia of Polymer Science and Technology, vol 8.
     New York:  Wiley and Sons,  pp 798-812.

19.   [Bikales, N.M., ed.,]  Forest W.A.  1973.  Water Soluble
     Polymers as Flocculants in Paper-making.  In:  Water-soluble
     Polymers.  New York:  Wiley-Interscience.  pp. 3-19.

20.   Bishop, J.B. and Kodell, R.L.  1980  The heritable
     translocation assay:  its relationship to assessment of
     genetic risk for future generations.  Teratoaen. Carcinogen
     Mutagen. 1:305-32.

21.   Bull,  R., Robinson, N., Laurie, R., Stoner, G., Greisiger,
     E., Meir, J., Stober, J.  1984a.  Carcinogenic Effects of
     Acrylamide in Sencar and A/J Mice.  Cancer Research 44:107-
     111.

22.   Bull,  R., Robinson, M., Stober, J.  1984b.  Carcinogenic
     Activity of Acrylamide in the Skin and Lung of Swiss-ICR
     Mice.   Cancer Letters 24:209-212.

23.   Burek, J.D., Albee, R., Beyer, J., Bell, T., Carreon, R.,
     Morden, D., Wade, C., Hermann, E., and Gorzinski, S.  1980.
     Subchronic toxicity of acrylamide administered to rats in
     the drinking water followed by up to 144 days of recovery.
     J. Env. Path. Toxicol. 4;157-182.

24.   Carlson, G. and Weaver, P.  1985.  Distribution and Binding
     of KC- acrylamide to Macromolecules in SENCAR and BALB/c
     Mice following Oral and Topical Administration.  Tox. App.
     Pharm. 79:307-313.

25.   Celtite.  1985.  Celtite, Inc. Telephone discussion between
     P. Switka and Mathtech, Inc.  April 17, 1985.

26.   Chem Purch.  1983.  Houston, C.W.  Chemicals for Enhanced
     Oil Recovery.  Chemical Purchasing,  pp.  37-48.  April
     1983.

27.   Cimino, M., Tice, R., Liang, J.  1986.  Aneuploidy in
     mammalian somatic cells in vivo.  Mutat Res. 167:107-122.
                               13-3
-------
28.  CMR.  1985.  Chemical Profile:  Acrylamide.  Chemical
     Marketing Reporter.  January 7, 1985.

29.  Collins, J.J.  1984.  A Proportional Mortality Ratio
     Analysis of Workers Exposed to Acrvlaroide at the Warners
     Plant.  Epidemiology Section, American Cyanamid Company.

30.  Conway, E., Petersen, R., Collingworth, R., Graca, J., and
     Carter, J.  1979.  Assessment of the need for and character
     of limitations on acrvlamide and its compounds.  Draft
     Report. May 1979 prepared for the Office of Pesticides and
     Toxic Substances (OPTS).  Washington, DC:  USEPA.  EPA
     Contract No. 68-01-4308.

31.  Cues.  1985.  Cues, Inc. Telephone discussion between W.
     Spain and Mathtech, Inc.  April 18, 1985.

32.  [Davidson, R., ed.,] Volk, H., and Friedrich, R.  1980.
     Polyacrylamides.  In: Handbook of Water-soluble Gums and
     Resins.  Chapter 16.  New York:  McGraw-Hill.  Pp. 1-26.

33.  Davis, L.N., Durkin, P.R., Howard, P.H, and Saxena, J.
     1976.  Investigation of Selected Potential Environmental
     Contaminants:  Acrylamides.  August 1976.  Washington, DC:
     USEPA.EPA 360/2-76-008.  PB 257-704.

34.  Dearfield, K.L., Abernathy, C.O.,  Ottley, M.S., Brantner,
     J.H., and Hayes, P.P.  1988.  Acrylamide:  Its metabolism,
     developmental and reproductive effects, genotoxicity, and
     carcinogenicity.  Mutation Res. 195:45-77.

35.  Delpire.  1985.  Acrylamide Exposure Estimates.  Memorandum
     to Richard Hefter,  ECAD, OPTS, USEPA  from Lynn Delpire, May
     30, 1985.  Environemntal Exposure Division (EED), OPTS,
     USEPA.

36.  Dixit, R., Mukhtar, H., Seth, P.,  and Murti,  C.  1981.
     Conjugation of acrylamide with glutathione catalyzed by
     glutathione-S-transferases of rat liver and brain.  Biochem.
     Pharmacol. 30f131;1739-1744.

37.  Dow Chemical Company.  1980.  Acrylamide:  Environmental
     Fate.  Midland, MI:  Dow Chemical Company.

38.  Durham, W.F. and Wolfe, H.R.  1962.  Measurement of Exposure
     of Workers to Pesticides.  Bull. WHO 26:75-91.
                               13-4
-------
39.  Edwards, P.M.  1975.  The distribution and metabolism of
     acrylamide and its neurotoxic analogs in rats.  Biochem.
     Pharmacol. 24:1277-1282.

40.  Edwards, P.M.  1976.  The insensitivity of the developing
     rat fetus to the toxic effects of acrylamide.  Chem.-Biol.
     Interact. 12;13-18.

41.  Environmental Science and Engineering, Inc.  1981.
     Environmental Assessment for Acrylamide. Draft Test Rule
     Support Document:  Chapters I, II, IV, and V.  Report No.
     68-01-6153.

42.  Eskin, T.A., Lapham, L., Maurissen, J., and Merigan, W.
     1985.  Acrylamide Effects on the Macaque Visual System.  II.
     Retinogeniculate Morphology.  Invest. Opthalrool. Vis. Sci.
     26111:317-329.

43.  Frantz, S.W., Dryzga, M., Freshour, N., and Watanabe, P.
     1985.  In vivo/in vitro determination of cutaneous
     penetration of residual monomer from polyacrylamides.
     Toxicologist 5(39);10pp. (Abst.).

44.  Garland, T. and Patterson, M.  1967.  Six cases of
     acrylamide poisoning.  Brit. Med. J. 4:134-138.

45.  GCA.  1980.  GCA Corporation:  Acrvlamide Technical Control
     Options Analysis. Draft Final Report. August 1980, 109pp.
     Prepared for Office of Chemical Control (OCC), OTS, USEPA,
     EPA Contract No. 68-01-5960, Report No. GCA-TR-80-75-G.

46.  Generoso, W.M., Bishop, J.B., Gisslee, D.G., Newell, G.W.,
     Shew, C.J and von Halle, E.  1980  Heritable translocation
     test in mice.  Mut. Res. 76:191-215.

47.  Geochem.  1985.  Geochemical Corporation.  Telephone
     discussion between W. Clarke and Mathtech, Inc.  April 18,
     1985.

48.  Going J.E.  1978if.  Environmental Monitoring Near Industrial
     Sites;  Acrvlamide - CH2=CHCONH2.  March 1978, 74pp.
     Prepared under Contract # 68-01-4115 by Midwest Research
     Institute for VJ DeCarlo, OTS, USEPA,  Washington, DC: EPA
     560/6-78-001.

49.  Going, J.E. 1978J|.  Sampling and Analysis of Selected Toxic
 :-    Substances.  Task 2 — Acrylamide.  Midwest Research
     Institute, Kansas City, MO, 1978.
                               13-5
-------
50.  Gorzinski, S.f Johnson, K., Campbell, R., Morden, D., and
     Toilet, J.  1984.  Acrylamide;  Data from Interim Sacrifices
     from a Two-Year Prinking Water Chronic Toxicitv-Oncoaenicity
     Study in Fischer 344 Rats.

51.  Hamblin, D.  1956.  The Toxicity of Acrylamide — A
     Preliminary Report.  Hommage au Doven Rene* Fabre (Paris)
     Sede 3;195-199.

52.  Haseman, J., Huff, J., Boorman, G.  1984.  Use of historical
     control data in carcinogenicity studies in rodents.   Tox.
     Path. 12(2);126-135.

53.  Hashimoto, K. and Aldridge, W.  1970.  Biochemical studies
     on acrylamide, a neurotoxic agent.  Biochem. Pharmacol.
     19;2591-2604.

54.  Hashimoto, K., Sakamoto, J., and Tanii, H.  1981.
     Neurotoxicity of acrylamide and related compounds and their
     effects on male gonads in mice.  Arch.  Toxicol. 47:179-189.

55.  Hashimoto, K. and Tanii, H.  1985.  Mutagenicity of
     Acrylamide and its Analogues in Salmonella typhimurium.
     Mutat. Res. 158;129-133.

56.  Void
                                                           •*
57.  Hozier, J., Sawyer, J., Clive, D., and Moore,  M.  1985.
     Chromosome 11 aberrations in small colony L5178/TK+/ -
     mutants early in their clonal history.   Mutation Res.
     141:237-242.

58.  Hsie, A., Recio, L., Katz, D., Lee, C., Wagner, M.,  and
     Schenley, R.  1986. Evidence for reactive oxygen species
     inducing mutation in mammalian cells.  Proc. Natl. Acad.
     Sci. (USA) 83;9616-9620.

59.  IARC.  1979.  IARC Monograph. Volume 19.  Some monomers,
     plastics, and synthetic elastomers, and acrolein. Pp 47-71.
     Lyon, IARC, WHO.

60.  IARC.  1986.  IARC Monographs. Vol. 39.  Acrylimide;  Some
     chemicals used in plastics and elastomers. Pp 41-66, Lyon,
     International Agency Res. Cancer (IARC), World Health
     Organization (WHO).
                               13-6
-------
61.   Igisu, H.,  Goto, I., Kawamura, Y., Kato, M., Izumi, K., and
     Kuroiwa, Y.  1975.  Acrylamide encephaloneuropathy due to
     well water pollution.  J. Neurol. Neurosurq. Psychiatry
     18:581-584.

62.   Ikeda, G. J., Miller, E., Sapienza, P.P., Michel, T.C., King,
     M.T., Turner, V.A., Blumenthal, H., Jackson, W.E., and
     Levin, S.  1983.  Distribution of   C-labelled acrylamide
     and betaine in foetuses of rats, rabbits, beagle dogs, and
     miniature pigs.  Ed. Chem. Toxicol. 21;49-58.

63.   Ikeda, G.J., Miller, E., Sapienza, P.P., Michel, T.C., King,
     M.T., and Sager, A.O.  1985.  Maternal-foetal distribution
     studies in late pregnancy.  II. Distribution of [I-14 C]
     acrylamide in tissues of beagle dogs and miniature pigs.
     Ed. Chem. Toxicol. 23:757-761.

64.   Institut d1Hygiene et d'Epidemiologie.  1985.  Brussels,
     Belgium.  TSCA FYI submission FYI-OTS-0885-044A.  Studies on
     the genetic toxicology of acrylamide monomer.  April 8,
     1985. Washington, DC: OTS, USEPA.

65.   Jacobs, P.A, Fackiewicz, A., taw, P., Hilditich, C.J. and
     Morton, N.E.^ 1975.  The effect of structural aberrations of
     the chromosomes on reproductive fitness in man.  II.
     Results.  Clin. Genet. 8:169-78.
                                                           •\
66.   Johnson, K.A., Gorzinski, S.J., Bodner, K.M., and Campbell,
     R.  1984.  Acrylamide;  A Two-year Drinking Water Chronic
     Toxicity - Oncogenicity Study in Fisher 344 Rats.  Final
     Report dated September 21, 1984.  46pp.  Dow Chemical
     Company.

67.   Johnson, K.A., Beyer, J., Bell, J., Schuetz, D., and
     Gorzinski,  S.  1985.  Acrylamide:  A Two-year Drinking Water
     Chronic Toxicity-Oncogenicity Study in Fischer 344 Rats.
     Electron. Microscopy Portion. August 13, 1985.  Sponsored by
     American Cyanamid Company, et al.

68.   Johnson, K.A., Gorzinski, S., Bodner, K., Campbell, R.,
     Wolf, C., Friedman, M. and Mast, R.  1986.  Chronic toxicity
     and oncogenicity study on acrylamide incorporated in the
     drinking water of Fischer 344 rats.  Toxicol. Appl.
     Pharmacol.  85;154-168.

69.   JRB Associates, Inc.  Level I Materials Balance -
     Acrvlamide.  Revised Draft Report, Prepared under EPA
     Contract No. 68-01-5793 for Survey and Analysis Division
     (SAD), OTS, USEPA, June 16, 1980.
                               13-7
-------
70.  Kaplan, M., Murphy, S., and Gilles, F.  1973.  Modification
     of acrylamide neuropathy in rats by selected factors;
     Toxicol. APP!. Pharmacol. 24:564-579.

71.  Karol.  1983.  Karol R.  Chemical Grouting.  New York:
     Marcel Dekkar. pp. 50-73.

72.  Kirk-Othmer.  1978.  Morris J.D. and Penzenstadler R.J.
     Acrylamide Polymers.  In:  Kirk-Othmer Encyclopedia of
     Chemical Technology. 3rd ed.. vol. 1.  New York:  Wiley-
     Interscience.  pp. 312-330.

73.  Ibid.  1979.  Karol R. and Welsh J.  Chemical Grouts.  In:
     Kirk-Othmer Encyclopedia of Chemical Technology.  3rd ed..
     vol. 5.  New York:  Wiley-Interscience.  pp. 368-374.

74.  Ibid.  1982a.  Clark R.K. and Nahm, J.J.  Drilling fluids.
     In:  Kirk-Othmer Encyclopedia of Chemical Technology. 3rd
     ed.. vol. 17.  New York:  Wiley-Interscience.  pp. 143-165.

75.  Ibid.  1982b.  Froning, H.R., Fussell, D.D., and Heffern,
     E.W. Petroleum (Enhanced Oil Recovery).  In: Kirk-Othmer
     Encyclopedia of Chemical Technology. 3rd ed.. vol. 17.  New
     York:  WileyiJnterscience.  pp. 168-181.
                  t
76.  Ibid.  1982c.  Resins, water-soluble.  In:  Kirk-Othmei^
     Encyclopedia of Chemical Technology. 3rd ed.. vol 20.  New
     York:  Wiley-Interscience.  pp. 212-214.               "  .

77.  Koppers Company Incorporated.  1979.  Flickingn C.W.  Status
     Report of Industrial Hygiene Monitoring at the Bridcreville,
     Pennsylvania. Plant. Chemical Division. Organic Chemical
     Group.  TSCA Section 8fd) Submission, [received November 2,
     1982], dated August 31, 1979.  OTS Document Cont. No. 87-
     820575. 40pp.

78.  Superman, A.  1958.  Effects of Acrylamide on the Central
     Nervous System of the Cat.  J. Pharmacol. Exp. Ther.
     123:180-192.

79.  Lande, S.S., Bosch, S.J., and Howard, P.H.  Metabolism and
     Transport of Acrylamide in Soil.  Final Report.  Prepared by
     Syracuse Research Corporation for the USEPA.  Contract No.
     68-01-2679.  July 1979.
                               13-8
-------
80.   Lapadula, D.M.,  Bowe, M., Carrington, C.D., Dulak, L.H.,
     Friedman, M.A.,  Abou-Donia, M.D.  1987.  Binding of [u C]
     acrylamide to cytoskeletal elements of the nervous system.
     Unpublished Report.  Duke Univ. Med. Center, Durham, NC. and
     American Cyanamid Company., Wayne, NJ.

81.   Lijinsky, W. and Andrews, A.  1980.  Mutagenicity of vinyl
     compounds in Salmonella typhimurium.  Teratogen. Carcinogen.
     Mutaaen 1:259-267.

82.   Lonati-Galligani, M., Lohman, P., and Berends, F.  1983.
     The Validity of autoradiographic method for detecting DNA
     repair synthesis in rat hepatocytes in primary culture.
     Mutation Res. 113:145-160.

83.   Marlowe, C., Clark, M., Mast, R.W., Friedman, M.A., and
     Waddell, W.J.  1986.  The distribution of [UC]acrylamide in
     male and pregnant Swiss-Webster mice studied by whole-body
     autoradiography.  Toxicol. Appl. Pharmacol. 86:457-465.

84.   Matsuda, F.  1977.  Advances in the Production of
     Acrylamide.  Yuki Grosei Kagaku Kyokaishi 35(3):212-216.

85.   McCollister, J)., Oyen, F., and Rowe, V.  1964.  Toxicology
     of acrylamide.  Toxicol. Appl. Pharmacol. 6:172-181.

86.   Merigan, W., Barkdoll, E., and Maurissen, J.P.  1982.
     Acrylamide-induced visual impairment in primates.  Toxicol.
     Appl. Phannacol. 62:342-345.

87.   Merigan, W., Barkdoll, E., Maurissen, J.P., Eskin, T.A., and
     Lapham, L.W.  1985.  Acrylamide Effects on the Macaque
     Visual System.  1. Psychophysics and Electrophysiology
     Invest. Qpthalmol. Visual Sci. 26(3):309-316.

88.   Miller, M.J. and McQueen, C.  1986.  The effect of
     .acrylamide on hepatocellular DNA repair.  Environ Mutaaen
     8:99-108.

89.   Miller, M.J., Carter, D., and Sipes, I.  1982.
     Pharmacokinetics of acrylamide in Fischer-344 rats.
     Toxicol. Appl. Pharmacol. 63:36-44.

90.   Miller, M.S., Miller, M.J., Burks, T., and Sipes, I.  1983.
     Altered retrograde axonal transport of nerve growth factor
     after single and repeated doses of acrylamide in the rat.
   :  Toxicol. Appl. Pharmacol. 69:96:101.
                               13-9
-------
91.  Miller, M.S. and Spencer, P.  1985.  The Mechanisms of
     Acrylamide Axonopathy.  Ann. Rev. Phannacol. Toxicol.
     25:643-66.

92.  Moore, M., Amtower, A., Doerr, C., Brock, K., and Dearfield,
     K.  1987.  Mutagenicity and clastogenicity of acrylamide in
     L5178Y mouse lymphoma cells.  Environ. Mutaqen. 9:261-267.

93.  Moore, M., Clive, D., Howard, B., Batson, A., and Turner, N.
     1985.  In situ analysis of trifluorothymidine-resistant
     (TFTr)  mutants  of L5178Y/TK+/ -  mouse  lymphoma  cells.
     Mutation Res. 151:147-159.

94.  Morton, N.E., Jacobs, P.A., Frackiewicz, A., Law, P., and
     Hilditch, C.J.   1975.  The effect of structural aberrations
     of the chromosomes on reproductive fitness in man.  I.
     Methodology.  Clin. Genet. 8:159-68.

95.  MRI.  1987.  Assessment of Airborne Exposure and Dermal
     Contact to Acrvlamide During Chemical  Grouting Operations.
     Final Report.  September 19, 1985, 123pp.  Midwest Research
     Institute.  Prepared under EPA Contract No. 68-02-4245, WA
     50 for BED, OPTS, USEPA.

96.1 Nalco Chemical Co.  1987.  Oak Brook,  Illinois.  TSCA 8(e)
     submission 8EHQ-1287-0560  Follow-Up.   Fiche #0509747-1.
     Combined Two Generation Reproduction Study and Dominant
     Lethal Assay in Fischer 344 Rats Administered Acrvlamide in
     the Drinking Water, dated November 25, 1987, 127pp.
     Washington, DC:  OTS, USEPA.

96.2 Ibid.  1987.  Oak Brook, Illinois.  TSCA 8(e)  submission
     8EHQ-1287-0560.  Follow-Up.  Fiche #0509747-1.   Combined Two
     Generation Reproduction Study and Dominant Lethal Assay in
     Fischer 344 Rats Administered Acrylamide in the Drinking
     Water, dated December 23, 1987,  6pp. [Corrections to 96.1
     above.] Washington, DC:  OTS, USEPA.

97.  NIOSH.  1980.  National Occupational Hazard Survey.
     USDHEW.

98.  NIOSH.  1985.  Report to EPA.  Monitoring at Manufacturing/
     Processing Sites. National Institute for Occupational Safety
     and Health (NIOSH), U. S. Dept.  of Health and Human Services
     (USDHHS)
                              13-10
-------
99.  NTP.  1985.  National Toxicology Program Fiscal Year 1985
     Annual Plan.  National Toxicology Program (NTP), Public
     Health Service (PHS), USDHHS.   Publication NTP-85-055.

100. Nyquist, E.B. and Yocum, R.H.  1973.  Applications,  in:
     Functional Monomers.  New York:  Marcel Dekker, Inc. pp. 78-
     99.

101. O'Donoghue, J.L.,  ed.,  1985.  Acrylamide and Related
     Substances.  In:   Neurotoxicity of Industrial and Commercial
     Chemicals  Vol. II. Boca Raton: CRC Press.

102. Pickrel, C., Rench, J.D., Tuckfield, R.C., and Wilson, J.L.
     1986.  Review of Proportional Mortality Ratio Analysis of
     Workers Exposed to Acrylamide at the Warners Plant.  Draft
     Report. Task 1-3,  Subtask 4, February 18, 1986, 20pp.
     Battelle, Washington, DC:  Prepared under EPA Contract # 68-
     02-4246 for EED,  OPTS, USEPA.

103. Ramsey, J. , Young, J., and Gorzinski, S.  1984.  Acrylamide;
     Toxicodynamics in rats.  Unpublished Report with cover
     letter dated February 21, 1986, 13pp.  Dow Chemical Co.
     Midland, MI.                 -

104. Roa, G.N., Edljnondson, J., and Haseman, J.K.   1988.
     Influence of viral infection on tumor incidences, body
     weight and survival of Fischer 344 rats.  Toxicoloqist>
     8_:166.

105. Robinson, M., Bull, R.J., Knutsen, G.L., Shields, R.P., and
     Stober, J.  1986.   A combined bioassay utilizing both the
     lung adenoma and skin papilloma protocols.  Environ. Health
     Persp. 68:141-145.

106. Rodricks, J. and Taylor, M.R.  1983.  Application of risk
     assessment to food safety decisionmaking.  Reg. Tox. Pharm.
     3:275-307.

107. Sakamoto, J. and Hashimoto, K.   1986.  Reproductive toxicity
     of acrylamide and related compounds in mice - effects on
     fertility and sperm morphology.  Arch. Toxicol. 59:201-205.

108. Schaumberg, H., Arezzo, J., and Spencer, P.   1982.  Short-
     latency somatosensory evoded potentials in primates
     intoxicated with acrylamide:  Implications for toxic
     neuropathies in man.  Abstract No. 490.  Presented at the
     1982 meeting of the Society of Toxicology, Boston, MA.
     Toxicoloqist 2:139.
                              13-11
-------
109. Schwager, W. and Vorchheimer, N. Polyelectrolytes for water
     and wastewater treatment.  In:  Synthetic Polyelectrolvtes.
     Boca Raton, FL.  CAC Press,  pp. 1-45.

110. Shelby, M.D, Cain, K.T., Cornett, C.V. and Genetoso, W.M.
     1987  Acrylamide: induction of heritable translocations in
     male mice.  Environ. Mut. 9;363-368.

111. Shelby, M.D., Cain, K.T., Hughes, L., Braden, P., and
     Generoso, W.  1986.  Dominant lethal effects of acrylamide
     in male mice.  Mutat. Res. 173;35-40.

112. Shiraishi, Y.  1978.  Chromosome aberrations induced by
     monomeric acrylamide in bone marrow and germ cells of mice.
     Mutat. Res. 57:313-324.

113. Sickles, D.W.  1987.  Neural specificity of acrylamide
     action upon enzymes associated with oxidative energy-
     producing pathways.  1. Histochemical Analysis of NADH-
     Tetrazolium reductase activity.  Neurotoxicolocrv 8:623-630.

114. Smith, M.K., Zenick, H., Preston, R., George, E., and Long,
     R.  1986.  Dominant lethal effects of subchronic acrylamide
     administration in the male Long-Evans rat.   Mutat. Res.
     173:273-277.*

115. Sobel, W., Bond,  G.G., Parsons, T.W., and Brenner, F.E>
     1986.  Acrylamide Cohort Mortality Study, Brit. J. Ind.-Med.
     4.3:785-788.  (This study was also submitted to USEPA under
     section 8(d) of the Toxic Substances Control Act on April
     28, 1986.)

116. SOCMA.  1986.  Rautio, A.W.  Comments regarding Acrylamide/
     polyacrylamide use in water treatment and sugar refining.
     Letter to SS Shapley, EED, OPTS, USEPA, dated May 23, 1986,
     9pp.

117. Solomon, J.J.,  Fedyk, J., Mukai, F., and Segal, A.  1985.
     Direct alkylation of 2'-deoxynucleosides and DNA following
     in vitro reaction with acrylamide.  Cancer Res 45:3465-
     3470.

118. Spencer, P.  1979.  A Neuropathologic Study .of Acrylamide
     Intoxication Unpublished final report.  Washington,  DC:
     National Institute for Occupational Safety and Health
     (NIOSH).  Contract No. OH 00535.  NIOSH, USDHEW.
                              13-12
-------
119.  Srivastava, S., Das, M.,  and Seth, P.  1983.  Enhancement of
     lipid peroxidation in rat liver on acute exposure to styrene
     and acrylamide.  A consequence of glutathione depletion.
     Chem.-Biol. Interact. 45;373-380.

120.  Srivastava, S., Seth, P., Das, M., and Mukhtar, H.  1985.
     Effects of mixed-function oxidase modifiers on neurotoxicity
     of acrylamide in rats.  Biochem. Pharmacol. 34:1099-1102.


121.  Sublet, V., Smith, M.K.,  Randall, J., and Zenick, H.  1986.
     Spermatogenic stages associated with acrylamide (ACR)
     induced dominant lethality.  Toxicoloaist 6;292.
     (Abstract).

122.  Textile Res. J. 1960.  Polymerization and Condensation
     Reaction of N-Methylol Acrylamide Within Cotton Fabric.
     Textile Research Journal.  September 1960.  pp. 774-781.

123.  Tilson, H.  1981.  The neurotoxicity of acrylamide:  An
     overview.  Neurobehavioral  Toxicol. Teratol. 3:445-461.

124.  USEPA.  1986.  Appendix A   —r Reference Doses (RfD):
     Description and use in health risk assessments.
                  «
125.  USEPA.  1986.  Guidelines for Carcinogen Risk Assessment.
     51 FR 33992-34003.                                    ->-

126.  USI?PA, ODW.  1985.  Final Draft for the Drinking Water   '
     Criteria Document on Acrvlamide.  EPA Contract No. 68-01-
     6750.

127.  USEPA, OTS, AD.  1980.  Assessment of Testing Needs;
     Acrvlamide Support Document for Decision Not to Require
     Testing for Health Effects.  June 1980, 55pp.  EPA  560/11-
     80-016. Assessment Division (AD).

128.  Vanhorick, M. and Moens,  W.  1983.  Carcinogen-mediated
     induction of SV40 DNA amplification is enhanced by
     acrylamide in Chinese hamster CO60 cells.  Carcinoqenesis
     4.: 1459-1463.

129.  Vasavada, H. and Padayatty, J.  1981.  Rapid transfection
     assay for screening mutagens and carcinogens.  Mutat. Res.
                              13-13
-------
130. Versar.  1986.  Results of telephone contracts for task 153
     - enumeration of populations exposed to acrylamide in
     polyacrylamide gel electrophoresis.  Memorandum from Jim
     Konz, Versar to Lynn Delpire, BED, OPTS, USEPA dated June
     27, 1986.

131. Waalkens, D., Joosten, H., Taalman, R.,  Scheres, J.,Yih, T.,
     and Hoekstra, A.  1981.  Sister-chromatid exchanges induced
     in vitro by cyclophosphamide without exogenous metabolic
     activation in lymphocytes from three mammalian species.
     Toxicol. Lett. 7:229-232.

132. Walden, R.,  Squibb, R., and Schiller, S.  1981.  Effects of
     prenatal and lactational exposure to acrylamide on the
     development of intestinal enzymes in the rat.   Toxicol,
     Appl. Pharmacol. 58;363-369.

133. Zaichkina, S. and Ganassi, Y.  1984.  Hicronuclearttest as a
     quantitative indicator of the structural disruptions of
     chromosomes induced by various agents.   Studia Biophysica
     99:203-210.

134. Zenick, H.,  Hope, E., and Smith,  M.K.  1986.   Reproductive
     toxicity associated with acrylamide treatment  in male and
     female rats. * J. Toxicol. Environ. Health 17:457-472.
                              13-14
-------