U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
PB-254 023
CRITERIA DOCUMENTS FOR BENZIDINE
ENVIRONMENTAL PROTECTION AGENCY
1 JUNE 1976
-------
EPA-440/9-76-017
CRITERIA DOCUMENT: BENZIDINE
U.S.EPA, Office of Water & Hazardous Materials,
Office of Water Planning and Standards
June 1976
-------
HiC DATA I1- K «?•»-•: No.
Criteria Documents for Benzidine
j^rorv/rrous
E. Pctioraia* O.-ganLcaclo,, U,,
U. S. ErivirorETsntal Protection Agency
Office of Water Planning and Standards
401 ?1 Street,' S.W.
n/ D.C. 20460
10.
IK Coatcace/Graat No.
12. Sp
nad
ress
c- J^ C:^.\aiii:ioa N
Office of V/ater Planning and Standards
. .. U. S. Enviroa«Rantal Protection Agsncy
401 M Street, S.V/.
tnnj P. C. 20460 _ _
3. Type of-Report & Period
CoTtred .
Interim
14.
Notes
I'eE^i'S
OSiis dccrnent surmarizes the physical/cheiracal properties, toxicological
infbrnation and environmental fate and. effects of: .Benzidine, with emphasis
its aquatic behavior; Fran these data criteria are developed for the
protection of aquatic life.
ori
W-
'-o.-ds and Dcxruasat Anlysis.
Criteria
.Toxicity
Aquatic animals
J^miatic biology
Huran ecology
Safety factor
J7o. Descriptor*
/-'i-
Tcxic. Pollutant: UfluiinL StanJards
Feuc-val V.'ater i'ollut.ion Control Act
j •
17c. C05ATI J.r
i-.
r.-.;.ort)
_ ____ ?.'NtJ A'j'..;i j; !»
/«.'. >•. .-i-rity C. !.«*-< ( I t,l
iix«" I.ASSU irn
-------
ATTENTION
AS NOTED IN THE NTIS ANNOUNCEMENT,
PORTIONS OF THIS REPORT ARE NOT LEGIBLE,
HOWEVER, IT IS THE BEST REPRODUCTION
AVAILABLE FROM THE COPY SENT TO NTIS;
DIRECT QUESTIONS RESULTING FROM
ILLEGIBILITY TO:
Mr. Kent. Valentine
EPA (Williftri)
401. M Sl.rc'ct, P..W.
W.-i 53!iiny ton , D. C. 20460
Thone: 202-245-3030
-------
CRITERIA DOCUMENTS FOR TOXIC POLLUTANTS
Scientific rationale and criteria developed pursuant to Section :*07(a)
of th<; l-'cderal Water Pollution Control Act. I1. L. !)2-T>00, 33 U.S. C. §§ I2.r>l
et seq.. (lf)72). for the development and establishment of effluent limitations
for toxic substances are set forth in the following chapters.
Section 307 (a )(2) states inter alia that a proposed effluent standard
"... shall take into account the toxicity of the pollutant, its persistence,
x
degradability, the usual or potential presence of the affected organisms in
any water, the importance of the affected organisms and the nature and extent
of the effect of the toxic pollutant on such organisms..." Thereafter, having
considered these factors, the Administrator is to set an effluent standard
for toxic pollutants which provides an ample margin of safety.
In the development of criteria which serve as both the basis and the goal
for the establishment of effluent limits, reliance was placed on the toxicity
data derived from laboratory studies on a range of organisms including
invertebrate, vertebrate, and mammalian test species. These studies
provided extensive acute and chronic toxicity data based on feeding experi-
ments for a wide range of aquatic organisms and consumers of aquatic
organisms. Environmental studies documenting bioaccumulation in the food
v/eb of the toxic material by the food chain organisms and bioconcer.tration
by organisms directly from water provided an important component data
base upon which criteria were derived. Appropriate human toxicity data
and mammalian carcinogenic studies, where available, were used also in
developing criteria.
ii
-------
Aquatic toxicity data generally are obtained by one of two basic
methods, the static and flow-through bioassay. The more traditional
static bioassay employs a tank in which the test organisms are living
and to which a given concentration of toxicant is added. Any water
loss due to evaporation is made up by the addition of fresh water. The
flow-through bioassay, which is a more recent development, reflects
more nearly the natural conditions. Concentrations of toxic substances
are constantly maintained and provide a more accurate test of sensitivity
of aquatic species. Water in a flow-through test is replenished constantly
through flushing. Comparative results using the static and flow-through
bioassays demonstrate that flow-through data yield lower toxicity values.
for a pollutant than a static bioassay. This fact is demonstrated by -
comparative studies as discussed in the endrin document. However,
most of the data available were developed using static bioassays.
Some toxic pollutants are extremely stable and degrade only slowly
or- form persistent degradation products. Those pollutants which degrade
rapidly pose a less severe long-term hazard unless their entry to
the environment is continuous. A parent compound, e.g., aldrin, may
rapidly degrade or be altered to a more toxic form, i. e., dieldrin.
Biocoricentration of toxic pollutants is a significant consideration in
the-development of criterion. The r='.e and degree of accumulation in
an animal and the rate of loss from the animal are factors that help
define the potential magnitude of the pollution load problem. As an
iii
-------
example, a pollutant which bioaccumulates presents a hazard both to
aquatic sy a terns and potentially to man or othor carnivores associated
with the <:«-oByntem. To satisfactorily manage a persistent or
non-degradable pollutant requires the maintenance of a ceiling for
ambient levels in water which will afford protection to the food chain
and the consumers of aquatic life (animals including humans). The
body burden of toxic pollutants in fish or food chain organisms may
have no outward effect on the species but will affect consumers of that food
level. As an example, the brown pelican, when feeding on endrin-contaminate
fish may die or suffer species depletion through reproductive impairment.
Data on toxic effects of pollutants are not available for all species
that may be exposed to the toxic pollutant in these complex societies.
Such data would be necessary to ensure protection of the most sensitive
species. It is desirable to know the relative sensitivity of a wide
variety of species in order to have a better estimate of the sensitivity
of the:untested. most sensitive species. Because such data are not
available on all species, the range in sensitivity of a small number of
tested species is used to provide a measure of the range of sensitivity
of all species.
The natural aquatic environment includes many kinds and life stages
of plants and animals that are intricately interrelated to form communities.
Criteria are developed to protect these interrelationships and incorporate
aquatic toxicity data for a phylogenetic cross section of organisms as well as
iv
-------
species representative of wide geographic distribution. Chronic
Mtudies arc an important consideration in establishing criteria nnd require
ntudies of at least one generation, i.e., one reproductive cycle. Use of
an application factor for persistent and bioaecumulated toxic pollutants
represents consideration of a safety factor. As discussed in the
National Academy of Science publication on water quality (p. 185 of
the NAS/NAE Water Quality Criteria -- 1972. GPO-5501-00520), the
use of an application factor of 0. 01 when applied to acute toxic values
is thought to provide an ample margin of safety for certain chlorinated
hydrocarbon pesticides.
Ecological importance of an organism is dependent on the
role the organism plays within the ecosystem and upon its relationship
to the food chain within the aquatic community and to consumers of
aquatic life, including man. Thus, toxicity data for the top carnivores
in a given ecosystem, as well as economically important species such
as trout, salmon, menhaden and shrimp are needed for the development
of a protective criteria level. Toxicity data for organisms such as the
stonefly and Daphnia are of equivalent importance since these organisms
are a food base for higher consumers and are representative of invertebrate
species found in most waters of the United States.
Invertebrate species, such as the stonefly and the Daphnia. are an
indication of the integrity of the aquatic food chain and their presence
may be the controlling factor for the abundance of economically and
-------
recreationally important predators such as trout, bass or pike. While
these fish may not directly consume the Daphnia or stonefly or, in
fact, even inhabit the same waters, these lower order organisms are
representative of the food chain base supporting predators.
Criteria levels, by their nature, are developed to protect aquatic
organisms and consumers of aquatic life from direct toxic effect when
placed on contact with the toxic pollutants; and, to protect from a
more insidious and even greater danger, e.g., chronic effects.
Chronic effects take the form of reproductive failures or the poisoning
of predators consuming food organisms which have bioaccumulated levels
of toxic pollutants as in the case of the brown pelican and consuming
endrin. loaded fish (see Attachment D, Endrin), and a variety of other
physiological effects as discussed in the various documents. Decreases
in aquatic organisms or consumers of aquatic life not always are coupled
to point source discharges of toxic pollutants at concentrations below
acute toxic levels; however, the addition of toxic levels which are not
acutely toxic can achieve the destruction or at least disruption of aquatic
systems by causing reproduction of failure. Hence, the need for application.
factors. The relationships between discharges of toxic pollutants and
effects on important organisms of economic and environmental importance
and consumers of these organisms are well documented in the criteria
documents.
vi
-------
BENZIDIKE
Table of Contents
II.
III.
IV.
V.
VI.
VII.
Prearnble 1
Physical and Chemical Properties ,1
•»
. Toxicological .Data 3
A. Invertebrates ....3
B. Fish 3
C. - Marmials 6
1. Biochemistry of Benzidine ........ 7
2. Physiological Effects .13
D. Humans . . . , . ^ . . . . ;i6
Environmental Fate and Effect .23
Criteria Formulation .33
References Cited 37
Appendix I 49
-------
I. ' PrcMinhL?
R--?nzidine is a proven human carcinogen with the site of
ir.'jmor i r< Ino1: ion h^inq the bladder. Th-^raf ore, in derivinq
an ambient water quality criteria considerations beyond the
aquatic toxicity arf! necessary. As calculated in Appendix
1, l»velr, of benzidine which could result in a cancer risk •
of no nor-"1 than one case per million people exposed, at the
95 percent confidence limit, during a lifetime, if allowed to reach
man through water, are projected to ranqes from 0.05 uq/1 to
1.0 uq/1. A level of C.I ug/1 is recommended based en the
calculated dose/response data, the use of extrapolative
methods to determine a risk level, and the possible
population at risk.
Benzidine has not been studied extensively for toxicity
to aquatic species since the main concern has been its
carcinogenic activity in man. However, initial fcioassays
with benzidine have produced 96 hour TL 's of 2.5 to
areater than 20.0 mq/1 for fish. A bioaccumulation study
with bonzieline has shown that edible tissue residues can
accumulate Uf* tiu'ps the. ambient water level. Use of an 0.01
application factor applied to the 96 hour TL-.. for fish
results in a level of 25 uq/1 which would protect aquatic
life.
TT. Physical and Chemical Properties
Benzidine (4,4'-diaminobiphenyl) is classified as an
aromatic amine. It has the empirical formula C H N and a
\-£ \_£. £.
-------
molecular weight of 184.24 grams (1). BenzifJrie exists as a
grayish-yellow, white, or reddish-gray crystalline powder
with a melting point of 128°C and a boiling point of 400°C
at 740 mm mercury (2). Benzidine's solubilities in water
range from 0.4 g/1 at 12°C to 9.4 g/1 at 10::C. Organic
solvents increase the solubility of benzidir^ with 20 g/1
dissolving in ethyl ether and 200 g/1 in absolute ethanol
(3).
As an aromatic amine, benzidine has an inherent
basicity. Although amines are much weaker bases than
hydroxide ion or ethoxide ion, they are much stronger bases
than water. Aqueous mineral acids or carbon/lie acids
readily convert amines into their salts; aqvieous hydroxide
ion readily converts the salt back into free amines. As
seen in figure 1, benzidine is easily converted to and from
its salt (4).
Figure 1
"X C
. . / s \ \ / /^^\ \ .,,, "
I
Ben/; idine Benzidine Salt
-------
Benzidine can readily undergo diazotization reactions.
This ability to undergo diazotization makes benzidine suited
as a basic building block for dyes. Diazotization is the
reaction of a primary aromatic amine with nitrous acid in
the presence of excess mineral acid to produce an azo
(-N=N-) compound (24) .
The azo compounds are strongly colored, with the color
^
depending upon the exact structure of the molecule. Because
of their color, the azo compounds are of importance as dyes;
about half of the dyes in industrial use today are azo dyes
CO.
III. Toxicological Data
A) Invertebrates
Benzidine' s toxicity to invertebrates was investigated
by Lerr:
-------
carcinogenic activity in man. However, initial bioansays
with benzicline have shown that it is r.ct highly toxic to
fish..
Lemks in 1973 (18) determined 96-hour TL 's for five
50
species of fish. Toxicity data for these five species are
shown in Table 1.
Table 1 (18)
TL Values for Static Biosssay with
50
Benzidine
Species
TL Values (mg/1)
24 hr. US hr. 72 hr. 96 hr,
Flagfish
(Jordanella 'floridae)
Fathead Minnows
(Pimephales promelas)
Red Shiner
(Not pop is lutrensis)
Lake Trout
(flalve Linus n-i
Rainbow Trout
+ 50
+20
+ 20
8.7
+ 20
32.5 . 25
20
10
5.0
ltt.1
+ 20
2. 5
«*.35
10
. 16.2
+ 20
2.5
4.35
-------
The Synthetic Organic Chemical Manufacturer's
Association (SOCMA) in 1975 (19) determined TI. values for
three different species of fish. Their results are listed
as Table 2. ;
Table 2 (19)
96-hour TL Values for Eenzidine
50
96-hour TL Value
Species .__ (mq/1)
Fathead Minnows 2
(Pimephales protnelas)
Emerald Shiners 5
(Notropis atherinoides)
Bluegills 15
{Lepornis macrochirus)
-------
C)
The history of cancer research shows that the
discoveries in man of carcinogenesis ty various chemicals
preceded studies of these substances in experimental
animals. Such was the case with benzidine. Animal studies
were used to gain more information about the carcinogenic
action of benzidine.
As early as 1895 , Rehn (20) related certain bladder.
tumors to exposure to aromatic amines. However, isolation
of the compound or compounds responsible for bladder tumors
was difficult since many of the workers studied were exposed
to an array of compounds. Through selective research the
major suspected carcinogens were identified by Barsotti and
Vigliani in 1949 (21) as beta naphthylamine and benzidine.
Early attempts to induce experimental tumors in animals
with benzidine had been without success (22, 23) , and it was
*~~-*J- *>*-.*- 4 ~l 1 O C r\ *, U -H *- O *-,-T -*- « ^ i, -* 1 / *7 \ ~l*-».-ns"knr^*-w-*4-X-«.'3 *. Vx -» JL.
J i W— C* I . w J_ A JL S ^ J Oi 1U O U £JJL V- i. CT I. CA JL. • I ^ / *-»^-"*'»«^itO t-i C4 C, CTVJL i_ 1 1'Ji l_
benzidine was carcinogenic. Tumors occurred in the liver,
intestine, and the acousti gland in 20 percent of the test
rats injected with a maximum cumulated amount of 1.28g of
technical-grade benzidine. Purified benzidine and benzidine
sulfate were also carcinogenic at a similar dosage level
(total dose: 0.96 and 0.94g, respectively). However, no
-------
tumors occurred in either rabbits or monkeys after injection
of benizidine. Only one case of bladder cancer was found in
dogs fed a total dose of 325g of benzidine in.five years.
No conclusions concerning the carcinogenicity of benzidine
for dogs was drawn from this single occurrence.
1) Biochemistry of Benzidine
x
The failure to obtain bladder tumors with benzidine in
previous experiments was thought to be due to the species of
the animals used, the mode of administration of the
chemical, or the degree of dosage. But, after Spitz's
successful experiment, the importance of the metabolism of
benzidine was recognized. Studies were conducted to
determine the fate of benzidine upon entering the body.
Work by Weber and Heideprim in 1928 (9) and Goldblatt in
1914-7 (8) indicated that benzidine may be oxidized in vivo.
Baker in 1950 (10) demonstrated that oxidation may take
placo to give 3,3'-dihydroxybenzidine. Further work by
Baker and Deighton in 1953 (11) determined the rate of
recovery of benzidine from tissue and the extent of
conjuqarion of the aromatic amino groups. Their studies
showed that 93 percent of the injected dose was recovered as
diazotizable material after U hours and 68 percent after 12
-------
8
hours. Of this, the conjugated forn accounted for 24
percent and 49 percent respectively of the recovered amount.
Sciarini in 1957 (12) identified free benzidine and the
metabolite 3-hydroxybenzidine in the urine of dogs following
intraperitoneal injection of benzidine. Troll and Nelson in
1958 (13) found the major metabolites of fcenzidine in dogs,
rabbits, and guinea pigs to be the sulfate and glucuromide
of 3- hydroxybenzidine.
Sciarini and Meigs in 1958 (14) verified the metabolites
and quantified the constituents in the dog. As shown in
Table 3, only about 2 percent to 9 percent of unmetabolized
benzidine was recovered in dog's urine.
-------
Table 3
Benzidine Recovery from Dog Urine Following Intraperitoneal
Injection at Different Dosage Levels
Dose Administered
(mg/kg Body Wt)
Percent Free Benzidine
Recovered in 24 hours
100
56
25
9
2.0
1.0
0.5
0.25
0.2
0.1
0,05
6.8
4.0
1.7
3.9
9.0
2.7
3.2
6.4
4.0
7.0
2.0
The average free benzidine recovered amounted to about 7
percer.t of the dose administered while the major metabolite
(3-hydroxybenzidine) was approximately 47 percent of the
dose. The feces of the dog were found to contain about 10
-------
10
percent as much benzidine and its metabolite as detected ir.
the urine.
Subsequent investigations by Sciarini and Meigs in 1961
(15) identified other metabolites in mice and benzidihe
workers. The recovery and identification of only a portion
»
of the dose of benzidine administered to the mice was
consistent with the hypothesis that additional metabolites
were present. Through analysis of the urine of workers
engaged in benzidine manufacture, four substances were
identified in the following proportions: free benzidine
3.6-5.6 percent; monoacetylbenzidine 1.6-5.4 percent; diacetylbenzidine
5.1-10 percent; and 3-hydroxybenzidine 78.5-89.7%.
Experiments with substituted benzidines conducted by .
Sciarini and Meigs in 1961 (16) were designed to broaden the
*• _ j
understanding of the behavior of benzidine. The three
disubstituted benzidines used were: dicrthotolidine,
dianisidine, and dichlorobenzidine. Structures of benzidine
and the disubstituted benzidines are shown in figure 2.
-------
11
Figure 2
V.ctf,
,s
NH,
OCH,
,,-•-,. ^\
[c i
x-
f ^
C
/
i
V S
.
-°CW3 V^<
-0/^ ^ ^-
NH^ /x/h^
Benzidine Diorthotolidine Dianisidine Dichlo
Both dianisidine and dichlorobenzidine were found to
tnet-.abclize in the human body along different pathways than
benzidine. Diorthotolidine follows a metabolic route
similar to that of benzidine.
Neumann's study in 1970 (17) of recognized carcinogenic
aromatic amines demonstrated that it is the metabolites or
reactive intermediate stages of benzidine formed during
metabolism whicn are considered to be the carcinogenic
substances.
Hovever, ev*n products such as azo dyes, which are
derived from benzidine, have been shown to undergo metabolic
reduction to free benzidine. Rinde and Troll (63) in 1975,
reportor] on thr metabolic reduction of four benzidine azo
-------
12
dyes in the rhesus monkey. By comparing the excretion of
i
free benzidine after feeding the monkeys either benzidine or
the dyes, it was determined that the metabolic reduction of
the dyes to free benzidine was nearly complete. Yet, P.inde
and Troll stipulate that they do not intend to imply that
all azo dyes are biologically reduced to carcinogens, but
those derived from carcinogenic amines should receive
particular attention.
-------
13
2) Physiological Effects
Concurrent with the investigation of the metabolic
pathways of benzidine, studies were undertaken to
experimentally produce tumors in mammals through exposure to
benzidine.
Boy land et al., in 1954 (25) fed ferrale Wistar rats a
diet containing 0.017 percent benzidine. Chlolagiomas and/
liver-cell tumors were reported, but urinary bladder
carcinomas were not found.
Bonser et al,, (26) injected benzidine into mice,
subcutaneously at a weekly dose of 6rng for 52 weeks.
Hepatonas and lymphomas occurred in the treated animals,
while lymphomas occurred in one-third of the untreated
control animals. Nevertheless, the presence of the
hepatomas caused benzidine to be classified as carcinogenic.
Loumonier and Laquerriere in 1962 (27) studied the
effect of daily applications of benzidine (3 percent in
benzene) painted to the back skin of Wistar rats. After 15
days of treatment, jaundice appeared due to toxic hepatitis.
After a latent period, of 2-1 months, the rats developed
malignant tumors, primarily hepatomata.
-------
m
In-1964, Pliss (20) studied the tu:-origenic properties
of ber.zidine in rats. A total dose of 3GOmg of benzidine
was subcutaneously injected during a 6 rroriths period,
however, there was a considerable loss of animals,
especially females, during the first few months of the
experiment due to acute toxic effects. Only 20 rats (40 percent)
including 5 females survived for 5.5 -or.ths. The deceased
animals showed profound degenerative changes in the liver
"•x
with ultimate necrosis. Hepatomata, cymbal gland tumors,
and sarcomata in the underlying fat at. the site of injection
of ber.zidine were reported in 70 percent of the surviving
rats.
Saffiotti et al., (29) and Sellakuxar ,et al., (30)
investigated the effects of ber.zidine on Syrian golden
hamsters. Benzidine was administered to hamsters in the
diet at concentrations of 0.1 percent for their lifespan.
Hepatomas, liver-cell and chloangiomatous tumors developed,
but none appeared in the bladder.
Griswold et al., in 1968 (31) reported that benzidine
produced breast cancer in female Sprarue-Dawley rats after
intraqastric feeding ranging from 12 to 50 mg/rat during a
30 day period.
-------
15
In 1970, Zabezhinskii (32) reported the results of
benzidine inhalation experiments on rats. The animals were
exposed to a concentration of benzidine in the air of 10-20
mg/cu m for 4 hours 'a day, 5 times per week, for 20 months.
After inhaling benzidine for 13 months, 29 percent of the
rats developed tumors consisting of myeloid leukemia,
fibroadenomata, carcinomata of the mammary gland (male and
female), and hepatoma.
s • ' •
The induction of hepatic tumors in mice by benzidine was
reported by Prokofjeva in 1971 (33). Single weekly doses of
6 mq per animal were subcutaneously injected into 181 mice
(C3HA line) over a period of 8-13 months (total dose:
210-336 mg) . Of the 46 mice that survived for 15-28 months,
31 (67.4 percent) had developed hepatomata. A previous control
colony of C3HA mice showed a 1 percent hepatoma frequency in
untreated animals.
A study was conducted by Vesselir.cvitch, et al.,
reported in 1975 (71), to assess the carcinogenicity of
benzidine hydrochloride in mice. In one carcinogenicity
study, 200 six-week old B6C3F mice were given benzidine
dihydrochloride intermittently by stomach intubation, twice
weekly, in the amounts of 0.5 or 1.0 mg/mouse at each
intubation for 8'4 weeks. All animals were killed at 90
weeks of age, at which time their tumor incidence was
-------
16
evaluated. Benzidine-treated mice developed liver tumors,
Harderian gland tumors, lung adenomas, and lymphoreticular
tumors. A total of 124 of the indicated tumors were
reported for both dose levels.
D) Human
The incidence of bladder tumors resulting from
•\
occupational exposure to aromatic amines was first described
by Rehn in Germany in 1895 (20). Between 1905 and 1932,
bladder tumors in dyestuff workers were reported in
Switzerland (36) , Great Britian (37) , Russia (38) , and
Austria (39). The first cases of this condition in the
United States were described by Ferguson in 1934 (40) .
Subsequently the disease has been recognized in Italy (41) ,
Japan (42) and France (43) .
The noted high occurrence of bladder tumors in the dye
industry was an .established fact, though the etiology of the
disease was still obscure. Early theories implicated
certain intermediate products used in the manufacture of
dyes (44). of those, aniline, benzidine, and alpha and beta
naphthylamines were the most frequently implicated (45, 46,
47, 48) .
-------
17
In 1919, Barsotti and Vigliani (U9) examined 186 workme.n
cystoscopically to follow the evolution of bladder lesions.
Through their study, the following conclusions were made:
1) Aniline has no appreciable carcinogenic potential.
2) Benzidine and beta-naphthylamine have the highest
carcinogenic potential among the aromatic amines
studied.
Scott in 1952 (5) investigated the incidence of bladder
tumors in an English dyestuffs factory*, Sixty-six cases of
bladder tumor were reported. Of the sixty-six cases there
were 30 (23 in the manufacturing section and seven in the
handling section) who were exposed to benzidine, and who had
never been exposed to beta naphthylamine.
In 1954, Case et al., (50) published the results of
their study of workmen engaged in the manufacture and use of
aniline, benzidine, and alpha and beta-r.aphthylamines in the
British chemical industry. They determined the incidence, of
tumors of the urinary bladder among such workers. The data
indicates that the incidence of bladder cancer among persons
exposed to benzidine greatly exceeded that among the general
population. The study also showed bladder cancer to be a
-------
18
fatal disease, only 20 percent of all cases having survived more
than 10 years from the first detection of the disease.
By 1957, based upon the findings cf Case et ajL., (50) ,
the manufacture of benzidine had been accepted as an
occupational hazard by the Association of British Chemical
Manufacturers. That year, plant and operating
recommendations were made by Scott and Williams (51) to
ensure the safe manufacture and use cf products causing
tumors of the bladder.
Loriq term retrospective studies were undertaken in order
to assemble additional information on the occurrence a-nd
natural history of malignant tumors ir. workers exposed to
dyes and dye intermediates.
In 1962, Goldwater et al. (52) studied the incidence of
urinary bladder tumors in workers exposed to
alpha-naphthylamine, beta-naphthylamina and benzidine, to
determine the average incubation period, the average
survival time, and the incidence of ralignant tumors other
than those of the bladder. The population studied consisted.
of 366 male workers in a coal tar dye.factory employed
bet-ween 1912-1962.. Of the 366 workers studied, 76 were
exposed to benzidine alone. Bladder car.cer developed in 17,
or 21.3 percent, of the 76 exposed to benzidine. The
induction of bladder cancer from benzidine had an average
-------
19
incubation period of 18.7 years, calculated from the first
exposure to diagnosis of malignancy. Details of the
calculated incubation periods are given in Table 4.
-------
20
Tnhle '>
Bladder Tumors Related to Years
of Exposure
A) First suggestive
genitourinary
complaints
Years of Benzidine Exposure
Range Mean
9-30
20.7
3) First positive
cystoscopy
H- 31
16.6
C) First diagnosis
of papilloma
U-28
18.
D) First diagnosis
of malignancy
5-33
18.7
-------
21
Following diagnosis of bladder malignancy in workers
exposed to benzidine, the longest period of survival was ten
years, the shortest was less than a year. The mean survival
time between diagnosis of cancer and death was approximately
3 years.
Among the 366 men studied, there were 11 recorded
malignant tumors other than cancer of the bladder, as
follows:
*
Lung 3
Prostate 3
Stomach 2
Brain 2
Lymphoma 1
There was no evidence of any unusual incidence of
malignancies other than those of the urinary bladder.
Rye et a1., in 1970 (53) discussed the use of benzidine
and its congeners, (diorthobenzidine, dianisidine, and
diorthotolidine) as curing agents in polyurethane
production. There was no evidence of cancer developing
among the workers exposed during production of the three
congeners. However, several bladder tumors had been found
-------
22
in man working in a plant that alternated production of
benzidine and its three congeners.
In 1973, Zavon et al. (54) published their findings
associating the manufacture of benzidir.e with a high risk of
developing bladder cancer. They observed a group of 25
workers engaged in manufacturing benzidine over a 13-year
period. Of the 25 workers, three had about one year of
exposure to beta-naphthylaminef three had exposure to
orthotoluidine, and seven had been on dichlorobenzidiae
production at some time. The mean exposure to benzidine was
10.6 years.
Thirteen of the group (52 percent) developed bladder tumors,
some multiple. The study group had an average induction
tine from first exposure to detection of the first tumor of
16.6 years. This time period agrees with previously
calculated induction times (50, 52, 55, 56) .
Ey 197G only one inember of the thirteen bladder cancer
patients had died, which gives a ten-year survival of 92
percent. However, this survival period was expected to
increase because no recurrences or life-threatening
complications were present at that tire. A mean survival of
thre« yoars was reported by Goldwater et al. (52) for
bladder cancer and the ten-year survival of less than 20
-------
23
percent was reported by Case et a1. (50). Increased
survival may be due to a number of causes. Earlier
diagnosis and improved therapy used in these patients are
possibilities.
It is important to note that decreasing the potential
exposure by limiting the'duration of employment on the
process only serves to increase the number at risk.
IV. Environmental Fate and Effect
In 1965, Takemura £t 'al_. (6) published their results of
a survey of the pollution of the Sumida River, in Tokyo,
Japan. Concentrations of aromatic amines found in the water
ranged between 205 to 562 ug/1. Concentrations of the
carcinogenic amine, benzidine, identified by chromatography
and determined quantitatively by the chloramine-T method,
ranged from a maximum of 233 ug/1 to a minimum of 82 ug/1.
The source ot these pollutants was suspected to be from
wastes discharged by dye and pigment factories along the
river. It was considered that the aromatic amines in the
river were not necessarily discharged directly in the
factories' wastewater but possibly were produced by the
reduction of azo-dyes in the wastes by H S or SO in the
river water. Laboratory demonstrations showed that if H S
-------
24
is bubbled for a few minutes through a pure azo dye
solution, aromatic amines are liberated from the azo dyes.
To determine whether benzidine persists in natural
waters for long periods of time or degrades relatively
rapidly, the Great Lakes Laboratory experimentally studied
the degradation of benzidine in lake water under laboratory
conditions (31). Lake Erie water was obtained from the
inlet of the Buffalo city water treatment plant. Sample
aliquots were spiked with initial benzidine concentrations
of 1, 2 or 5 ug/1. The samples were treated under several
conditions: 1 or 2 mg/1 available chlorine, stirred, aerated
and undisturbed and all samples were shielded from light.
Results of the degradation of benzidir.e under various
conditions have been tabulated in Table 5.
-------
25
Table 5
Degradation of Benzidine Under Varying
Conditions
BENZIDINE IN CHLORINATED LAKE WATER (1 mg/1 Cl )
(hrs .)
1/2
u
12
24
48
72
3KNZ. CONC.
in SOLN. A
(ug/1)
0.84
0.48
C.10
N.D.
BEV'Z. CONG.
in SCLN. B
ug/1)
1.90
1.17
1.17
N.C.
BENZ. CONC.
in SOLN. C
(ug/1)
4.42
4.C6
2.17
N.D;
BENZIDINE IN CHLORINATED LAKE WATER (2 mg/1 Cl )
TIK2
(hrs.)
1/2
4
12
2U-
48
72
where
BSNZ. CONC.
in SOLN. A
0.79
0. 39
0.04
N.D.
BENZ. CONC.
in SOLN. B
ug/1)
1.73
1.09
0.31
N.D.
BENZ. CONC,
in SOLN. C
(ug/1)
4.37
4.10
1.87
N.D.
Soln. A = Solution with initial benzidine
concentration of 1 ug/1.
Soln. B = Solution with initial tenzidine
concentration of 2 uq/1.
ooli. C = Solution with irit-ial benzidine
concentration of 5 uq/1.
= Not detectable K0.2 ug/1) .
-------
26
TIME
(hrs.)
1/2
u
12
24
48
72
Table 5 (Coat.)
BENZIDINE IN AERATED LAKE WATER
TIME
(hrs.)
1/2
U
12
24
48
72
BENZ. CONG.
in SOLN. A
(ug/1)
0.83
0.69
0.31
N . D .
-
..
BENZ. CCNC.
in SCL:-J. B
ug/1)
1.9-
l."2
0.72 .
N . D .
-
-
BENZ. (
in SOL;
(ug/
4.61
4.27
2.01
N.D.
—
—
BENZIDINE IN STIRRED LAKE WATER
TIME
(hrs.)
1/2
4
12
24
48
72
BENZ. CONC,
in SOLN. A
(ug/1)
0.90
0.70
0.48
N.D.
-
-
BENZ, CCNC.
in SCLN. B
ug/1)
1.57
1,37
0.50
N.D.
-
-
BENZ. (
in SOD
(ug/!
4.70
4.21
2.15
N.D.
-
-
BENZIDINE IN UNDISTURBED U:XE WATER
BENZ. CONC,
in SOLN. A
(ug/1)
0.89
0.71
0.47
0.02
* • r~»
u . L/.
BENZ. CONC.
in SCLU. B
uc/1)
1,85
. 1.C5
0.75
BENZ. CONC.
in SOLN. C
(ug/1)
4.81
4.03
2.07
N.D.
-------
27
The results show that the decay of ber.zidine in .lake
water is relatively rapid. At all three concentrations,
regardless of the system, benzidine was at or below the
detection limits after 2U hours. However, the experiments
did not indicate the by-products, kinetics and/or mechanisms
of this decay.
.A further investigation by the Great Lakes Laboratory
>.
(57) determined that the rate of degradation of benzidine in
natural waters was near a first-order reaction with a rate
constant of 0.175 per hour.
To further evaluate the fate of benzidine in the aqueous
environment and to define the role of sunlight, a series of
studies were conducted to determine the photodegradation
effect by light (58). The photodegradation effect was
qualitatively determined by comparing the observed and
calculated concentrations at various times during the
reaction. The calculated values were determined by using
the first-order decay equation and a rate constant of 0.175
per hour as previously determined (57). Results of the
experiment have been tabulated in Tatle 6.
-------
28
Table- 6
PHOTODEGRADATION OF 3ENZIDINE IN WATER
Reaction Time
(hours)
P
0.5
1.5
4.0
12.G
24.0
27.0
Series A
Benzidine Concern ration
Observed Calculated
1.0 mg/1
0.73
0.66
0.45
0.087
ND
ND
1.0 mg/1
0. 91
0.77
C.50
0.12
ND
ND
Series B
0.5
1.5
'4.0
12.C
2 4 . C
29
48
53
72
77.0
10.00 mq/1
8.70
0.05
6.45
4.5
2.49
1.80
0.50
C.29
0.018
0.018
10.00 mg/1
9.16
7.69
4 .96
1.22
1.50
C.06
ND
ND
ND
Series C
1.5
a. 0
12.C
24.0
23.0
56.-?
100.00 mg/1
91.95.
86.93
95.61
75.41 .
63.68
52.13
31.15
100.00 mg/1
91.62
76.91
^9. 65
12. 24
1. 50
0.69
ND
ND
MD - Mo-. Detectable
ug/1)
-------
29
For series A the observed values were always less than
the calculated values, indicating that photodegradation
enhanced the rate of degradation of benzidine. For series B
and C, a reddish-brown material precipitated, serving as a
light" screen which diminished the rate of degradation. For
the preliminary studies presented, the data indicate that
for low concentrations of benzidine (less than 1 mg/1) ,
photodegradation enhances the rate of degradation of
benzidine in aqueous solutions.
However, in the photodegradation study (Table 6) , the
dependence of decay on the ambient water temperature was not
studied. The mean temperature of the samples in the
fadometer was 48°C. This temperature is atypically high of
natural waters.
The biodegradation of benzidine was studied by Malaney
et aj.. (35) and Lutin et al_» (59) using Warburg
respirometers which contained samples of activated sludge
obtrH. r.srl from three different wastewater treatment plants.
The studios conducted used sludge concentrations of 2500 and
5000 m.g/1 mixed liquor suspended solids (MLSS) and a
ber.zidine concentration of 500 mg/1. The results showed
th'it benzidine was inhibitory to bio-oxidation by each of
the three sludg.es. The extent of removal of benzidine by
bio-oxidation in an activated sludge aeration tank was found
-------
30
to be insignificant at the concentration studied. It was
postulated that benzidine may be removed partially or
completely by adsorption. However, this mechanism was not
investigated.
A similar study was conducted during 1973 by Ryckman et
al. (60) using Warburg respirometers and two different types
of sludges: An aniline acclimated sludge and a municipal
wastewater treatment plant sludge. The sludge samples were
spiked with two different concentration levels of benzidine,
1 and 5 mg/1. The results showed that benzidine appeared to
be slightly inhibitory to both the aniline acclimated and
non-acclimated municipal sludges. Still, both of the
sludges showed rather large uptakes of benzidine during the
12-hour studies. Approximately 78 and 60 percent uptakes
were noted for the acclimated and non-acclimated sludges,
respectively for the samples with a concentration of 1 mg/1
benzidine. For the samples containing 5 mg/1 benzidine,
uptakes of 64 and 50 percent were observed for the
respective sludges. It was hypothesized that an adsorption
mechanism was involved in removing the ber.zidine. However,
no conclusive proof was obtained to edther support or reject
the hypothesis of an adsorption mechanism.
An investigation of the bio-oxidation of benzidine was
conducted for SOCWA' by International Hydronics, Inc. (61).
-------
31
Studies were conducted to determine if benzidine car. be
biochemically oxidized at a concentration of 50 ug/i- The
fate of benzidine was studied through a series of tescs
u'sing acclimated and non-acclimated sludges. As interpreted
by SOCMA, the tests produced the following results:
1) Benzidine can be biologically cxidized at a level
of 50 ug/1. Oxidation of about 90 percent of the
^> •
compound is accomplished.
2) Air oxidation of benzidine in a sterilized
activated sludge will occur. About 30 percent of
the benzidine concentration, 50 ug/1, can be
reduced in this manner.
'' " • . "S
Ho-vever, neither of these experiments demonstrated '*
oxidation as such (for example, recovery of 14C carbon
dioxide after treatment of labeled benzidine) , but depended
on measuring the decrease in benzidine across the system.
During 1975r Aquatic Toxicology Laboratory of Bionomics
(62) conducted studies to determine the bioconcentration of
14C-benzidine in bluegill populations. The radiometeric
method employed measured only lie-residues. These residues
were not identified to determine if the 14C was in the form
of bor.zinine or a metabolite. During the six week
-------
32
observation period following the ir.treduction of bluegill to
the test units, the cumulative mortality was 6 percent for
fish exposed to 14C-benzidine at a mean water concentration
of 191.02 ug/1, 8 percent for fish exccsed to a mean water
concentration of 98.07 ug/1, and 2 percent for the fish in
the control unit. Bioconcentration factors of 38 times the
mean concentration of 191.02 ug/1 and "U times the mean
concentration of 98.07 ug/1 were four.c! in the edible
portions of the bluegills after the 42 days of exposure.
Following the 42 days of benzidine exposure there was a
14-day depuration period to determine the rate and extent of
the elimination of the 14C-residues. within the first week
of-depuration, the edible portions of fcluegill exposed to
the high treatment level (191.02 ug/1) of 14C-benzidine
eliminated 47 percent of the 14C-residue present at the end
of exposure. By the end of the depuration period, 70
percent of these residues had been eliminated. Within seven
days after commencing depuration, the bluegills exposed to
the low treatment level '98.07 xy/l) or 14C—benzidine
eliminated 42 percent of the 14C-residue present at the end
of exposure. By the end of the depuration period, 73
percent of these 14C-residues had beer, eliminated from the
edible portions of the exposed bluegills.
-------
33
Biqconcentration stu'dios show the edible portir-s of
fish bioconcentrate 14C-residues from labeled benziciir.e to a
factor of 44 times the water concentration, thus exhibiting
an exposure route to man. However, following a 14 day
depuration period, 73 percent of the bioccncentrated
-residues from labeled benzidine had beer, eliminated.
V. Criteria Formulation
Benzidine has not been studied extensively for toxicity
to aquatic species since the main concern has been its
carcinogenic activity in man. However, initial bioassays
with benzidine have produced 96 hour TL *s of 2 mg/1 for
the fathead minnow and 5.0 for the emerald shiner (19) and
2.5, 4.35, and 7.4 mg/1 for the red shiner, lake trout and
rainbow trout respectively (18). A bioaccumulation study
with benzidine has shown that edible tissue residues
accumulate 44 times the ambient water level. Use of an 0.01
application factor applied to the 96 hour TL for the red
shiner results in a level of 25 ug/1 which would provide an
ample margin of safety for aquatic life. ,
However, the principal objective in regulating benzidine
under soction 307 (a) of the Federal Water Pollution Control
Act is to minimize the risk of human exposure to benzidine
which would result from its discharge to the navigable
-------
waters in light of its proven carcincaenicity. See 38 Fed.
Rag. 243U3 (September 7, 1973). Accordingly, the
recommended ambient water criterion is to provide adequate
protection for human health as well as for aquatic
organisms. It must minimize any unreasonable risk of human
exposure.
Observations of the incidence of human bladder cancer
v
due to occupational exposure have demonstrated benzidine to
be a human carcinogen, , To date no minimum effect level or
maximum safe level has been established for humans.
Extrapolation through mathematical modeling is one approach
the Agency has used to determine levels of exposure which
are likely to result in a low risk, or incidence, of'cancer.
The two most widely utilized models for this approach are
the "one-hit" method (70) and the Mantel-Bryan method (64).
In developing and applying these formulae, animal studies
' tt
were used to experimentally demonstrate the relationship
between various dose levels and the frequency of resultant
tumors. Based on these experimental cose/effect data, it is
possible to theoretically extrapolate the dose for low
effect levels when a carcinogen is administered at a given
daily dose over a lifetime.
There are several caveats which apply in using this type
of mathematical modeling. First, this type of modeling is
-------
35
not accepted by some sectors of the scientific community.
It is not sufficiently advanced or proven as to justify use
of this approach alone as the sole basis for establishing
Hsaf eM levels of exposure. If one is to have a high degree of
confidence in. predicting the likely dose/effect level, a larger
statistical data base than that which is available should
be utilized. However, the risk extrapolative rr^thods rray
furnish an indication of risk levels associated with a
particular exposure level or exposure levels which result in
a predetermined level of risk, e.g., one cancer in a million of the
exposed population.
With these qualifications in mind, each of the two
mathematical models was applied to the three best available
sets of dose/effect data (7, 33, 71). The calculations are
set forth in Appendix 1, and are summarized in the following
table.
-------
36
Table 1 shows levels of benzidine which, based on a daily
.exposure over an entire lifetime, would produce no more than
one cancer in one million population at risk at the "95
percent confidence limit:
Table 1
• .
Extrapolative Method
Observed Dose-Pesponse
Data One-Hit (7C) Mantel-Bryan (64)
Spitz (7) 0.056 ug/1 0.608 ug/1
Prokof jeva (33) 0.048 ug/1 0.2-12 ug/1
Vesselinovitch (71) 0.061 ug/1 0.843 ug/1
The Agency does not possess data as to the presence of
benzidine in waters at the levels indicated in the above
chart. Such measuring as has been clone fails to show
benzidine in the water columns at detectable levels. As
noted earlier, the Agency lacks evidence of" direct human
exposure to benzidine from drinking water on the aquatic
environment.
these ci re urn stances, and in light of the
previously expressed caveats concerning use of mathematical
modeling, the Agency has concluded that: a level of 0.1 ug/1,
which approximates the low side of the middle range of the
exposure levels reflected in the mathematical modeling
exercises (0.05-0. 84 ug/1), should provide the requisite
margin of safoty for human health as well as for aquatic
organisms. This level is recommended as the ambient water criterion.
-------
37
REFERENCES CITED:
1. Weast, R.C., editor, 1972. Handbook cf chemistry and
Physics, 53rd Ed.
2. Standen, A., editor, 1972. Kirk-Cthmer Encyclopedia of
Chemical Technology. 3:108.
• . ."*• x
3. Stecher, P.G., editor, 1968. The Merck Indes, 8th Ed.
tt. Morrison, R.T. and R.M. Boyd, 1972. Organic Chemistry,
2nd Ed. • •
5. Scott, T.S., 1952. The incidence of bladder tumours in
4
a dyestuffs factory. Brit. J. Industr. Med., f1
9:127-132. .
6. Takemura, N., T, Akiyamo, and C. Nakajima, 1965. A
survey of the pollution of the Sumida River,
especially on the aromatic amines in the water.
Int. J. Air Wat. Poll., 9:665-670.
7. Spitz, S., W.H. Kaguigan, and K. Cofcriner, 1950. The
carcinogenic action of benzidine. Cancer,
3:789-80*4.
-------
38
8. Goldblatt, H.W., 1947. Occupational cancer of the
bladder. Brit. Med. Bull., 4:405-417.
9. Weber, H., and C. Heiderprim, 1923. Zur Kenntnis 'des
verhaltens aromatischer ainine in organismus.
Zeritralblf Gewerbehyg, 5:269-272.
10. Baker, R.K., 1950. The carcinogenic activity of
dihydroxoybenzidine. Acta Uniori internat. contra
Cancer, 7:16-51.
11. Baker, R.K., and J.G. Deighton, 1953. The metabolism of
benzidine in the rat. Cancer Res., 13:529-531.
12. Sciarini, L.J., 1957. 3-Hydroxyber.zidine, a metabolite
of benzidine. Arch. Biochem. and Biophys.,
71:437-441.
13. Troll, W., and N. Nelson, 1958. Sruaies on aromatic
aminea. I. preliminary observations on
metabolism. Industrial hygiene Journal,
19:499-503.
-------
39
14. Sciarini, L.J., and J.W. Meigs, 1958. The
biotransformation of benzidine, (4,
4'-diaminobiphenyl), an industrial carcinogen, in
the dog. A.M.A. Arch. Indust. Health, 18:521-530.
15. Sciarini, L.J. and J.W. Meigs, 1961. The
biotransformation of benzidine II. Studies in mouse
and man. Arch. Environ. Health, 2:423-428.
16. Sciarini, L.J., and J.W. Meigs, 1961. Biotransformation
of the benzidines in. studies en diorthotolidiner
dianisidine, and dichlorobenzidine: 3,
3 *-disubstituted congeners of henzidine. Arch.
Environ. Health, 2:108-112.
17. Neumann, H.G., 1970. Uber die a ktiverung
krebserzuegender aromatischer amine im
stoffwechsel. Klin. Wschr., 48:959-966.
IS. LemkG, A.E., 1973. Static bioassays v?ith benzidino,
unpublished data. National Water Quality
Laboratory.
19. 3OCMA, 1975. Acute fish toxicity. Second Submission by
SOCMA Benzidine Task Force to Environmental
Protection Agency.
-------
no
20. Rehn, L., 1895.. Arch. Klin. Chir., 50:588.
21. Barsotti, M., and E.G. Vigliani, 19U9. Bladder^lesions
from aromatic amines. Med. Lavoro, 40:129-138.
22. Brenblum, I. and GnM. Bonser, 1937. xperimental
investigation of "aniline cancer". J. Indust.
Hygiene and Toxicol., 19:86-92.
23, Hartwell, J.L., 1941. Survey of compounds which have
been tested for carcinogenic activity. U.S. Public
Health Service.
24. Hawley, G.G., editor, 1971. The condensed chemical
dictionary, 8th Ed.
25. Boyland, E., J. Harris, and E.S. Horning, 195U. The
induction of carcinoma of the bladder in rats with
acetamidoflurene. Brit. J. Cancer, 8:617-654.
26. Bonser, G.H., D.B. Clayson, and J.'-J. Jull, 1956. The
induction of tumours of the subcutaneous tissues,
liver and intestine in the rrcuse by certain
dyestuffs and their intermediates. Brit. J.
Cancer, 10:653-667.
-------
27. Laumonier, R., and R. Laguerriere, 1962. Changes in DNA
content during hepatic regeneration. Clinical and
experimental data. Rev. Int. Hepat, 12:633-654.
28. Pliss, G.,B., 1964. Carcinogenic properties of
benzidine. Vop. Onkol., 10:529-532,
29. Saffiotti, U., F. Cefis, R, Montesano, and A.R.
v
Sellakumar, 1967. Induction of bladder cancer in '
hamsters fed aromatic amines. In: Bladder Cancer,
A Symposium. Edited by K.F. Lampe.
30. Sellakumar, A.R., R. Montesano, and U. Saffiotti, 1969.
Aromatic amines carcinogenicity in hamster. Proc,
Am. Assoc. Cancer Res., 10:78. •
31. Griswold, D.P., A.E. Casey, E.K. Weisturger, and J.H.
Weisburger, 1968. The carcinogenicity of multiple
intragastric doses of aromatic and heterocyclic
nitro or amino derivatives in young female'
sprague-dawley rats. Cancer Research, 28:924-933.
32. Zabezhinski, M.A., 1970. Effectiveness of inhalation.as
•
a method of administration of atomized carcinogens.
Byull EKsp Biol. Med., 69:72-74.
-------
42
33. Prokofjeva, O.G., 1971. Inducticr. of hepatic tumors in
mice by benzidine. Vop Onkol., 17:61-64.
34. SOCMA, 1975. Exhibit III, Benzidine degradation.
Submission by SOC.MA Benzidine Task Force to
Environmental Protection Agency.
35. Malaney, G.W. , P. A. Latin, J.J. Citulka, and L.H.
Hickerson, 1967. Resistance of carcinogenic
organic compounds to oxidation by activated sludge.
*
J. Water Poll. Control Fed., 12:2020-2029.
36. Schedler, R., 1905. Zur Kasuistik der blasentumoren bei
f acharbeitern. Inaug. diss., Easel.
37. Ross, H.C., 1918. Occupational Cancer. J. Cancer
Research, 3:321-356. .
38. Rosenbauin, N., and J. Gottlich, 1926. Neubildungen der
arnbleise be: textilarbeiterr.. Gigiona trudn,
30
39. Sch'.iellar, H., 1932. Blasentumoren bei anilinarbeitern
Ztschr. Urol., 26:284-286.
-------
43
40. Ferguson, R,5., G.H. Gehrman, D.I-'.. Gay, L. Anderson, and
V. Washburn, 1934. Symposium on aniline tu.--.ors of
the bladder. J. Urol., 31:121-126.
41. di Maio, G., .1936. Tumore della vesica tfa amine, detti
da anilina. Urologia, 3:239-242.
42. Nagayo, M., and R. Kinosita, 1940. Seme significant
-»
features of cancer incidence in Japan.' Yale J.
Biol. and Med., 12:301-303.
43. Billiard-Duchesne, J.L., 1946. Les amine-tumeurs de la
vessie. J. d'durol., 53:401-422.
44. Leuenberger, S.G., 1912. Die unter dem einfluss der
synthetischen farbenindustric. beobachtete
geschwulstent-wicklung. Beitr. Klin. Chir., 80:208.
45. Curchmann, F., 1920. Statistische erhebungen uber
blasenrumoren bei arbeitern in der chemischen
Industrie. Gewerbehyg, 8:145-169.
i
46. Nassauer, M., 1919. Uber bosartige blasengeschwulste
bei arbeitern der organisch-chemischen.
grosindustric. Atsch.f. Pathol., 22:353.
-------
44
47. Oppenheimer, R., 1927. Uber die bei arbeitern
chemischer betriebe beobachtetian erkrankungen des
harnapparates. Atch.f. urol . Chir., 21:336.
48. Schwerin, 1920. Blasengeschwulste fcei arbeitern in
chemischen betrieben. Abl. f. Gewerbehyg., 8:64.
49. Barsotti, M. ,- and E.G. Vigliani, 1952. Bladder lesion^
i » V
from aromatic amines. Statistical considerations
and prevention. Arch. Indust. Hyg. Occup. Med.,
50. Case, R.A.M., M.E. Hasker, D,B. McDonald, and J.T.
Pearson, 1954. Tumors of the urinary bladder in
workmen engaged in manufacture and use of certain
dyestuff intermediates in the British chemical
industry. Part I. The role of aniline, benzidine,
alpha-naphthylainine, and beta-naphthylamine. Brit.
J. Ind. Med., 11:75-104.
51. Scott, T.S., and M.H.C. Williams, 1957. The control of
industrial bladder tumours. Brit. J. Industr.
Med., 14:150-163.
-------
52. Goldwater, L.J., A.J. Rosso, and M. Kleinfeld, lr:>5.
Bladder tumors in a coal tar dye plant* Arch.
Environ. Health, 11:814-817.
53. Rye, W.A., P.P. Woolrich, and R.F. Zanes, 1970. Facts
and myths concerning aromatic diamine curing
agents. J. Occupt. Med. , 12:211-215.
x.
^ x
54. Zavon, M.R., U. Goegg, and E. Bingham, 1973. Benzidine
exposure as a cause of bladder tumors. Arch.
Environ. Health, 27:1-7.
55. Ehrlicher, H., 1958. Benzidine in arbeitsmedizinischer
sicht. Zentrabl Arbeitsmed, 8:201-207.
j
'
56. Scott, T.S., 1962. Carcinogenic and Chronic Toxic
Hazards of Aromatic Amines.
57. SCCMA, 1975. Determination of degration rate for
benzidine under different laboratory conditions.
Second submission by SOCMA Benzidine Task Force for
the Environmental Protection Agency.
-------
46
58. SOCMA, 1975. Degradation of benzidine by exposure to
• simulated. June sunlight. Second Submission by
SOCMA Benzidine Task Force for the Environmental
Protection Agency.
59. Lutin, P.A., J.J. Cibulka, and G.W. Malaney, 1966.
Oxidation of selected carcinogenic compounds by
activated sludge. Proc. 21st Ind. Waste Conf.,
Purdue Univ.
60. Ryckman, Edgerley, Tomlinson and Associated, Inc., 1974.
Fate of benzidine in the aquatic environment: A
scoping study.
61. SOCMA, 1975. Biological treatability of benzidine.
Third Submission by SOCMA Benzidine Task Force; for
the Environmental Protection Agency.
62. SOCMA, 1975. Exposure of fish to lUC-benzidine:
Accumulation; distribution, and elimination of
14C-residues. Third Submission by SOCMA Benzidine
Task Force for the Environmental protection Agency.
63. Rinde, E. and W. Troll, 1975. Metabolic reduction of
benzidine azo dyes to benzidine in the rhesus
monkey. J. Nat. Cancer Inst., 55:181-182.
-------
47
64. Mantel, N., and W.R. Bryan, 1961. "Safety" testing of
carcinogenic agents. J. Nat. Cancer Inst., 27:455*
470.
65. Freireich, E.J., E.A. Gehan, D.P. Rail, L.H. Schmidt,
and H.E. Skipper, 1966. Quantitative comparison of
toxicity of anticancer agents in mouse, rat,
hamster, dog, monkey, and man. Cancer Chemotherapy
Reports, 50:219-244.
66. Huggins, C., G. Briziarelli, and H. Sutton, Jr., 1959
Rapid induction of mammary carcinoma in the rat and
the influence of hormones on the tumors. J. Exptl.
Med., 109:25-41.
67. Shay, H., E.A. Aegerter, M. Grunenstein, and S.A.
Komarov, 1949. Development of adenocarcinoma of
the breast in the wistar rat following gastric
instillation of methylcholanthrene. J. Natl.
Cancer Inst., 10:255-266.
68. Altraan, P.L. and D..S. Dittmer, editors, 1971. Section
III - development and growth. Biology Data Book,
2nd Ed.
-------
69. U.S. Environmental Protection Agency. Statement .of
Basis and Purpose for the national Interim Primary
Drinking Water Regulations.
70. Hoel, D.G., D.W. Gaylor, R. L. Krischstein, U. Saffiotti,
and M.A. Scheiderman, 1975. Estimation of risks of
irreversible, delayed toxicity. J. Toxicology and
Environ. Health, 1:133-151.
71. Vesselinovitch, S.D., K.V.N. Rao, and N. Mihailovich,
1975. Factors modulating benzidine carcinogenicity
bioassay. Cancer Research, 35:2814-2819.
72. Earnest, R. 1970. Effects of pesticides on aquatic
animals in the estuarine and irarine environment.
Unpublished data. In: Annual Progress Report.
Bureau of Sport Fisheries and wildlife. U.S. Dept.
of the Interior.
7"^ VT*-»--» M IQ^T A /•»! »-f-£> •f-/~\-VT/1>->4-tr r\f Cr/-\rv/i /->>~rT=»r^Ty-»
-^ » i \ »-* — <^ ^ t. * • ^. ^ \J *. * *»Vxi^.v-v^ «^«^«*^_VU.u.J' »_»*. •*•*•*.* t \ • \~, VX.A-^-flAllJ.V'
insecticides to three species of salmonids and the
threespine stickleback. Trans. Amer. Fish. ,Soc.,
90 (3) :26H-268.
-------
U9
Appendix 1
Various mathematical models can be used for deriving a
level from the experimental dose/effect data which will
ensure a limit of no more than one cancer in one million
(1x10 ) of the human population at risk at a 95 percent
confidence level. The mathematical models employed to
calculate this level for benzidine were the Mantel-Bryan
probit model (64) and the one-hit model (70).
The Mantel-Bryan probit model is based on the assumption
that the susceptibility to the carcinogenic effect of a
chemical is distributed log normally among the population,
and that the incidence of cancer will decrease with
decreasing dose at a rate of no less than one normal
deviation (probit) per 10-fold decrease in dose.
The one-hit model assumes that a single molecule can
cause cancer. If the molecules are distributed at random,
the probaoility of inducing ccmccr, as corrected for
repeated cancer production, is proportional to the dose.
-------
50
The relationship of an observed response, P, at a h:.u
dose, d, is expressed by the equation:
P=l-e-kd
where:
P=observed response
d=dose
k=ccnstant, calculated from experimental
dose-response data.
With the calculated constant, the dose level at the limit of
one cancer per one million at risk can then be calculated, i
The dose-carcinogenic response studies reviewed for
extrapolation to calculate human dose levels at the limit of
no more than one cancer per one million at risk were
reported by Griswold, et a 1«, (31) using Sprague-Dawley
rats; Spitz, et a1., (7) using Sherman rats; Prokofjeva (33)
using C3HA mice; the experiment conducted by Spitz, et al.,
(7) with dogs as reported in Bonser, et
-------
51
test animals employed. Because of the defined sensitivity
of the female Sprague-Dawley rats and the statistically
insignificant total of ten test animals used, the Griswold
data was rejected for mathematical extrapolation.
Spitz's published conclusion (7) concerning their
experiment with dogs was that no conclusion concerning the
carcinogenicity of benzidine could be drawn from their
•<.
reported one case of papillary carcinoma from seven test
dogs. Six years after the publication of the test results
from Spitz's experiment, Bonser, et al., in 1956 .(26)
reported a personal communication frorr. Spitz stating that a
total of three dogs developed papillorr.as and carcinomas of
the bladder. However, because the necessary specific data
has not been published, this experiment could not be used in
the extrapolative methods.
The experiments reported by: Spitz, et al,, (7) ;
Prokofjeva (33); and Vesselinovitch (71), although not using
random populations,- are the best available data for
calculation of human dose levels corresponding to no more
than one cancer per one million at risk.
The dose administered to the experimental animals is
first converted to an equivalent human dose. A conversion
method is base-1 upon the observation that for acute toxicity
-------
52
of a number of anti-cancer agents, parameters such as the ID
and the maximum tolerated dose are quantitatively
proportional in mice, rats, hamsters, dogs, monkeys and man
(65 >. This proportionality is based on a weight
relationship, which is calculated by:
human dose = Average human weight \ 2/3 ^^
[Average animal weight) x amlnaj- dose
In the work reported by Spitz, et al., (7) a total of
385 Sheman strain rats were injected subcutaneously in the
bac< with a weekly dose of 15 mg of benzidine throughout the
lif a of ';he animals until grossly obvious tumors appeared.
DOS-SB were not administered if the animals suffered loss of
weirht or obvious illness. Due to the high mortality of
botl the control and experimental series in the first 200
day of the experiment, only those animals which survived
200 or more days were counted as "at risk." Out of the
ori< inal 385 animals injected with benzidire, a total of 206
sur-'ived 20( days or more. There were three tumor types in
the test an.mals caused by the benzidine. These tumor types
wer*: hepa-; omas; carcinomas arising in the specialized
auditory sebaceous glands; and adenocarcir.crra of the colon.
A total of 100 test animals produced turners of the three
ind .cated types with an average incubaticr. time of 222 days.
The administered dose of benzidine was 15 mg per week or
2.1 mg/day. The average weight of the experimental animals
-------
53
was 265 gm, while the reported average weight of a human is
70kg (68) . Calculating the equivalent human dose according
to the 2/3 power of the weight ratio, as previously
described/ is demonstrated in the following equation:
' sexvalent h^an a,se . feg ^ff f* x dose Ja anin*ls
human dose = ^_2yi_y/3 x 2.14 ng/day
human dose = 88.16 mg/day
Assuming no sources of benzidine intake other than the
average intake of two liters of water per person per day
(69) , this dose corresponds to a concentration of 44.08 rag/1
in water.
The :jicidence of tumors from benzidine was 0.4JO (48%) ,
with the upper 95 percent confidence limit bei ig 0. 543
(54.3%) . According to the one hit model, with a dose (d) of
44.08 mg/1, a response (P) of 0.543 (54.3%) at the upper 95
percent confidence limit, the constant k is calculated as?
P = 1 - e~kd
k = -ln(l - P) = -ln(0.457) = 0.7831
d 44.08 44.08
k = 1.78 x 10~2
Deriving a concentration which, at the 95% confidence
1 jrdt, will result in no more than one cancer per million of
- 6
t le population at risk (1x10 ) yields:
d = -ln(l - P) _ -In (0.999999) 10"6
k 1.78 x 10~2 1.78 x 10~2
d = 5.62 x 10~5 mg/1
-------
54
At a 95 percent confidence level, the limit of one
cancer per million at risk corresponds to a concentration Of
0.056 ug/1, as extrapolated from the Spitz data calculated
by the one-hit model.
Using the Mantel-Bryan method, a risk level of one in
one million at risk is equivalent to a probit of -4.75.
While the upper 95 percent confidence limit of the observed
response corresponds to a probit of 0.108. Therefore,
extrapolating from the experimental response of 0.543
(54.3%) to a response of no more than one cancer per million
at risk corresponds to a probit change from 0.108 to -4.75
equal to -4.86.
According to the ten-fold reduction in dose per one
probit, a probit change of -4.86 corresponds to a
concentration of benzidine which will produce no more than
one cancer per million in the population at risk of: .
concentration = 44.08 mq/1 x 10~4-86
— /i n no ^r, /•»-. n 10 v "in — 5\
— -»T»\/»-»III-^'J-/V ij_*jyiyvJL\// ••
6.08 x 10~4 mq/1
0.6C8 ug/1
At the 95 percent confidence limit, the concentration
which will result in no more than one cancer per million at
risk is equal to 0.608 ug/1, as extrapolated.by the Mantel-
Bryan method using the Spitz data.
-------
55
In the experiment reported by Prokofjeva, a total of 181
C2 iA mice were given subcutaneous injections of benzidine at
a single weekly dose of 6 mg per mouse over a pericxl of 8-13
mciths. Hepatomas appeared in 31 of 46 surviving test mice
v;i^±i the first occurence being within 15 to 16 months. In a
deiined control population, a one percent hepatoma frequency
wa; observed. The administered dose of benzidine
(6 mg/week) was equal to 0.856 mg/day. Calculating the
eq ivalent human dose, assuming the average weights of 30
gr ms for an adult C3HA. mouse (68) and 70 Kg for an adult
hu an (68), the following dose is:
quivalent human dose = (j^*^ x 0.856 mg/day
human dose = 150.48 mg/day
Assuming no sources of benzidine intake other than the
average intake of two liters of water per person per day
(69), this dose corresponds to a concentration of 75.24 mg/1
in water.
The incidence of tumors from benzidine v.-as 0.674 (67.4%)
with the upper1 95 percent confidence limit equal to 0.788
(7c .8%). Calculating the dose corresponding to the level of
-------
56
no more than one cancer per one million at risk using the
one-hit model, the rfollowing level is calculated:
Solving for k according to the Prokof jeva data:
P=l-e-kd
k = -ln(l - 0.788) _ 1.55117
75.24 mg/1 75.24
k = 2.06 x 10~2
Solving for the dose (d) when the response (P) is no more
•chan one cancer per million at risk (10 ) :
d = -ln(l - 10"6X _ 10"6 _,
2.06 x 10 ^ " 2.06 x 10"^
d = 4.85 x 10~5 mg/1
At the 95 percent confidence level a limit of no more
than one cancer in one million at risk corresponds to a
concentration level of 0.048 ug/1 as calculated by the one-
hit theory based upon the Prokof jeva data..
The corresponding probit for 0.788 (78.8%), which is the
95 percent confidence limit of the observed response, equals
0.80 probits. The probit change from an observed incidence
of 0.788 (78.8%) to one in one million at risk equals -5.55,
a probit change from 0.80 to -4.75.
According b> the ten-fold reduction in dose pir one
probit change, i probit change of -5.55 corresponds to a
concen -.ration o: benzidine which will produce no nore than
one cc.-icer per :iillion in the population at risk of:
concentration = 75.25 mg/1 x 10~5*55
75.24 mg/1 x (2.82 x 10~6)
2.12 x 10~4 mg/1
concentration = 0.212 ug/1
-------
57
At the 95 percent confidence limit, the concentration
which will result in no more than one cancer per million at
risk is equal to 0.212 ug/1, as.extrapolated by the Mantel-
Bryan method using the Prokofjeva data.
The study by Vesselinovitch, et_ al. (71), consisted of
tin integrated series of studies to assess the
carcinogenicity of benzidine dihydrochloride in mice. In
one carcinogianicity study, 200 six-week old B6C3F mice were
given benzid-'me dihydrochloride intermittently by stomach
intubation, rwice weekly, in the amounts of 0.5 or 1.0
mg/mouse at oach intubation for 84 weeks. All animals were
killed at 90 weeks of age, at which time their tumor
incidence was evaluated. Benzidine-treated mice developed
liver t\mors, Harderian gland tumors, lun : adenomas, and
lymphoreticular tumors. A total of ]24 of the indicated
tumors were reported for both dose levels. The average dose
level is calculated to be 0.428 mg/day/mcuse. Calculating
the equivalent hu nan dose, assuming the cverage weights of
30 grams for an adult B6C3F mouse (f8) ar.3 70 kg for an
adult human (68) , the following dose is:
equivalent hurrvm dose =^™_|2.J x 0.428 mg/day
human dose = 75.24 irg/day
Assuming no sources of benzidine inta:e other than the
average intaVe of two liters of water per person per day
-------
58
(69), this dose corresponds to a concentration of 37.62 mg/1
in water.
The incidence of tumors from benzLdine was 0.413 (41.3%)
with the upper 95 percent confidence .Limit equal to 0.460. .
The dose corresponding to a level of no more than one cancer
per one million at risk, using the one-hit modal, is
calculated as follows:
Solving for k according to the Vesseilinovitch data:
v _ -ln(l - P) = -ln(0.540) = 0.61619
d 37.62 37.62
k = 1.64 x 10~2
Solving for the dose (d) when the response (P) is no more than
one cancer per million at risk (10~^):
d = -Ind - 10"*) _ 10"6 .,
1.64 x 10"? ~ 1.64 x 10 '•
d = 6.10 >: 10~5 mg/1
A ; the 95 percent confidence level a limit of no more
than one cancer in one million at risk corresponds to a
concentration level of 0.061 uc/1 as calculated by the one-
hit tieory based upon the Vesselinovitch data.
Tie corresponding probit for 0.460 (46.0%), which is the
upper 95 percent confidence limit of the observed response,
equal.1- -0.100. The probit change frcrr an observed incidence
of 0.--60 (46.0%) to one in one million at risk equals -4.65,
a proi it change from -1.00 to -4.75.
-------
59
According to the ten-fold reduction, in dose per one
probit change, a probit change of -4.65 corresponds to a
concentration of benzidine which will produce no more than
one cancer per million in the population at risk of:
concentration = 37.62 mg/1 x 10~4'65
= 37.62 mg/1 x (2.24 x 10~5)
= 8.43 x 1CT4 mg/1
concentration = 0.8U3 ug/1
At the upper 95 percent confidence limit, the
concentration which will result in no more than one cancer
per one million at risk is equal to 0.843 ug/1, as
extrapolated by the Mantel-Bryan method using the
Vesselinovicch data.
Table 7 presents a summary of the calculated
concentrations which will result in nc more than one cancer
per million at the upper 95 percent confidence limit as
extrapolated by the Mantel-Bryan and the one-hit methods,
from dose-response data as reported by Spitz, et a1., (7),
Prokofjcv-a (33), arid Veaaeiiriovitch, et _al. (71).
-------
60
Table 7
Summary of Calculated Concentrations
Extrapolative Method
Observed Dose-Response
Data * One-Hit (70) Mantel-Bryan (64)
Spitz (7) 0.056 ug/1 0.608 ug/1
Prokofjeva (33) 0.048 ug/1 0.212 ug/1
Vesselinovitch (71) 0.061 ug/1 0.843 uq/1
-------
GLOSSARY
Acutely toxic: Causing death or severe damage to an organism by
poisoning during a brief exposure period, normally ninety-six
hours or less.
Anadromous fishes: Irishes thai spend a part of their lives in seas
or lakes, but as<-end rivers and streams al. certain intervals to .
spawn. Kxamples are sturgeon, shad, salmon. Lroul, and
'^
striped bass. • „• -
Application factor: The ratio of the safe concentration to the lethal
concentration as determined for potential aquatic pollutants
administered to species of interest.
Bioaccumulation (Bioconcentration): The phenomenon wherein elements
or compounds are stored in living organisms because elimination .
E1
• •' . fc
fails to match intake. r
('urcinogenic : Producing < 'anci-r.
Catadromous fishes; Fishes that Iced and grow in fresh water, but
return to the sea to 'spawn. The best example is the American
eel.
Chronically toxic; Causing death or damage to an organism by
poisoning during prolonged exposure, which, depending on the
organism tested and the test conditions and purposes, may range
from several days, to v/eeks, months, or years, or through a
reproductive cycle.
-------
EC50; The concentration at which a specified effect is observed
under the It-si conditions in a specified time in fifty percent of
1.1 ie. organisms tested. lOxamples of specified effects arc hcmor--
rhaging, decreased feeding, dilation of pupils, and altcr"d
swimming patterns.
Epilimnion: That region of a body of \vater that extends from the
surface to the top of the thermocline and does not have a permanent
temperature stratification. . ".
(•'low-through bioassay: An assay system in which aquatic species
arc: exposed to toxicants in a constantly flowing system, and where
th<: toxicant is replenished continuously or dise.ontinuously.
Hardness (Water): The concentration of the polyvalent metallic ions
dissolved in water. Unually it is reported as the equivalent
concentration of calcium carbonate (CaC'O ).
:i
llyperplasia: Abnormal multiplication or'increase in the number
of normal cells in normal arrangement in a tissue.
Hypolimnion: The region of a body of water that extends from the
bottom of the thermocline to the bottom of the water body and
is essentially independent of most surface phenomena.
-------
LC25: The concentration ol' a toxicant that is lethal (fatal) to twenty-
five percent of the organisms tested under the test conditions in
a specified time.
l/lfiO: The concentration of a toxicant which is lethal (I'alaP to
fifty percent of the organisms tested under tin; lest conditions
in a specified time. |l is virtually identical with Thru and Tl,f>0.
LUfiO: The dose of a toxicant that is lethal (fatal) to fifty percent
* '
of the organisms tested under the test conditions in a specified
time. A dose is the quantity actually administered to the
organism and is not identical with a concentration, which is the
amount of toxicant in a unit of test medium rather than the
amount ingested by or administered to the organism.
Liter (I).- The volume occupied by OIK; kilogram of water at a pressure
0 i
ol 7f>0 mm of mercury and a temperature ol 4 ('. A liter is
1. 057 quart.
Methylmercury; Mercury which has been methylated, usually through.
some biological agent, such as bacteria.
Microgram per liter (ug/1): The concentration at which one millionth
of a gram (one microgram) is contained in a volume of one liter.
Where the density of solvent is equal to one, one ug/1 is equiva-
lent to one part per billion (ppb) or one microgram per kilogram
(ug/kg).
-------
lUicrogram per kilogram (ug/kg): The concentration at which one
millionth of a gram (one micrograrn) is contained in a mass of
one kilogram. A kilogram is 2. 2046 pounds.
Milligram per kilogram {mg/kg): The concentration at which one
thousandth of a gram (one milligram) is contained in a mass «>f
one kilogram. A gram contains 1000 milligrams.
Milligram per liter (nig/1): The concentration at which one milligram
is contained in a volume of one liter. Where the density of the
solvent is equal to one, one mg/1 is equivalent to one part per
million (ppm) or one milligram per kilogram (mg/kg).
Milliliter (ml): A volume equal to one thousandth of a liter.
Nanogram per liter (ng/1): The concentration at which one billionth
•of a gram (one nanogram) is contained in a volume of one liter.
'Where the density of the solvent is equal to one, one ng/1 is
equivalent to one part per trillion or one nanogram per kilogram
(ng/kg).
Neoplastic: Describing any new and abnormal growth, such as a tumor.
Part per million (ppm): A concentration in which one unit is contained
in a total of a million units. Any units may be used (e.g., weight,
volume) but in any given application identical units must be used
(e.g. , grams per million grams or liters per million liters).
Where the density of the solvent is one, one part per million is
equivalent to one milligram per liter.
-------
1'arts per thousand (o/oo): A concentration at which one unit is
contained in a total of a thousand units. The rules for using
this term are the same as those for parts per million. Normally.
this term is used to specify the salinity of estuarine or sen waters.
Piscicide; A substance used for killing fish.
Static bioassay: A bioassay in which the toxicant is not renewed during
the test.
Thermocline; That layer in a body of water where the temperature.
difference is greatest per unit of depth. It is the layer in which
the drop in temperature is 1 {.'.. or greater per meter of. depth.
TLrn - Median Tolerance Limit; The concentration of a test material
at which fifty percent of the test animals are able to survive-
under test conditions for a specified period of exposure. It is >
virtually synonymous with LC50 and TL50.
TJ,T)0: Synonymous with TLm and virtually synonymous with LC50.
Turnorigenic: Causing or producing tumors.
------- |