External Review Draft
October 1977
BENZENE HEALTH EFFECTS
ASSESSMENT
NOTICE
This document is a preliminary draft. It has been
released by EPA for public review and comment
and does not necessarily represent Agency policy.
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, D.C. 20406

-------
External Review Draft
October 1977
BENZENE HEALTH EFFECTS
ASSESSMENT
NOTICE
This document is a preliminary draft. It has been
released by EPA for public review and comment and
does not necessarily represent Agency policy.
U.S. Environmental Protection Agency
Office of Research and Development
Washington, D.C. 20460
U S EPA LIBRARY REGION 10 MATERIALS

-------
CONTENTS
Page
ACKNOWLEDGMENTS	iv
SUMMARY AND CONCLUSIONS	1
1.	INTRODUCTION	4
2.	BENZENE METABOLISM AND CYTOGENETIC EFFECTS	6
Benzene Metabolism	6
Cytogenetic Effects	18
3.	CHRONIC BENZENE TOXICITY IN ANIMALS	37
Exposures by Inhalation	38
Exposure by Other Routes	44
Evaluation and Comments	5 0
4.	BENZENE TOXICITY IN MAN	64
Introduction	64
Pancytopenia	6 6
Leukemia	90
Other Benzene-Associated Disorders	124
Summary	130
iii

-------
ACKNOWLEDGMENTS
This document was prepared by EPA's Office of Research
and Development with extensive help from a team of con-
sultants led by Bernard D. Goldstein, M.D. Major contribu-
tions were by Carroll A. Snyder, Ph.D., Robert Snyder,
Ph.D., and Sandra R. Wolman, M.D. The views represented in
this document are those of the EPA and not necessarily those
of the consultants.
The final document will in addition incorporate, as
appropriate, comments and contributions from many sources,
and especially those from EPA's Scientific Advisory Board.
iv

-------
SUMMARY AND CONCLUSIONS
This report presents the research findings on benzene
toxicity relevant for assessing human health risks at en-
vironmental exposure levels. The principal conclusions to
be drawn from this report are:
1.	Benzene exposure by inhalation and other exposure
routes is strongly implicated in three pathological condi-
tions that may be of public health concern at environmental
exposure levels:
1)	leukemia, especially acute myelogenous leukemia
2)	pancytopenia (including aplastic anemia)
3)	chromosomal aberrations.
2.	The epidemiological data, from occupational ex-
posure studies, argue convincingly that benzene is a human
leukemogen. The exposure data in these studies do not allow
a scientific derivation of a dose-response curve. Most
studies in which exposure levels were determined involved
exposures in the range of 100 to 500 ppm, though in some the
benzene concentrations were lower.
3.	The data do not indicate that any population seg-
ment is particularly susceptible.
1

-------
4.	Currently there is no convincing evidence that
benzene causes neoplasias, including leukemia, in animals.
Failure to induce leukemia in animals could be due to an as-
yet-unknown cocarcinogen required to evoke the leukemogenic
response initiated by benzene.
5.	Hematotoxicity, particularly pancytopenia, has been
observed in both humans and animals, following exposure to
benzene. The toxicity does not follow exposure to other
compounds such as toluene and xylene commonly associated
with benzene environmentally.
6.	Humans who develop hematologic abnormalities due to
benzene exposure have a greatly increased probability of
developing leukemia and aplastic anemia, a finding con-
sistent with the thesis that benzene is leukemogenic.
7.	Long-term occupational exposures of workers to
benzene at levels as low as 20 ppm but generally at levels
greater than 100 ppm have resulted in various signs of
hematotoxicity.
8.	Two effects, as yet unconfirmed, of potential
significance have been reported at occupational exposure
levels of 3 to 15 ppm. The effects are 1) an increase in
red blood cell levels of deltaaminolevulinic acid, a
precursor in the meme biosynthetic pathway, and 2) a de-
crease in the mean serum complement of the blood.
2

-------
9. Available data from studies where measurements
ranged from 25 to 150 ppm strongly suggest that chromosome
breakage and rearrangement can result from chronic exposure
to benzene. These aberrations have been observed to persist
in lymphoid and hematopoietic cells after removal from
benzene exposure. Since a favored mechanism for leukemia
development is somatic mutation, the persistence of chro-
mosomal aberrations, coupled with clinical observations of
chromosomal abnormalities in human leukemic cells, support
the thesis that benzene is a leukemogen. A dose-response
relationship has not been demonstrated for benzene-induced
chromosome aberrations. This lack may result from varia-
tions in individual susceptibility.
10.	Benzene toxicity probably occurs via a toxic metab-
olite .
11.	In animals, benzene accumulates in lipoid tissue
such as fat and bone marrow, and benzene metabolites con-
centrate in the liver and bone marrow. The concentration of
metabolites in the bone marrow exceeds that in the blood.
12.	The accumulation of benzene metabolites in bone
marrow along with the coincidental covalent bonding of
benzene to solid residues of bone marrow is consistent with
a phenomenon of toxico- and carcinogenesis shared by many
other chemicals.
3

-------
SECTION 1
INTRODUCTION
There is substantial evidence that concentrations of
benzene encountered in the work place (in the United States
and elsewhere) have caused diseases of the blood and bone
marrow in general (e.g., blood dyscrasia, pancytopenia) and
leukemia in particular (especially acute myelogenous leuke-
mia) . Because current policy of the Environmental Protec-
tion Agency (EPA) states that there is no zero risk level
for carcinogens, benzene has been listed by EPA under Sec-
tion 112 of the Clean Air Act as a hazardous air pollutant.
As an aid in determining what regulatory action (if
any) should be taken by EPA on benzene, three reports have
been prepared:
1.	A health effects assessment,
2.	An environmental exposure assessment, and
3.	A risk assessment based on the data in the first
two assessments.
This report is the health effects assessment; it is
largely a review and evaluation of the scientific literature
relevant to determining the human health effects of environ-
4

-------
mental exposures to benzene. Most of what is known con-
cerning the effects of benzene on human health has been
learned by studies of persons exposed to benzene in the
workplace. Virtually no information is available that
describes the health effects of nonoccupational exposures of
the general populace to benzene. Our evaluation of poten-
tial environmental health effects, then, must be based upon
what we know of the mechanisms of benzene toxicity and its
genetic implications and of the effects of benzene on animals
and on human beings. This report is structured accordingly.
Section 2 introduces some of the major biomedical
concepts that are pertinent to assessment of the health
effects of benzene. Following a brief discussion of benzene
metabolism in animals and in humans, the cytogenetic effects
of benzene are considered, particularly its effects on
chromosomes.
The major portion of the report deals with assessments
of benzene toxicity in animals (Section 3) and in man (Sec-
tion 4). These latter analyses focus on two forms of ben-
zene-induced disorders: 1) pancytopenia, defined as the
diminution of all formed elements in the blood, and 2)
leukemia, defined as a proliferation and accumulation of
mature and immature white blood cells (leukocytes) in blood
and/or in bone marrow, leading to the impairment of normal
function.
5

-------
SECTION 2
BENZENE METABOLISM AND CYTOGENETIC EFFECTS
BENZENE METABOLISM
Metabolism in Animals
Most of the benzene that enters the body is excreted
3 42
via the lungs in exhaled air. ' Study of the distribution
of benzene and its metabolites in animal organs shows that
free benzene accumulates in lipoid tissue such as fat and
bone marrow. High concentrations of benzene metabolites can
be observed in liver tissue and in bone marrow. It is
particularly significant that the concentration of metabo-
3
lites in bone marrow exceeds that in blood. Repetitive
administration of benzene leads to accumulation of both
benzene and its metabolites in these organs and to covalent
binding of benzene metabolites to liver and to solid resi-
• w	53
dues in bone marrow.
The metabolic pathway of benzene in liver is shown in
54
Figure 1. The initial step appears to be a reaction
mediated by the mixed-function oxidase. This enzyme is
inducible, so that pretreatment with benzene, phenobarbital,
or 3-methylcholanthrene can increase the rate of benzene
6

-------
Benzene (100%)
phenyl mercapturic
ocid (0.5%)
glutathione
<
NHAc ®P°V
transferase
0
Jahh
benzene
oxide
O
^ Expired unchanged (40%)
benzene glycol (03%)
CH.—CH
iooH
hydroquinol (5%)
r^^iOH
HoO'
PAPS,
sulpho-conjugates
^jOSOjK
potassium phenylsulfafe
(50-100%)
epoxide
hydrase
spontaneous

phenol (23-50%)
-O"-
OH
(dehydrogenase)
catechol (3-25%)
r«^N0H
Cis-Cis
muconic acid
CCOOH
COOH.
Trans-Trans
muconic ocid (13%)
COOH
+C08 +Hp

OH
Conjugations *-
hydrosyhydroquinol (03%)
.UDPG
glucuronic-conjugales

alkaline salts
elimina

0-CHt(CHOH),CHCOfK

AHH»orjl hydrocarbon hydiwylow
UDPC»undinfl diptaiptatt gtvcuren^
Iransferaso
WPS * 3'-photphtf-
5'-phoiphoiuHat«
ed in urine
phenyl glucuronide
(0-50%)
Figure 1. Metabolic pathway of benzene in liver.

-------
13 44 51
metabolism. ' ' The direct product of the interaction
of benzene with mixed-function oxidase is probably an arene
oxide that is highly reactive. It can spontaneously re-
arrange to form phenol, undergo enzymatic hydration followed
by dehydrogenation to form catechol or a glutathione deriva-
tive (phenylmercapturic acid), or bind covalently with
cellular macromolecules. Formation of hydroquinone or of
trihydroxylated compounds probably is the result of several
reactions with hydroxylating enzymes.
Benzene Metabolism in Man
Most metabolic studies of benzene in man have been
concerned either with uptake and excretion of unchanged
benzene via the breath or with measurements of benzene
40
metabolites m urine. Nomiyama and Nomiyama exposed
volunteers to a series of solvent vapors and found that
among six subjects exposed to benzene at 52 to 62 ppm for 4
hours, retention of benzene in the respiratory system de-
creased and then became constant after 3 hours at 30.2
percent of the inspired dose. There was no distinction
attributable to sex of the subjects. Excretion, as measured
in exhaled air after removing the subject from the benzene-
laden atmosphere, was about 16.8 percent. Net uptake, i.e.
the sum of uptake and excretion, was 46.9 percent. These
8

-------
39
authors went on to show that when the logarithm of the
benzene concentration in expired air was plotted against
time, the excretion pattern described a hyperbole that could
be expressed mathematically and that yielded three rate con-
stants to describe the phenomenon. The subjects continued
to excrete benzene in the exhaled air for as long as 15
18
hours. Hunter, in studies of people exposed to 100 ppm
benzene, detected benzene in the expired air 24 hours later.
He suggested that measurements of benzene in expired air
could be used to estimate benzene content of the inspired
air by extrapolation.
Phenol content of the urine is often measured after
benzene exposure. Maximum concentrations are thought to
18
occur within 2 hours after exposure. The major conjugated
form appears to be ethereal sulfate until phenol levels of
the urine reach 400 mg/liter, at which point glucuronide
w'	50
begins to appear.
57
Teisinger et al, who exposed humans to benzene at 100
ppm for 5 hours, reported that 46 percent of the dose was
retained. Of that amount, 61 percent was recovered as
phenol, 6.3 percent as catechol, and 2.4 percent as hydro-
quinone. In these studies the majo~ monohydroxyiated metab-
olite and the two major dihydroxylated metabolites observed
42
by Parke and Williams in rabbits were also observed in
man.
9

-------
Relationship of Benzene Metabolism to Benzene Toxicity
42
Since Parke and Williams suggested in 1954 that a
metabolite of benzene is responsible for benzene toxicity,
3 7
evidence to support that hypothesis has mounted. Nomiyama
demonstrated that inhibition of benzene metabolism protected
3
rats against benzene-induced leukopenia. Andrews reported
that when benzene metabolism was inhibited with toluene the
subjects were protected against benzene-induced reduction of
red cell production. Animals have been protected against
benzene toxicity when pretreated with phenobarbital, ^
probably because phenobarbital stimulates benzene metabolism
in liver, which leads to detoxification and thereby reduces
the amount of benzene available for formation of the toxic
agent in bone marrow. The specific metabolite that produces
benzene toxicity has not yet been identified, but likely
candidates are benzene oxide, catechol, and hydroquinone, or
the corresponding semiquinones.
The demonstration that reduction of red cell count dur-
ing benzene treatment is accompanied by accumulation of
benzene'metabolites in marrow and coincidental covalent
53
binding of benzene to solid residues of marrow suggests a
phenomenon in toxico- and carcinogenesis shared by a variety
of other chemicals, such as acetaminophen,^ bromobenzene,^
32	35	22
hydrazine derivatives, parathion, and many others.
10

-------
Although further studies are required to prove the
hypothesis, it seems likely that benzene, like many other
chemicals, exerts its toxicity by formation of a toxic
metabolite.
11

-------
REFERENCES: Benzene Metabolism
1. Aksoy, M.( Dincol, L., Erdem, S., Akgun, T. , Dincol, G.
Details of blood changes in 32 patients with pancytopenia
associated with long-term exposure to benzene. Brit.
J. Industr. Med., 29:56-64, 1972.
2.	Amiel, J.-L. Essai negatif d1induction de leucemies
chez les souris par le benzene. Rev. Franc. Etudes
Clin. Biol., 5:198-199, 1960.
3.	Andrews, L.S., Lee, E.W., Witmer, C.M., Kocsis, and
Synder, R. Biochem. J., 26:293, 1977.
4.	Boje, H., Benkel, W., Heiniger, H.J. Untersuchungen
zur leukipoese im knocherimark der ratte nach chronischer
benzol-inhalation. Blut, 21:250-257, 1970.
5.	Brandino, G. Osservazione istologische nel'intossicazione
acute e croniche da benzolo. Gazz. Med. Lomborda,
81:141, 1922.
6.	Cheng, S.C. Amer. J. Hyg., 11:449, 1930.
7.	Deichmann, W.B., MacDonald, W.E., Bernal, E. The
hemipoietic tissue toxicity of benzene vapors. Tox.
Appl. Pharm., 5:201-224, 1963.
8.	Diwan, B.A., and Meier, H. Can. Lett., 1:249, 1976.
9. Drew, R.T., Fouts, J.R., Harper, C. The influence of
certain drugs on the metabolism and toxicity of benzene.
IN: Symposium on Toxicology of Benzene and Alkyl-
benzenes, Braun, D., ed., Industrial Health Foundation,
Pittsburgh, pp. 17-31, 1974.
10. Drew, R.T., Fouts, J.R. The lack of effects of pre-
treatment with phenobarbital and chlorpromazine on the
acute toxicity of benzene in rats. Tox. Appl. Pharm.,
27:183-193, 1974.
12

-------
11.	Gerarde, H.W., Ahlstrom, D.B. Toxicologic studies on
hydrocarbons. XI. Influence of dose on the metabolism
of mono-n-alkyl derivatives of benzene. Tox. Appl.
Pharm., 9:185-190, 1966.
12.	Laskin, S., Goldstein, B.D., Snyder, C.A. Unpublished
observations, 1977.
13.	Gonasun, L.M., Witmer, C., Kocsis, J.J., Snyder, R.
Benzene metabolism in mouse liver microsomes. Tox.
Appl. Pharm., 26:398-406, 1973.
14.	Harris, C., Burke, W.T. Am. J. Path., 33:931, 1957.
15.	Hartmen, H.A., Miller, E.C., Miller, J.A., Morris, F.K.
Can. Res., 19:210, 1959.
16.	Hough, V.N., Guinn, F.D., Freeman, S. Studies on the
toxicity of commercial benzene and of a mixture of
benzene, toluene and xylene. J. Industr. Hyg.,
26:296-306, 1944.
17.	Huggins, C.B., Sugiyama, T. Proc. Natl. Acad. Sci.,
55:74, 1966.
18.	Hunter, C.G. Solvents with reference to studies on the
pharmacodynamics of benzene. Proc. Roy. Soc. Med.,
61:913-915, 1968.
19.	Ikeda, M. Enzymatic studies on benzene intoxication.
J. Biochem., 55:231-243, 1964.
20.	Ikeda, M., Ohtsuji, H., Imamura, T. In vivo suppres-
sion of benzene and styrene oxidation by co-administered
toluene in rats and effects of phenobarbital. Xenobiotica,
2:101-106, 1972.
21.	Jenkins, L.J., Jr., Jones, R.A., Siegel, J. Long-term
inhalation screening studies of benzene, toluene, o-
xylene, and cumene on experimental animals. Tox. Appl.
Pharm., 16:818-823, 1970.
22.	Jollow, D.J., Kocsis, J.J., Snyder, R., Vainio, H.
(Ieds.), Biological Reactive Intermediates, Plenum
Press, New York, 1977.
13

-------
23.	Jollow, D.J., Mitchell, J.R., Potter, W.Z., Davis,
D.C., Gillette, J.R., Brodie, B.B. J. Pharm. Exp.
Ther., 187:195, 1973.
24.	Kirschbaum, A., Strong, L.C. Influence of carcinogens
on the age incidence of leukemia in the high leukemia F
strain of mice. Cancer Res., 2:841-845, 1942.
25.	Kissling, M., Speck, B. Chromosone aberrations in
experimental benzene intoxication. Helv. Med. Acta.,
36:59-66, 1972.
26.	Kissling, M., Speck, B. Further studies on experimental
benzene induced aplastic anemia. Blut Z Gesamte Blut
Forsch, 25(2):97-103, 1972.
27.	Laerum, O.D. Reticulum cell neoplasms in normal and
benzene treated hairless mice. Acta Path Microbiol.
Scand. 81:57-63, 1973.
28.	Latta, J.S., Davies, L.T. Effects on the blood and
hematopoietic organs of the albino rat of repeated
administration of benzene. Arch. Path., 31:55-67,
1941.
29.	Lefe, E.W., Kocsis, J.J.., Snyder, R. Acute effects of
benzene on Fe incorporation into circulating erythrocytes.
Tox. Appl. Pharm., 27:431-436, 1974.
30.	Li, T.W., Freeman, S., Gunn, F.D. J. Physiol.,
145:158-166, 1945.	~~
31.	Lignac, G.O.E. Die benzolleukamie bei menschen und
weissen mausen. Klin. Wschr., 12:109-110, 1932.
32.	Mitchell, J.R., Nelson, S.D., Snodgrass, W.R., Timbress,
J.A. IN: Biological Reactive Intermediates, D.J.
Jollow, J.J. Kocsis, R. Snyder, H. Vainio (editors),
p. 271, Plenum Press, New York, 1977.
33.	Moeschlin, S., Speck, B. Experimental studies on the
mechanism of action of benzene on the bone marrow
(radioacutographic studies using thymidine). Acta
Haemat., 38:104-111, 1967.
34.	Nau, C.A., Neal, J., Thornton, M. Arch. Env. Health,
12:382, 1966.
14

-------
35.	Neal, R.A., Kamatakii, T., Lin, M., Ptashne, K.A.,
Dalvi, R.R., Pooge, R.E. IN: Biological Reactive
Intermediates, D.J. Jollow, J.J. Kocsis, R. Snyder, H.
Vainio (editors), p.320, Plenum Press, New York, 1977.
36.	Nomiyama, K. Studies on the poisoning by benzene and
its homologues (6) oxidation rate of benzene and
benzene poisoning. Med..-J. Shinshu Univ., 7:41-48,
1962 .
37.	Nomiyama, K. Experimental studies on benzene poisoning.
Bull. Tokyo Med. Dental Univ., 11:297-313, 1964.
38.	Nomiyama, K., Minai, M. Ind. Health, 7:54, 1969.
39.	Nomiyama, K., Nomiyama, H. Int. Arch. Arbeitsmed.,
32:85, 1974.
40.	Nomiyama, K., Nomiyama, H. Respiratory retention,
uptake and excretion of organic solvents in man.
Benzene, toluene, n-hexane, trichloroethylene, acetone,
ethyl acetate, ethyl alcohol. Int. Arch. Arbeitsmed.
32:75-83, 1974.
41.	Ogui, T., Nakadate, M., Odashima, S. Can. Res.,
36:3043, 1976.
42.	Parke, D.V., Williams, R.T. Detoxication. XLIX.
Metabolism of benzene containing (C^) benzene.
Biochem. J., 54:231-238, 1954.
43.	Reich, C., Dunning, W.F. Cancer Res., 3:248, 1943.
44.	Saito, R.U., Kocsis, J.J., Snyder, R. Effect of
benzene on hepatic drug metabolism and ultrastructure.
Tox Appl. Pharm, 26:209-217, 1973.
45.	Santesson, C.G. Ueber chronische vergiftungen mit
steinkohlentheerbenzin; vier todesfalle. Arch. Hyg.
Berlin, 31:336-376, 1897.
46.	Schalm, O.W., et al. Veterinary Hematology, 3rd
Edition, Lea & Febiger, Philadelphia, 1975.
47.	Secchi, P. Ricerche ematologiche nelle intossicazione
acute e chroniche da benzolo. Rif. Med., 30:995, 1914.
15

-------
48.	Selling, L. Benzol as a leucotoxin. Studies on the
degeneration and regeneration of the blood and hemato-
poietic organs. John Hopkins Hospital Reports,
17:83-142, 1916.
49.	Shay, H. Gruenstein, M.G., Marx, H.E., Glazer, L.
Can. Res. 11:29, 1951.
50.	Sherwood, R.J. Benzene: The interpretation of moni-
toring results. Annals of Occupational Hygiene,
15:409-421, 1972.	—
51.	Snyder, R., Uzuki, F., Gonasun, L., Bromheld, E.,
Wells, A. The metabolism of benzene in vitro. Tox.
Appl. Pharm., 11:346-360, 1967.
52.	Snyder, R., Kocsis, J.J. Current concepts of chronic
benzene toxicity. CRC Critical Reviews in Toxicology,
pp. 265-288, June 1975.
53.	Snyder, R., Lee, E.W., Kocsis, J.J., Unpublished, 1977.
54.	Snyder, R., Lee, E.W., Kocsis, J.J., Witmer, C.M. Life
Sciences (in press), 1977.
55.	Steinberg, B. Bone marrow regeneration in experimental
benzene intoxication. Blood, 4:550-556, 1949.
56.	Svirbely, J.L., Dunn, R.C., von Oettingen, W.F. The
chronic toxicity of moderate concentrations of benzene
and of mixtures of benzene and its homologues for rats
and dogs. J. Industr. Hyg. Toxicol., 26:37-46, 1944.
57.	Teisinger, J., Jiserova-Bergerova, V., Kudrna, J. The
metabolism of benzene in man. Pracov. Lek., 4:175-188,
1952.
58.	Uyeki, E.M., Ashker, A.E., Shoeman, D.W., Bisle, T.U.
Toxicol. Appl. Pharm., 40L49, 1977.
59.	Ward, J.M., Weisburger, J.H., Yamamoto, R.S., Benjamin,
T., Brown, C.A., Weisburger, E.K. Long-term effect of
benzene in C57BL/6N mice. Arch. Environ. Health,
30:22-25, 1975.	— -
60.	Weiskotten, H.G., Schwartz, S.C., Steensland, J. Med.
Res., 33:127, 1915.
16

-------
61. Weiskotten, H.G., Schwartz, S.C., Steensland, H.S. The
action of benzol. The deuterophase of the diphasic
leukopenia and intogen-antibody reaction. J. Med.
Res., 35:63-79, 1916.
62. Weiskotten, H.G., Schwartz, S.C., Steensland, H.S. J.
Med. Res., 41:425, 1920.
63.	Wolf, M.A., Rowe, V.K., McCollister, D.D., Hollingsworth,
R.L., Oyen, F. Toxicological studies of certain
alkylated benzenes and benzene. AMA Arch. Ind. Health,
14:387-398, 1956.
64.	Zampaglione, N., Jollow, D.J., Mitchell, J.R., Stripp,
B., Hamrick, M., Gillette, J.R., J. Pharm. Exp. Ther.,
187:218, 1973.
17

-------
CYTOGENETIC EFFECTS
Concepts
Mutagens and Carcinogens - Benzene is believed to
affect chromosomes, and chromosomal aberrations have been
sought as indications of a biologic response to benzene for
logical reasons. Somatic mutation has long been accepted as
a critical event in the initiation or maintenance of malig-
nant change, although the concept is not unchallenged.
Focus on sites of genetic damage is based partly on observa-
tions of the prolonged delay from the time of exposure to a
carcinogen until the advent of malignancy, such delay being
consistent with perpetuation of the original damage in the
genetic system. Further, many lines of evidence indicate
that most, if not all, carcinogens are mutagens.
Rapid, convenient, accurate, and inexpensive systems
for mutagen testing are available for evaluation of point
mutations in prokaryotic cells;"'" nevertheless, the assess-
ment of damage to mammalian chromosomes is probably more
directly relevant to estimations of human health hazards
from mutagens. If a cell shows sufficient chromosome altera-
tion that further cell division is interrupted, then from a
reproductive point of view that cell is d^ad and the damage
is toxic. If, on the other hand, the chromosome alteration
does not interfere with cell division and can be replicated,
18

-------
then it constitutes a mutation, a structural change in the
genome that presumably alters cell function. Chromosomal
breaks, which may be repaired, are not mutational events (in
the sense of being heritable). However, each occurrence
increases the probability of formation of a structural
aberration and therefore of a mutation.
Investigations aimed at evaluating effects of benzene
have appropriately concentrated on changes in cell nuclei,
metabolism of deoxyribonucleic acid (DNA), cell division,
and chromosome alterations. All these constitute direct
measures of changes in quantity, structure, organization, or
function of the cellular DNA. Moreover, some of these
changes are heritable and imply permanent changes in the
genome of the affected cell.
Clastogens and Mitotic Poisons - The use of chromosome
studies to monitor possible environmental mutagens should
not be limited to evaluations of chromosome-breaking or
"clastogenic" effects on cells arrested in metaphase. If
the cells are analyzed without conventional pre~reatment
with mitosis-arresting agents or hypotonic solutions, abnor-
malities in the anaphase can be identified. These include
multipolar mitoses, imperfect or unequal separation of
chromosomes, and bridges interfering with reconstitution of
the daughter nuclei. Some abnormalities may be detectable
19

-------
only in cells recovering from the effects of a chemical.
During exposure the affected cells may be totally blocked
from entering mitosis. Thus, in evaluating the potential
action of a chemical as a chromosomal mutagen the investiga-
tor must look for both clastogenic and antimitotic effects.
The latter may be especially important in chemicals that do
not induce point mutations.
The effectiveness of benzene as a mitotic poison has
been amply demonstrated. Decrease in DNA synthesis has
2-4
occurred in cultured human cells and in bone marrow of
5-10
rats and rabbits after treatment in. vivo.	The total
numbers of nucleated cells, and, in some cases, the mitotic
indices have declined. Inhibition of cell proliferation has
been shown most often by decrease in uptake of radioactive-
labelled thymidine, a DNA precursor. Although these may not
be the most sensitive of indices, they are clearly and
directly relevant to cell survival and reproductive fitness.
Furthermore, both numerical and structural chromosome aber-
rations have been described that could be interpreted as
either toxic or mutational damage. These include loss or
gain of parts of chromosomes, whole chromosomes, or chro-
mosome sets, in addition to exchanges that result in morpho-
logically aberrant chromosomes.
20

-------
Anaphase studies on human cells have not yet been
reported but are under way in several laboratories. Morishima
and his colleagues have described appropriate conditions for
testing human material."'""'"
Cytogenetic Aberrations in Leukemia - The assumption that
chromosomal mutation is etiologically important in the
development of leukemia has been strengthened by observa-
tions of abnormalities in human leukemic cells. The close
association of the Ph"*" translocation with chronic myelog-
12 13
enous leukemia is well-known, ' and specific chromosomal
abnormalities have been reported with other forms of leuke-
mia."^	These abnormalities appear to be specific to each
disease entity, confined to the leukemic cells, and clonal
(indicating a probable single-cell origin). Therefore, it
is clearly important to investigate the actions of potential
leukemogens with particular emphasis on their ability to
cause site-specific chromosomal lesions. It is, however,
even more likely that the initial damage caused by most
carcinogens is nonspecific, causing a genetically more
variable population of cells. This, in turn, increases the
probability that an abnormal proliferative state will arise
(or be selected).
Cytogenetic Studies of Animals
Studies of benzene effects have been conducted in many
species, including rats, mice, rabbits, and newts. These
21

-------
studies have included whole-animal exposures and effects of
benzene on cells .in vitro; they have been based on either
acute or chronic and repeated exposures. The results of
such studies are difficult to evaluate since they differ not
only in the biologic end-point chosen, but also in species,
routes of administration, and dosage. Since few of the
studies have involved inhalation exposure, their relevance
to problems of human disease may be questioned.
Unpublished studies by Wolman et al evaluated chro-
mosomal findings in rats chronically exposed to 300 or 100
ppm benzene. Within 10 weeks there was a striking and
persistent increase in chromosome breaks and aneuploidy
(deviation from the normal diploid chromosome number) in the
bone marrow of treated (300 ppm) animals. The increase
following 100 ppm exposure was not as great and not of clear
statistical significance.
Increased chromosome breakage in several species has
been reported. Rats exposed to benzene sabcutaneously over
a period of 12 days showed highly significant increases in
chromosome aberrations of bone marrow cells over untreated
17
and toluene-treated controls. Classification of both gaps
and breaks as aberrations complicates interpretation of
these findings (since gap rates vary more with the inter-
preter and with preparation) and inflates the aberration
22

-------
rate. For example, although significant increases i.n aber-
rations were also found in the toluene-treated controls in
this study, the benzene-treated group was the only one in
which breaks were more common than gaps as aberrations.
Exchange figures such as might result from abnormal repair
after breakage (i.e. ring forms, translocations, and di-
centrics) were rarely seen. Another, more acute exposure
(2.0 ml benzene/kg body weight for 12 to 72 hours) in rats
18
produced similar findings. Increased numbers of chromatid
breads were found at almotet every exposure interval, al-
though the responses of individual animals varied consid-
erably. Chronic exposure to injection of 0.2 mg/kg per day
in rabbits (up to 18 weeks of treatment) also resulted in a
high frequency of aberrations; since less than 15 percent of
the aberrations reported were breaks, the significance of
9
this study is not established. Again, exchange figures,
dicentrics, and hyperploid cells were rare. In each of
these animal-exposure studies only a single dosage of
benzene was used. Thus, although different exposure times
in different species can induce increases in chromosome
aberrations, there is no clear evidence for a dose-dependent
response to benzene exposure. Furthermore, none of these
studies presents data suggestive of mutational rather than
toxic damage. Very few experiments have addressed the
23

-------
question of direct interference with benzene-induced abnor-
malities and possible therapeutic routes.
Studies of dividing erythroblasts taken from the am-
19 20
phibian newt (Molge vulgaris L.) ' are of particular
interest because of their demonstration of anaphase abnor-
malities. Young animals injected with water-saturated
solutions of benzene were bled 6 to 12 hours later and a
drop of tail blood was used for coverslip culture. At the
time of sampling, 38 percent of mitoses were arrested in
late metaphase. Another 28 percent showed evident anaphase
abnormalities, of which 20 percent were migration arrests, 3
percent were anomalies of numerical distribution, and 3
percent were anaphase bridges. The remaining 2 percent
showed small subgroups of chromosome condensations outside
the two poles of the newly forming nuclei. Observations
over several hours showed that these mitotic abnormalities
resulted in unequal nuclear divisions, polynucleated cells,
and atypical nuclei. Prophase and early metaphase anomalies
were never found at the doses used in these studies (up to
54 mg of benzene per animal).
Cytogenetic Studies of Man
Experiments - A few experimental observations have been
made on the responses of cultured human cells to addition of
benzene to the culture medium. Increased incidence of
24

-------
breaks and gaps was observed in leukocytes (white blood
cells) and cancerous cells after brief exposure to 1.1 or
-3	2-4
2.2 x 10 M benzene in vitro.	At the higher dose a
decrease in DNA synthesis interfered with clear correlation
of dose to the incidence of breaks. These findings were
21
considered to be toxic damage. Another experiment in
which peripheral blood lymphocytes stimulated by phytohemag-
glutinin (PHA) were exposed to benzene during 72-hour cul-
ture revealed both numerical and structural alterations in
the treated cells. Aneuploidy was seven times more frequent
in the treated populations than in controls, and chromosome
breakage was seen in 11 percent of the treated cells as
compared with 1 percent in the controls.
Chromosome Studies and Hematologic Disease - In con-
trast to the paucity of experimental data, there is an abun-
dance of reports on chromosome studies in exposed popula-
tions and of case reports on leukemia patients. The case
reports are particularly difficult to evaluate and compare.
Some individuals were exposed to benzene vapor above per-
22
missible levels, but in many cases the exposure levels
were unknown. The total periods of exposure ranged from
23
brief and acute to as long as 22 ye^rs. The times between
exposure and the development of disease or death also varied
greatly. Most of all, the endpoints of disease were not
25

-------
comparable. The various reports include diagnoses of acute
22
intoxication, death with massive bleeding and extramedul-
22	23
lary hematopoiesis, benzene leukemia, acute myeloid
24	25	26
luekemia, acute erythroblastosis, erythroleukemia, ac-
27 28	29
quired aplastic anemia, ' acute lymphoblastic leukemia,
myelofibrosis, and chronic myelogenous leukemia. Chro-
mosome abnormalities have been present in industrial workers
21
in association with hematologic pathology or pancyto-
32
penia.
Prior blood transfusions, the use of PHA-stimulated
lymphocytes in some cases and bone marrow in others, and the
lack of chromosome breakage rates in controls also hamper
interpretation of the results of chromosome studies. Never-
theless, certain trends appear amid this mass of data.
22 23 29 33
Additional chromosomes were identified in several cases, ' ' '
and in two cases the additional chromosome was identified as
a member of the C group. Both were cases of acute leukemia,
in which additional C-group chromosomes - usually number
8 - are frequently found; therefore this does not, as one
34
reviewer suggested, constitute evidence of benzene eti-
ology. Persistence of abnormal chromosomes long after
35
exposure and illness was also reported. Tetraploidy or
23 25 27 28
polyploidy was found in several instances. ' ' '
Increased chromosome breakage was reported but not well-
documented .
26

-------
It is important to emphasize that the end stage of
exposure to benzene was as variable in alterations of the
karotype, or chromosome "package," as it was in clinical
manifestations. Indeed, the karyotypic changes may well
have reflected the disease state more than they reflected
the (presumed) inducing agents.
Occupational Exposures - The clearest picture of the
relationship between benzene exposure and chromosome changes
emerges not from experimental studies but from studies of
occupationally exposed workers. Tough and Court-Brown
observed unstable* chromosome damage in cultured lymphocytes
3 6
from workers exposed to benzene solvents. They and their
37
collaborators expanded the study to include groups from
three factories and sex-matched controls. The first group
of 20 men had been exposed to benzene at factory A for
periods of 1 to 20 years and were tested 2 or 3 years after
exposure ended. The second group of 12 had worked periods
of 6 to 25 years in areas where benzene was present (factory
B), the exposure ending approximately 4 years prior to the
study. The third group of 20 had worked for periods of 2 to
26 years in a closed distillation plant (factory C). In
each instance controls were selected from nonexposed indi-
* Unstable aberrations include open breaks, fragments, ring
and multicentric chromosomes, and exchange figures. Stable
aberrations include deletions, translocations, inversions,
trisomies, and other alterations of chromosome number.
27

-------
viduals in the same factories. Available measurf-r.-nts of
atmospheric benzene were 25 to 150 ppm in factories A and B
and approximately 12 ppm in factory C. The results indi-
cated significant increases in unstable aberrations in
exposed workers from factory A, in both exposed workers and
controls in factory B, and in neither group in factory C.
Furthermore, the exposed workers at factory A were older
than their controls, and the authors demonstrated signifi-
cant increases in aberrations in the general population with
increasing age.
Several other investigators have reported increases in
chromosome breakage or in stable and unstable aberrations in
healthy workers.^ ^ In one report"^ the atmospheric
concentrations of benzene were less than 25 ppm. More
41
compelling results were obtained by Forni and co-workers,
who compared data on 34 workers in a rotogravure plant with
those of matched controls. The group of workers was sub-
divided; 10 individuals had been exposed to benzene for
periods of 1 to 22 years (measurements of benzene in the
plant during a brief single period ranged from 125 to 532
ppm). These 10, with the remaining 24 workers, were then
exposed to toluene for periods up to 14 years at levels
ranging up to 82 4 ppm. The age- and sex-matched controls
had no history of exposure to either solvent. The findings
28

-------
in workers exposed only to toluene were not significantly
different from those in the controls, but the group that had
been exposed to benzene showed increases in both stable and
4 2
unstable chromosome aberrations (p < 0.01). Another study
of 25 subjects who had recovered from clinical "benzene
hemopathy" indicated that increases in both types of aber-
rations persisted several years after cessation of exposure,
although there was some decrease in the proportion of un-
stable aberrations. Average values of unstable abnormal-
ities in the exposed group were 3 times greater than in the
controls, and of stable chromosome abnormalities, 30 times
greater.
Most of these industrial studies were systematic to
some degree, including controls and statistical evaluations
of results. These studies of workers from several European
countries all present similar results; that is, statisti-
cally significant increases in both numerical and structural
chromosome alterations in populations exposed to benzene.
PHA-stimulated lymphocytes showed both stable and unstable
chromosome changes in the absence of detectable alterations
of the bone marrow, and aneuploidy or polyploidy was re-
ported frequently. In studies wher^ little or no clinical
symptomatology resulted from exposure, there was consid-
40
erable variation among individuals. For example, m Girard's
29

-------
and Forni's studies a few individuals withir. each benzene-
exposed group were responsible for the significantly higher
chromosome breakage rates in the exposed populations.
Moreover, it is clear that these changes persisted for many
years after exposure, particularly in persons who showed
clear evidence of clinical illness from benzene. The
persistence of damage has been likened to that occurring
after exposure to ionizing radiation. The few reported in-
4 2 35
stances of abnormal clone formation ' are important in
terms of possible leukemogenesis. There is no correlation,
however, between the degree or length of exposure to ben-
zene, the clinical symptoms, and the persistence or extent
of chromosomal aberrations.
Summary
/
The available documentation strongly suggests that
chromosome breakage and rearrangement can result from
exposure to benzene and that damage may persist in hemato-
poietic and lymphoid cells. The aberrations in human cells
appear nonspecific; that is, they are random within the
genome and unrelated to the aberrations associated with
various forms of leukemia. A dose-dependent relationship
between exposure to benzene and amount of chromosome damage
has not been demonstrated. Evidence that benzene causes
disturbance in DNA synthesis suggests that its mutagenic
30

-------
action could involve interference with mitosis. Cytogenetic
analysis of anaphase and postmitotic damage has not been
evaluated adequately.
Theoretical considerations and some clinical observa-
tions suggest a relationship between chronic benzene ex-
posure, chromosome damage, and leukemia. Chromosomally
aberrant clones are typical of some but not all human leuke-
mias, and aberrant cells and clones have been observed in
individuals exposed to benzene who have later developed
leukemia. Many authors have suggested that the lack of an
observed dose-response relation in benzene-induced chro-
mosome damage is due to variation in individual suscepti-
bility. Some studies have recorded biological effects at
(chronic) exposure levels below 25 ppm. The report of a
recent international workshop on the toxicology of benzene
has commented on this literature: "No dose-effect rela-
tionship has so far been demonstrated for benzene-induced
chromosome aberrations. In workers chronically exposed to
levels in the range of 5 to 25 ppm of benzene, both positive
and negative reports involve small numbers of workers and
confirmation of negative data is required on larger groups."
Increased susceptibility to chemical clastogens has been
found in human cancer syndromes that are genetically deter-
43
mined. The variable response to benzene may be attributed
31

-------
also to such possibilities as activation of virus, sup-
pression of immune surveillance, or cocarcinogenic activity
of other chemicals.
More detailed evaluation of the cytogenetic effects of
benzene will require definitive data on dose/response rela-
tionships, relating the frequency and severity of chromosome
damage to the amount and duration of benzene exposure. Both
clastogenic and antimitotic measures of chromosomal muta-
genicity should be evaluated. Benzene dosage should be
correlated with clinical effects as well as with the various
measures of chromosome damage. When an appropriate animal
model becomes available, the evolution and sequence of
chromosome changes with initiation and progression of leuke-
mia may become clear.
32

-------
REFERENCES: Cytogenetic Effects
1.	Ames, B.N., "A Bacterial System for Detecting Mutangens
and Carcinogens", in 'Mutagenic Effects of Environmental
Contaminants' pp. 57 - 66, Eds. H.E. Sutton, and M.I.
Harris; New York, Academic Press, 1972.
2.	Koizumi, A., Dobashi, Y., Tachibana, Y., Tsuda, K.,
Katsunuma, H. Cytokinetic and cytogenetic changes in
cultured human leucocytes and hela cells induced by
benzene. Ind. Health 12:23-29, 1974.
3.	Dobashi, Y. Influence of benzene and its metabolites
on mitosis of cultured human cells. Jap. J. Ind. Health
16:453-461, 1974.
5.	Matsushita, T. Experimental studies on the disturbance
of hematopoietic organs due to benzene intoxication.
Nagoya Journal Med. Sci. 28:204-234, 1966.
6.	Moeschlin, S., Speck, B. Experimental studies on the
mechanism of action of benzene on the bone marrow
(radioautographic studies using ^H-Thymidine). Acta
Haemat. 38:104-111, 1967.
7.	Speck, B., Schnider, Th., Gerber, U., Moeschlin, S.
Experimentelle untersuchungen uber den wirkungsmechan-
ismus des benzols auf das knochenmark. Schweizerische
Medizinische Wochenschrift 3 8:127 4-1276, 1966.
8.	Speck, B., Moeschlin, S. Die wirkung von toluol,
xylol, chloramphenicolund thiouracil auf das knochen.
Experimentelle autoradiographische studien mit ^H-
Thymidin. Schweizerische Medizinische Wochenschrift
98:1684-1686, 1968.
9.	Kissling, M. , Speck, B. Chromosome aberrations in
experimental benzene intoxication. Helv. Med. Acta.
35:59-66, 1972.
10. Boje, H., Benkel, W., Heiniger, H.J. Untersuchungen
zur leukipoese im knochenmark der ratte nach chronischer
benzol-inhalation. Blut 21:250-257, 1970.
33

-------
11.	Morishime, A.; Henrich, R.T., Jou, S., and Nahas, G.G.
"Errors of Chromosome Segregation" in 'Marijuana:
Chemistry, Biochemistry, and Cellular Effects' pp.
265-271, Eds. G.G. Nahas, W.D.M. Payton, and J.
Idanpaan-Heikkila; Springer-Verlag, 1976.
12.	Nowell, P.C., and Hungerford, D.A. "A Minute Chromo-
some in Human Chronic Granulocytic Leukemia", Science
132:1497, 1960.
13.	Rowley, J.D. "A New Consistant Chromosmal Abnormality
in Chronic Myelogenous Leukemia Identified by Quinacrine
Fluorescence and Giemsa Staining" Nature, 234:290-293,
1973 .
14.	Sandberg, A.A., and Hossfeld, D.K. "Chromosomes in the
Pathogenesis of Human Cancer and Leukemia", in 'Cancer
Medicine1 pp. 165-177, Eds. J.M. Holland and E. Frei;
Philadelphia, Lea and Febiger, 1973.
15.	Trujillo, J.M., Cork, A., Hart, J.S., Geroge, S.L., and
Freireich, E.J. "Clinical Implications of Aneuploid
Cytogenic Profiles in Adult Acute Leukemia", Cancer
33:824-831, 1974.
16.	Rowley, J.D. Non-random chromosomal abnormalities in
hematologic disorders of man. PNAS (USA) 72 (1) :152-
156, 1975.
17.	Lyapkalo, A.A. Genetic activity of benzene and toluene.
Gig. Tr. Prof. Zabol. 17(3):24-28, 1973.
18.	Philip, P., Jensen, M.K. Benzene induced chromosome
abnormalities in rat bone marrow cells. Acta Path.
Microbiol. Scand. 78 (A)': 489-490 , 1970.
19.	Rondanelli, E.G., Gorini, P., Magliuio, E., Vannini,
V., Fiori, G.P., Parma, A. Benzene induced anomalies
in mitotic cycle of living erythroblasts. Sangre
9:342-351, 1964.
20.	Rondanelli, E.G., Gorini, P., Pecorari, D., Trotta, N.,
Colombi, R. Effets du benzene sur la mitose erythro-
blastique. Investigations a la Microcinematographie en
contraste de phase. Acta Haemat. 2f:281-302, 1961.
21.	Haberlandt, W., Mente, B. Deviation in number and
structure of chromosomes in industrial workers exposed
to benzene. Zbl. Arbeitsmed. 21 (11) :338-341, 1971.
34

-------
22
23
24
25
26
27
28
29
30
31
32
33
Buday, M., Labant, M., Soos, G. Benzolmergezes okozta
panmyelopathiabol fejlodott acut myelosis. Orv.
Hetil. 112:2415-2416, 1971.
Forni, A., Moreo, L. Cytogenetic studies in a case of
benzene leukaemia. Europ. Journal Cancer 3:251-255,
1967 .
Hartwich, G., Schwanitz, G., Becker, J. Chromosome
anomalies in a case of benzene leukaemia. German Med.
Monthly 14:449-450, 1969.
Marchal, G. A propos de la communication de R. Andre
et B. Dreyfus maladie de Di Guglielmo. Sangre 23:682,
1952 .
Forni, A., Moreo, L., Chromosome studies in a case of
benzene-induced erythroleukaemia. Europ. Journal Cancer
5:459-463, 1969.
Cobo, A., et al. Cytogenetic findings in acquired
aplastic anemia. Acta Haemat. 44:26-32, 1970.
Pollini, G., Colombi, R. Lymphocyte chromosome damage
in benzene blood dyscrasia. Med. Lavoro 55:641-654,
1964.
Aksoy, M., Erdem, S., Erdogan, G.M. Dincol, G. Acute
leukaemia in two generations following chronic exposure
to benzene. Human Heredity 24:70-74, 1974.
Wurster-Hill, D.K., Cornwell, G.G., Mclntyre, O.R.
Chromosomal aberrations and neoplasm - a family study.
Cancer 33(1):72-81, 1974.
Aksoy, M., Erdem, S. Ann. N.Y. Acad. Sci. 165 (30) :15,
1969.
Erdogan, G., Aksoy, M. Cytogenetic studies in thirteen
patients with pancytopenia and leukaemia associated
with long-term exposure to benzene. New Istanbul
Contrib. Clin. Sci. 10:230-247, 1973.
Pollini, G., Strosselli, E., Colombi, R. Relationship
between chromosomal alterations and severity of benzol
blood dyscrasia. Med. Lavoro 55:735-751, 1964.
35

-------
34.	Benzol-Leukamie. G. Hartwich. Ser, Haemat. 7(2):211-
223, 1974.
35.	Pollini, G. , Biscaldi, G.P., Robustelli della Cuna, G.
Le alterazioni cromosomiche dei linfochiti rilevate
dopo cinque anni in soggetti gia'affetti da emopatia
benzolica. Med. Lavoro 60(12) :743-758, 1969.
36.	Tough, I.M. Chromosome aberrations and exposure to
ambient benzene. Lancet. 1:684, 1965.
37.	Tough, I.M., Smith, P.G., Court Brown, W.M., Harnden,
D.G.	Chromosome studies on workers exposed to atmos-
pheric benzene. Europ. Journal Cancer 6:49-55, 1970.
38.	Khan, H., Khan, M.H. Cytogenetic studies following
chronic exposure to benzene. Arch. Toxikol. 31(1) :
39-49, 1973.
39.	Hartwich, G., Schwanitz, G. Chromosomenuntersuchungen
nach chronischer benzol-exposition. Deutsche Medizinsche
Wochenschrift 97:45-49, 1972.
40.	Girard, R. , Mallein, M.L., Bertholon, J., Coeur, P.,
Cl. Evreux, J. Etude de la phosphatase alcaline
leucocytaire et du caryotype des ouvriers exposes au
benzene. Arch. Mai. Prof. 31(1-2):31-38, 1970.
41.	Forni, A., Pacifico, D., Limonta, A. Chromosome studies
in workers exposed to benzene or toluene or both.
Arch. Environ. Health 22:373-378, 1971.
42.	Forni, A.M., Cappellini, A., Pacifico, E., Vigliani,
E.C.	Chromosome changes and their evolution in subjects
with past exposure to benzene. Arch. Environ. Health
23:385-391, 1971.
43.	Auerbach, A. and Wolman, S.R. "Susceptibility of Human
Fibroblasts with Chromosome Instability to further
Damage by Chemical Carcinogens" Nature, 261494-496,
1976.
36

-------
SECTION 3
CHRONIC BENZENE TOXICITY IN ANIMALS
The earliest report of experimental work on chronic
45
benzene toxicity was published by Santesson in 1897. He
had been asked to determine the toxic agent in an industrial
solvent that had caused purpura hemorrhagica in a group of
female employees. He succeeded in producing similar effects
in rabbits after treating them with benzene by poultice and
by subcutaneous injection.
4 8
Selling, who had studied several cases of benzene-
induced purpura hemorrhagica, leukopenia, and anemia in
young girls exposed to benzene in a canning factory, ex-
tended his investigation of the disease to rabbits. He
administered benzene daily by subcutaneous injection and
demonstrated a dramatic decrease in leukocytes, a smaller
decrease in red cells, and degenerative changes in bone
marrow. This study was a landmark because it was the first
in which the investigators measured the depleting effect of
benzene on numbers of circulating blood cells and related
the decreases to bone marrow damage.
37

-------
In subsequent studies of benzene toxicity, the solvent
either has been administered as a vapor, which the animals
inhale in an exposure chamber, or has been injected as the
pure solvent or as a mixture with a carrier such as oil.
The injections are often subcutaneous. The following dis-
cussion focuses first on experiments with inhalation, which
is the most common route of industrial exposure, then com-
pares results of exposures by inhalation with those re-
sulting from injection of benzene.
EXPOSURES BY INHALATION
45	48
Santesson and Selling were not successful in
attempts to produce benzene toxicity by the inhalation
6 2
route. Weiskotten, who exposed rabbits to benzene at a
52
concentration calculated to be approximately 240 ppm,
produced leukopenia, slight anemia, and hemorrhages. Dif-
ferential blood cell counts suggested that small mononuclear
white cells were depressed more than "polynuclear ampho-
philes." The effects were observed within 2 weeks of ex-
posures for 10 hours per day.
56
Svirbely et al exposed rats, dogs, and mice to 1000
ppm of benzene 7 hours per day, 5 days per week for 28
weeks. Although measurements indicated intermittent leuko-
38

-------
penia and lymphocytopenia in the rats, the values returned
to control levels by the time the experiment ended. The
dogs demonstrated lymphocytopenia throughout the study.
Hough et al^ exposed nine dogs to a mean concentration of
800 ppm benzene for 4782 hours over a 123-week period and
observed a decrease in leukocyte counts from 15,917 + 28 63
(M+SD)* to 6272 + 3481, i.e., a reduction to 39 percent of
control values. Differential cell counts were not per-
formed, and no anemia was observed.
Li et al^ investigated the effects of varying the
protein and fat contents of the diet of dogs on their sus-
ceptibility to the depressant action of benzene on bone
marrow. The dogs were fed equicaloric diets described as 1)
low fat, high protein; 2) high fat, high proteir.; 3) low
fat, low protein, and 4) high fat, low protein. Controls
included a group' fed high fat, no protein and a group that
was not exposed to benzene and was fed high fat, low pro-
tein. The exposed animals inhaled benzene at 600 ppm for 42
hours per week over varying periods of time depending on
their responses to benzene. Those least affected were
continually exposed for periods longer than 1 year, whereas
others were sacrificed when they became moribund after
periods as short as 5 to 6 weeks. Monitoring of leukocyte
* Mean value plus or minus the standard deviation.
39

-------
and thrombocyte levels indicated that benzene exposure
reduced the numbers of bot..i types of cells but that protein
deficiency produced greater reductions, which were in turn
exacerbated by high-fat diets.
6 3
In a complex series of studies, Wolf et al used 7-
hour exposure periods. They exposed rats (9400 ppm, 1 to 10
times over 1 to 19 days; 6600 ppm, 70 times over 93 days;
4400 ppm, 28 times over 38 days; 2200 ppm, 133 times over
212 days; 88 ppm, 136 times over 204 days), guinea pigs (two
studies: 88 ppm, 193 times over 269 days; 99 ppm, 23 times
over 32 days) and rabbits (80 ppm, 175 times over 243
days). Although details of cell counts were not reported,
the authors claim to have observed leukopenia in animals
exposed to levels as low as 80 ppm and also to have observed
histopathological changes in bone marrow.
7
Deichmann et al exposed rats to benzene at various
doses, usually for 5 hours per day, 4 days per week. A
decrease in white cells was observed at benzene concentra-
tions of 831, 65, 61, 47, and 44 ppm over periods ranging
from 2 to 8 weeks. No leukopenia was observed at concentra-
tions of 31, 29, or 15 ppm. Females were more sensitive
than males, but differential counts were not reported.
Splenic hemosiderosis was a prominent, but not dose-related,
observation.
40

-------
34
Nau et al exposed rats to benzene at three dose
levels (1000, 200, and 50 ppm) for various periods of time.
At 1000 ppm, rats exposed for 23.5 hours per day seriously
deteriorated after 183 hours, showing distended stomach,
empty gastrointestinal tract, and engorgement of lungs,
liver, kidneys, intestines, and omental tissues. In addi-
tion, the leukocyte count dropped markedly. Lymphocyte
levels appeared to decrease, while polymorphonuclear leuko-
cytes increased. Levels of DNA in the bone marrow were
depressed, and the proportion of red cell precursors was
increased. When exposure was reduced to 19 hours per day
for up to 1782 hours, similar effects were observed. When
the rats were removed from benzene exposure, the blood
analyses showed a return to normal, except that levels of
DNA in the bone marrow remained depressed. In exposures at
200 ppm for 8 hours per day, 5 days per week, leukopenia
occurred after 750 hours, with equal reduction in lympho-
cytes and polymorphs. Myelocytic activity of the bone
marrow was reduced, and erythroid activity was increased.
Similar effects were observed after exposures at 50 ppm.
21
Jenkins et al detected no significant changes in
leukocyte, hemoglobin, or hematocrit values in rats, guinea
pigs, or dogs exposed to benzene concentrations of 255, 30,
41

-------
or 17.5 ppm during repeated exposures for 8 hours per day
for 30 days or during continuous exposures for 90 or 127
days.
19
Ikeda exposed rats to benzene concentrations of 1000
ppm for 60 days, 7 hours per day, 5 days per week. These
exposures indicate that age and sex may affect suscepti-
bility to benzene toxicity in rats, since the leukocyte
levels decreased in the following order: adult males, young
20
males, adult females, young females. Both Ikeda and Ohtsuji
and Drew et al10 reported that pretreatment of rats with
phenobarbital protected against depression of leukocyte
levels during exposure to atmospheric benzene at levels of
4
1000 ppm and 1650 ppm, respectively. Boje et al exposed
rats to 400 ppm of benzene 7 hours per day for 13 weeks and
observed marked leukopenia. Radioautography of samples of
bone marrow from benzene-exposed rats given tritiated
thymidine showed that benzene-treated animals displayed less
incorporation of radioactivity than did control animals.
These authors hesitated to suggest a direct inhibition of
DNA synthesis by benzene because they recognized that cel-
lular damage occurring at any point in the cell cycle might
be manifested as a reduction in thymidine uptake.
5 8
Uyeki et al studied two parameters of benzene tox-
icity not previously reported. The colony forming cell
42

-------
(CFC) assay is a measure of granulocyte precursor activity
of the bone marrow. In these studies, the assay involved
culturing marrow cells from benzene-treated mice in an
appropriate medium. The authors used the number of colonies
of granulocytic cells formed as a measure of the number of
granulocyte precursors in the marrow. The colony forming
unit (CFU) assay, which measures stem cells, also was used
in these studies. The assay involved injecting marrow cells
from benzene-treated animals into X-irradiated mice and
evaluating the subsequent formation of colonies on the
spleen of the receptor mouse. One day after the mice were
exposed to 468 0 ppm of benzene for 8 hours, both CFC and CFU
activity were reduced to 40 to 45 percent of that measured
in controls. Repeated exposure by inhalation further
reduced CFC activity.
The most recent studies on inhalation of benzene in
i
rats and mice, performed over a 2-year period by Laskin et
12
al, are as yet unpublished. Mice and rats were exposed to
benzene at 100 and 300 ppm and were examined periodically
for signs of change in blood cell levels and for indication
of aplastic anemia and/or leukemia. Although evaluation of
these data is not yet complete, some observations may be
made. Although depressions of white cell levels were
observed and animals that died during the study often ap-
43

-------
peared to undergo depletion of bone marrow, other animals
that died displayed no signs of bone marrow toxicity. Of
special interest is the observation that decreased levels of
white cells generally were reflective of lymphocytopenia but
not of granulocytopenia. Further evaluation of the data is
ongoing, and additional useful information is expected.
Preliminary assessment indicates that in C-57 Black mice and
Sprague-Dawley rats that were exposed to 300 ppm of benzene
for approximately one year, lymphocyte counts were reduced
to 25 and 60 percent of those in controls, respectively,
during the first 5 to 10 weeks of exposure. The counts
remained close to those levels for the remainder of the
year. The mice, but not the rats, also displayed a reduc-
tion in erythrocyte counts, which became apparent at the
same time as the lymphocytopenia and remained fairly con-
stant at about 68 percent of control levels throughout the
remainder of the period.
EXPOSURE BY OTHER ROUTES
4 5	4 8
Both Santesson and Selling, administering benzene
subcutaneously as described above, produced benzene toxic-
ity. Selling administered the benzene dissolved in olive
oil to rabbits at a dose of 1 ml/kg per day. He monitored
leukocytes until they were reduced to a level of 200 to 800
44

-------
cells per mm . Because of differential sensitivity among
the animals, periods of 4 to 9 days were required to reach
these low levels. When additional injections were given,
most of the animals died. Since these experiments were of
short duration, red cell levels did not vary significantly.
In a similar series of studies Weiskotten et al^ treated
the animals until the white cell count dropped to 1000 per
3
mm . Studies of recovery after cessation of treatment by
SellingWeiskotten et al,^ Brandino,^ and Eecci^ showed
that the white cell counts initially rise, then go through a
secondary depression, which Weiskotten described as the
deuterophase, and then return gradually to control values.
4 8
It is emphasized that Selling produced complete aplasia of
the bone marrow when giving benzene parenterally, an effect
that corresponds well with the observation of aplastic
anemia in humans exposed via inhalation to benzene in in-
dustrial environments.
Gerarde^ performed a similar set of experiments in
which rats were treated daily for 2 weeks with 1 ml/kg of
benzene as a 50 percent solution in olive oil. He continued
to observe the animals for an additional 3-week period,
during which treatments were discontinued. Leukopenia was
observed during the initial 2-week treatment period and was
45

-------
accompanied by decreases in the levels of nucleic acid and
of nucleated cells in the marrow of the femur. When admin-
istration of benzene was discontinued, each of these three
2 8
parameters returned to normal. Latta and Davies treated
rats with a solution of 50 percent benzene in olive oil in
doses of 2, 3, and 4 ml/kg daily for up to 60 days. Fol-
lowing a temporary stimulation of white cell production, the
circulating leukocyte levels dropped and the data suggested
impairment of leukocyte maturation. Lymphocyte production
appeared to be more dramatically impaired than was granu-
locyte production. At the higher doses no myelocytes were
observed after 24 days, and the marrow was completely aplas-
tic after 60 days.
Steinberg"^ observed degeneration of bone marrow in
benzene-treated rabbits. This was followed by regeneration
after the benzene treatments were stopped. Although extrap-
olation of doses from his data is difficult because the
weights of the animals are not given, it is calculated that
the dose was approximately 2 ml of benzene per day per
rabbit. The animals were dosed subcutaneously for 6 to 7 0
days. Under these conditions leukocyte counts were reduced
but not in a dose-related fashion. Throughout the treatment
period white cell counts in the dosed animals averaged 22.6
+ 7.3 percent of control counts, the calculation being based
46

-------
on all values from day 6 through day 70. As part of these
studies, Steinberg determined the ability of the marrow to
regenerate by comparing regeneration in rabbits from which
marrow had been surgically extracted with regeneration in
rabbits that had been intoxicated with benzene. Unlike
control animals, benzene-treated animals displayed regenera-
tion only to the extent of formation of primitive reticular
cells.
36 3 8
Nomiyama, ' who performed daily subcutaneous injec-
tions of rats with benzene (1 gm/kg), reported reduction of
leukocyte counts. In similar studies with several strains
of mouse, he found a strain-dependent variation in sensi-
tivity to benzene.
2 5 26 33
Speck et al ' ' administered benzene to rabbits at
doses of 0.2 and 0.3 ml/kg per day and produced severe
leukopenia within 1 to 9 weeks. Although reductions in
hemoglobin and red cells were more pronounced, reductions in
reticulocytes and thrombocytes also were observed. Pancyto-
penia was a common finding. Among 19 marrow smears, 21
percent were very hypoplastic, 32 percent were hypoplastic,
26 percent appeared normal, and 21 percent were hypercel-
lular. No correlation between cellularity and duration of
exposure to benzene was apparent. In general, myeloid
precursors were diminished more than erythroid cells. After
47

-------
treatment of the animals with tritiated thymidine, radio-
autography of the marrow indicated that incorporation of
radioactivity into DNA was reduced. The authors interpreted
these findings to indicate that benzene toxicity leads to a
decrease in DNA synthesis. These authors also demonstrated
chromosomal aberrations in rabbits similarly treated with
benzene. Finally, they reported a decrease in incorporation
of tritiated cytidine into RNA, which they interpreted as an
inhibition of RNA synthesis.
29
Lee et al demonstrated that benzene depressed the
59
incorporation of a radioactive iron isotope ( Fe) into
59
circulating erythrocytes. Radioactivity from Fe dis-
appears from the blood soon after parenteral administration
and reappears when incorporated into hemoglobin. The rate
of reappearance reflects the maturation and proliferation of
the components of bone marrow cells. Single doses of
benzene in corn oil at 400 mg/kg and 2200 mg/kg were given
to mice. In measurements of 2 4-hour uptake, the reappear-
ance of radioactivity reflects only reticulocyte activity,
whereas measurements of 72-hour uptake reflect the entire
cycle from stem cell to mature red cells. Thus, the 24-hour
uptakes that were measured 24 or 48 hours after giving
benzene were lower than those of controls, whereas they were
not lower when iron was given 1, 12, or 72 hours after
48

-------
administration of benzene. Based on current knowledge of
red cell maturation in the mouse, these data suggest that
the cells most sensitive to benzene are early precursors
called pronormoblasts and normoblasts. In similar experi-
ments in which 72-hour uptakes were measured, the reductions
in iron utilization were less dramatic, perhaps because the
longer uptake period allowed sufficient time for compensa-
tory mechanisms to come into play.
53
In more recent studies not yet published, Lee et al
administered benzene to mice in multiple doses over several
days. They found that when mice were given a single sub-
cutaneous dose (440 mg/kg) of benzene per day, incorporation
59
of Fe was reduced 50 to 6 0 percent in 20 days. When mice
were given 880 mg/kg in single daily doses, iron utilization
decreased gradually to 80 percent of controls in 10 days.
When doses of 440 mg/kg were given twice per day (i.e. a
total daily dose of 880 mg/kg given as two doses) the reduc-
tion in iron uptake occurred more rapidly, and by the tenth
day almost no iron was taken up. Finally, when two doses of
benzene, at 880 mg/kg per dose, were given each day, incor-
poration of iron appeared to terminate by the sixth day.
These data suggest that a dose-times-time relationship
governs the cumulative effects of benzene. Giving the dose
49

-------
twice per day (440 mg/kg each) exacerbated the effect over
that observed when the same amount was given as a single
dose, perhaps because the single dose schedule allows for
essentially complete excretion of the benzene either as free
benzene in exhaled air or as water-soluble metabolites in
the urine. When the dose is divided, the animals apparently
retain some benzene the morning after, to which the next
dose adds. Therefore, the animals accumulate more and more
benzene as the experiment progresses.
The oral route is the least investigated in studies of
6 3
benzene toxicity. Wolf et al administered 132 feedings of
benzene to rats at doses of 1, 10, 50, and 100 mg/kg over a
187-day period. No effect was observed after the 1 mg/kg
dose, slight leukopenia was observed after the 10 mg/kg
dose, and both leukopenia and anemia were observed after the
higher doses.
EVALUATION AND COMMENTS
The major problem in extrapolation of data obtained in
animal studies to benzene toxicity in humans relates to
deficiencies in use of controls and in blood counting
techniques in the animal studies. Normal values of total
leukocytes in common laboratory animals such as the mouse,
rat, and rabbit range widely from approximately 4000 to 5000
50

-------
3
cells per ram to over 20,000, with occasional values m the
46
rabbit as low as 3200. Thus, it may be argued that reduc-
tions in white cell count within that range may not reflect
true toxicity but may reflect successful protective re-
sponses by the organism. In contrast, decreases in leuko-
3
cyte levels to below 2000 cells per mm may reflect insults
severe enough to impair the health of the animal. In many
reports that claim the production of leukopenia, however, it
is implied that leukopenia is defined as a reduction of
white cell levels below some control value. When the con-
trol values are set by measurements of the test animals
prior to exposure, an error may be introduced because of
natural variations in white cell counts. For example,
Cheng*' reported that leukocyte counts in rabbits increase
gradually from 2000 in the first week of life to between
4500 and 6000 cells at 100 to 200 days of life. Consider-
able variability was observed among 240 adult rabbits in
which white blood cell (WBC) counts averaged 7000 + 5000
cells. Although age may not be a factor in studies with
mice, the time of day of sampling, the site from which the
sample is taken, and the strain of mouse all affect the cell
46
count. The effect of age on the leukocyte count in rats
43
remains an open question. Reich and Dunning studied eight
14
different strains of rat, and Harris and Burke studied the
51

-------
Wistar rat, a strain not examined by Reich and Dunning.
Both groups reported that age did not affect leukocyte
counts. It was suggested that the neutrophil-lymphocyte
12
ratio increases with age in rats. Recently, Laskin et al
found that a relative decrease in leukocytes occurs within
the first 3 months of life of Sprague-Dawley rats. These
observations indicate that claims for benzene-induced
leukopenia should be supported by proper controls for age,
as well as for exposure.
Although the extreme decreases in leukocytes observed
4 8	61
by Selling, Weiskotten and co-workers, and Kissling and
2 6
Speck clearly demonstrated leukopenia, the studies of
Hough et al,"*"^ Wolf et al,^ Deichman et al,^ Nau et al,^
19 20	9
Ikeda et al, ' Drew et al, and others must be evaluated
with respect to the age effect and other factors.
A second question regarding changes in white cell
counts concerns the interpretation that reductions in
numbers of white cell must, of necessity, reflect bone
marrow damage. Relatively few investigators have performed
differential counts of the white blood cells. Laskin et
12
al have shown that during extended exposure to benzene,
the total WBC counts in rats and mice decreased but the
numbers of granulocytes apparently did not. Since granulo-
cytopenia reflects bone marrow damage, these animal models
52

-------
may not reflect the huma.n disease in which granulocytopenia
is a prominent feature.
Even with proper controls and differential cell counts,
however, proper interpretation of the effects of benzene in
animal experiments will be difficult if evaluation of peri-
pheral blood factors remains the sole focus of attention.
The problem can be exemplified by comparing the results of
Speck and associates^'^ with those of Boje et al.^ The
Speck group treated rabbits with benzene subcutaneously
3
until white cell levels decreased below 1000 per mm .
Under these conditions mitosis in bone marrow was impaired
and uptake of tritiated thymidine into DNA was reduced
4
significantly. When Boje et al exposed rats to benzene by
inhalation at 400 ppm for up to 13 weeks, they demonstrated
a similar decrease in DNA synthesis in the bone marrow, but
in this case the white cell levels were decreased to only 46
percent of controls, which represents a considerably higher
3
count than Speck's 100 0 per mm . Thus, significant damage
to marrow may occur when leukocyte levels remain relatively
high. It must be argued then that reductions in white cell
levels, even when the counts remain in the normal range, may
be indicative of marrow damage and _Lt is not necessary to
demonstrate leukocyte levels below 2000 cells to suspect
benzene-induced marrow damage. Careful studies of the
53

-------
effects of benzene on circulating blood cell elements should
be supported with information concerning the effects of
benzene on bone marrow.
Benzene and Leukemia in Animals
Leukemia is a general term for a group of diseases
usually characterized by large increases in numbers of white
blood cells in blood and/or bone marrow or the appearance of
unusual leukocyte precursors in the blood. Leukemia is
associated with neoplasms, i.e., new growths of tissue
serving.no physiologic function. It is usually accompanied
by a variety of defects in the hematopoietic, or blood-cell
forming, system. Leukemias are known to occur spontaneously
in some strains of mice, and there is abundant evidence that
, . ,	.	, , . . , , .	, , 8,15,17,31,41,49
chemicals can induce leukemia in both mice and rats.
31
Lignac reportedly produced leukemia in mice by treating
them with benzene subcutaneously for 17 to 21 weeks. Of
the 44 mice that survived treatment, 8 were described as
having developed leukemia or lymphosarcoma. Failure to
include control mice and to provide details concerning the
strain of mouse and the diagnostic criteria leaves this
report open to question.
In a specific attempt to duplicate the results of
2
Lignac, Amiel treated mice of the AKR, DBA2, C3H' ancl C57B1
strains with weekly injections of benzene (30 mg/kg) through-
54

-------
out their lifetimes but observed neither aplastic anemia nor
59
leukemia. Ward et al studied a group of C5731 mice for a
total of 104 weeks, during which time the dosage schedule
was varied but in general was increased from 45 0 mg/kg to
1.8 gm/kg. Although a number of mice died of the toxic
effects of benzene, there was not a statistically signif-
icant increase in incidence of neoplastic disease in ben-
zene-treated mice over that observed in controls.
The evidence for production of leukemia in animals by
injection with benzene must be considered nonconclusive.
6 3
Moreover, neither oral dosing, skin painting with ben-
24,27	. . . J.. 21,63	^ ^ ^
zene,	nor inhalation	has been demonstrated to
produce leukemia or any other type of neoplastic disease in
rats, mice, guinea pigs, or rabbits.
In contrast, with respect to possible benzene-asso-
ciated leukemia in humans, evidence from industries that
have used benzene heavily implies a direct relationship
between benzene exposure and development of leukemias.
Furthermore, the leukemias have been observed mainly in
cohorts of workers among whom many showed signs of benzene-
induced bone marrow damage of variable severity. Despite
the inability of investigators to demonstrate leukemia in
groups of animals exhibiting bone marrow damage, the studies
55

-------
of humans provide compelling evidence that benzene is in-
volved in leukemogenesis.
Several theories might be offered to explain the in-
ability yof researchers to induce leukemia in animals by
treatment with benzene. Man may be the only species yet
observed that is susceptible to benzene-induced leukemia for
a variety of reasons such as 1) a novel metabolic pathway
that produces a unique reactive metabolite not formed in
other animals, 2) relative inefficiency in repair of DNA. of
benzene-induced damage, or 3) relative ineffectiveness of
immune surveillance following benzene-induced insult. Also,
the phenomenon may involve other mechanisms of which we are
not aware. It is possible that the latency period in
animals is long enough to preclude the appearance of leuke-
mia during the lifetime of the animal, or an as-yet-unknown
cocarcinogen may be required to evoke the leukemogenic
response initiated by benzene. If bone marrow damage is a
prerequisite for benzene-induced leukemia, the appropriate
animal experiments may not have been done. For example, it
may be necessary to induce bone marrow damage and then to
allow for recovery during the remainder of the litre of the
animal. Regular monitoring might then show some animals
with leukemia. The researcher then should demonstrate that
a similar control population displays significantly fewer
56

-------
cases of leukemia. In any attempts to disclose a cause and
effect relationship between benzene exposure and leukemia,
it is essential that the studies include sufficient numbers
of control and treated animals. The need for large numbers
of animal subjects is underscored by the exceedingly low
incidence of benzene-induced leukemia in human workers.
Meaningful statistics on these cases have been collected
only when large groups of workers were observed.
57

-------
REFERENCES: Chronic Benzene Toxicity in Animals
1.	Aksoy, M., Dincol, L., Erdem, S., Akgun, T., Dincol, G.
Details of blood changes in 32 patients with pancytopenia
associated with long-term exposure to benzene. Brit.
J. Industr. Med., 29:56-64, 1972.
2.	Amiel, J.-L. Essai negatif d1induction de leucemies
chez les souris par le benzene. Rev. Franc. Etudes
Clin. Biol., 5:198-199, 1960.
3.	Andrews, L.S., Lee, E.W., Witmer, C.M., Kocsis, and
Synder, R. Biochem. J., 26:293, 1977.
4.	Boje, H., Benkel, W. , Heiniger, H.J. Untersuchungen
zur leukipoese im knochenmark der ratte nach chronischer
benzol-inhalation. Blut, 21:250-257, 1970.
5.	Brandino, G. Osservazione istologische nel'intossicazione
acute e croniche da benzolo. Gazz. Med. Lomborda,
81:141, 1922.
6.	Cheng, S.C. Amer. J. Hyg., 11:449, 1930.
7.	Deichmann, W.B., MacDonald, W.E., Bernal, E. The
hemipoietic tissue toxicity of benzene vapors. Tox.
Appl. Pharm., 5:201-224, 1963.
8.	Diwan, B.A., and Meier, H. Can. Lett., 1:249, 1976.
9. Drew, R.T., Fouts, J.R., Harper, C. The influence of
certain drugs on the metabolism and toxicity of benzene.
IN: Symposium on Toxicology of Benzene and Alky1-
benzenes, Braun, D., ed., Industrial Health Foundation,
Pittsburgh, pp. 17-31, 1974.
10. Drew, R.T., Fouts, J.R. The lack of effects of pre-
treatment with phenobarbital and chlorpromazine on the
acute toxicity of benzene in rats. Tox. Appl. Pharm.,
27:183-193, 1974.
58

-------
11.	Gerarde, H.W., Ahlstrom, D.B. Toxicologic studies on
hydrocarbons. XI. Influence of dose on the metabolism
of mono-n-alkyl derivatives of benzene. Tex. Appl.
Pharm., 9:185-190, 1966.
12.	Laskin, S., Goldstein, B.D., Snyder, C.A. Unpublished
observations, 1977.
13.	Gonasun, L.M., Witmer, C., Kocsis, J.J., Sr.yder, R.
Benzene metabolism in mouse liver microsomes. Tox.
Appl. Pharm., 26:398-406, 1973.
14.	Harris, C., Burke, W.T. Am. J. Path., 33:931, 1957.
15.	Hartmen, H.A., h'ller, E.C., Miller, J.A., Morris, F.K.
Can. Res., 19:210, 1959.
16.	Hough, V.N., Guinn, F.D., Freeman, S. Studies on the
toxicity of commercial benzene and of a mixture of
benzene, toluene and xylene. J. Industr. Kyg.,
26:296-306, 1944.
17.	Huggins, C.B., Sugiyama, T. Proc. Natl. Acad. Sci.,
55:74, 1966.
18.	Hunter, C.G. Solvents with reference to studies on the
pharmacodynamics of benzene. Proc. Roy'. Soc. Med.,
61:913-915, 1968.
19.	Ikeda, M. Enzymatic studies on benzene intoxication.
J. Biochem., 55:231-243, 1964.
20.	Ikeda, M., Ohtsuji, H., Imamura, T. In vivo suppres-
sion of benzene and styrene oxidation by co-administered
toluene in rats and effects of Dhenobarbital. Xenobiotica,
2:101-106, 1972.	"
21.	Jenkins, L.J., Jr., Jones, R.A., Siegel, J. Long-term
inhalation screening studies of benzene, toluene, o-
xylene, and cumene on experimental animals. Tox. Appl.
Pharm., 16:818-823, 197C.
22.	Jollow, D.J., Kocsis, J.J., Snyder, R. , Vainio, H.
(Ieds.), Biological Reactive Incermediates, Plenum
Press, New York, 1977.
59

-------
23.	Jollow, D.J., Mitchell.,. J. R. , Potter, W.Z., Davis,
D.G., Gillette, J.R., Birodie, B.B. J. Pharm. Exp.
Thsr., 187:195, 1973.
24.	Kirschbaum, A., Strong, L.C. Influence of carcinogens
on the age incidence of leukemia in the high leukemia F
strain of mice. Cancer Res., 2:841-845, 1942.
25.	Kissling, M., Speck, B. Chromosone aberrations in
experimental benzene intoxication. Helv. Med. Acta.,
36:59-66, 1972.
26.	Kissling, M., Speck, B. Further studies on experimental
benzene induced aplastic anemia. Blut Z Gesamte Blut
Forsch, 25 (2) : 97-103, 197.?.
27.	Laerum, O.D. Reticulum cell neoplasms in normal and
benzene treated hairless mice. Acta Path Microbiol.
Scand. 81:57-63, 1973.
28.	Latta, J.S., Davies, L.T. Effects on the blood and
hematopoietic organs of the albino rat of repeated
administration of benzene. Arch. Path., 31:55-67,
1941.
29.	Lee, E.W., Kocsis, J.J., Snyder, R. Acute effects of
benzene on 5 Fe incorporation into circulating erythrocytes.
Tox. Appl. Pharm., 27:431-436, 1974.
30.	Li, T.W., Freeman, S., Gunn, F.D. J. Physiol.,
145:158-166, 1945.
31.	Lignac, G.O.E. Die benzolleukamie bei menschen und
weissen mausen. Klin. Wschr., 12:109-110, 1932.
32.	Mitchell, J.R., Nelson, S.D., Snodgrass, W.R., Timbress,
J.A. IN: Biological Reactive Intermediates, D.J.
Jollow, J.J. Kocsis, R. Snyder, H. Vainio (editors),
p. 271, Plenum Press, New York, 1977.
33.	Moeschlin, S., Speck, B. Experimental studies on the
mechanism of action of benzene on the bone marrow
(radioacutographic studies using ^H-thymidine). Acta
Haemat., 38:104-111, 1967.
34.	Nau, C.A., Neal, J., Thornton, M. Arch. Env. Health,
12:382, 1966.
60

-------
35.	Neal, R.A., Kamatakii, T., Lin, M. , Ptashne, K.A.,
Dalvi, R.R., Pooge, R.E. IN: Biological Reactive
Intermediates, D.J. Jollow, J.J. Kocsis, R. Snyder, H.
Vainio (editors), p.320, Plenum Press, New York, 1977.
36.	Nomiyama, K. Studies on the poisoning by benzene and
its hornologues (6) oxidation rate of benzene and
benzene poisoning. Med. J. Shinshu Univ., 7:41-48,
1962.
37.	Nomiyama, K. Experimental studies on benzene poisoning.
Bull. Tokyo Med. Dental Univ., 11:297-313, 1964.
38.	Nomiyama, K., Minai, M. Ind. Health, 7:54, 1969.
39.	Nomiyama, K., Nomiyama, H. Int. Arch. Arbeitsmed.,
32:85, 1974.
40.	Nomiyama, K., Nomiyama, H. Respiratory retention,
uptake and excretion of organic solvents in man.
Benzene, toluene, n-hexane, trichloroethylene, acetone,
ethyl acetate, ethyl alcohol. Int. Arch. Arbeitsmed.
32:75-83, 1974.
41.	Ogui, T., Nakadate, M. , Odashima, S. Can. Res.,
36:3043, 1976.
42.	Parke, D.V., Williams, R.T. Detoxication. XLIX.
Metabolism of benzene containing (C^) benzene.
Biochem. J., 54:231-238, 1954.
43.	Reich, C., Dunning, W.F. Cancer Res., 3:248, 1943.
44.	Saito, R.U., Kocsis, J.J., Snyder, R. Effect of
benzene on hepatic drug metabolism and ultrastructure.
Tox Appl. Pharm, 26:209-217, 1973.
45.	Santesson, C.G. Ueber chronische vergiftungen mit
steinkohlentheerbenzin; vier todesfalle. Arch. Hyg.
Berlin, 31:336-376, 1897.
46.	Schalm, O.W., et al. Veterinary Hematology, 3rd
Edition, Lea & Febiger, Philadelphia, 1975.
47.	Secchi, P. Ricerche ematologiche nelle intossicazione
acute e chroniche da benzolo. Rif. Med., 30:995, 1914.
61

-------
48.	Selling, L. Benzol as a leucotoxin. Studies on the
degeneration and regeneration of the blood and hemato-
poietic organs. John Hopkins Hospital Reports,
17:83-142, 1916.
49.	Shay, H. Gruenstein, M.G., Marx, H.E., Glazer, L.
Can. Res. 11:29, 1951.
50.	Sherwood, R.J. Benzene: The interpretation of moni-
toring results. Annals of Occupational Hygiene,
15:409-421, 1972.
51.	Snyder, R. , Uzuki, F. , Gonasun, L. , Bromhelr1, E. ,
Wells, A. The metabolism of benzene in vitro. Tox.
Appl. Pharm., 11:346-360, 1967.
52.	Snyder, R., Kocsis, J.J. Current concepts of chronic
benzene toxicity. CRC Critical Reviews in Toxicology,
pp. 265-288, June 1975.
53.	Snyder, R., Lee, E.W., Kocsis, J.J., Unpublished, 1977.
54.	Snyder, R., Lee, E.W., Kocsis, J.J., Witmer, C.M. Life
Sciences (in press), 1977.
55.	Steinberg, B. Bone marrow regeneration in experimental
benzene intoxication. Blood, 4:550-556, 1949.
56.	Svirbely, J.L., Dunn, R.C., von Oettingen, W.F. The
chronic toxicity of moderate concentrations of benzene
and of mixtures of benzene and its homologues for rats
and dogs. J. Industr. Hyg. Toxicol., 26:37-46, 1944.
57.	Teisinger, J., Jiserova-Bergerova, V. , Kudrna, J. The
metabolism of benzene in man. Praccv. Lek., 4:175-188,
1952.
58.	Uyeki, E.M., Ashker, A.E., Shoeman, D.W., Bisle, T.U.
Toxicol. Appl. Pharm., 40L49, 1977.
59.	Ward, J.M., Weisburger, J.H., Yamamoto, R.S., Benjamin,
T., Brown, C.A., Weisburger, E.K. Long-term effect of
benzene in C57BL/6N mice. Arch. Environ. Health,
30:22-25, 1975.
60.	Weiskotten, H.G., Schwartz, S.C., Steensland, J. Med.
Res., 33:127, 1915.
62

-------
61. Weiskotten, H.G., Schwartz, S.C., Steenslar.d, H.S. The
action of benzol. The deuterophase of the diphasic
leukopenia and intogen-antibody reaction. J. Med.
Res., 35:63-79, 1916.
62. Weiskotten, H.G., Schwartz, S.C., Steenslar.d, H.S. J.
Med. Res., 41:425, 1920.
63.	Wolf, M.A. , Rowe, V.K., McCollister, D.D., Hollingsworth,
R.L., Oyen, F. Toxicological studies of certain
alkylated benzenes and benzene. AMA Arch. Ind. Health,
14:387-398, 1956.
64.	Zampaglione, N., Jollow, D.J., Mitchell, J.R., Stripp,
B., Hamrick, M., Gillette, J.R., J. Pharm. Exp. Ther.
187:218, 1973.
63

-------
SECTION 4
BENZENE TOXICITY IN MAN
INTRODUCTION
The world medical literature contains hundreds of
references describing thousands of cases of human hemato-
logical toxicity associated with benzene exposure. This
literature has been reviewed relatively recently by a number
,	24,39,77,87,143,146,170	.	,
of authors.	Although numerous hema-
tological disorders have been reported in association with
benzene exposure, only two entities are clearly related to
benzene. These are 1) pancytopenia and its variants, in-
cluding anemia, leukopenia, thrombocytopenia and aplastic
anemia; and 2) acute myelogenous leukemia and its variants,
such as acute myelomonocytic leukemia and erythroleukemia.
Unfortunately, there is a relative paucity of information
concerning the doses of benzene to which affected indi-
viduals were exposed.
This section deals with the evidence presented in the
literature concerning the human hematological toxicity of
benzene. The emphasis is primarily on studies that have
evaluated relatively large numbers of occupationally exposed
64

-------
individuals, particularly where measurement of dose has been
attempted. Most of the test of the reports describe one or
a few cases of hematotoxiciity associated with benzene ex-
posure. Such reports do have cumulative weight, partic-
ularly in relation to the causative role of benzene in acute
myelogenous leukemia. Individually, however, they provide
little information not given elsewhere and therefore they
are not discussed here.
In the reported cases of benzene hematotoxicity, ex-
posure has usually occurred in a workplace where benzene was
used as a solvent or manufactured as a product. In some
instances, benzene was used because it is inexpensive and
has excellent solvent properties. In other instances,
benzene was an inadvertent contaminant of aromatic hydro-
carbon solvents. Whatever the reason for its use, in almost
all cases in which benzene hematotoxicity has been reported,
the patient has been exposed also to some other solvent or
chemical product. That it is benzene, rather than some
other commonly associated agent such as xylene or toluene,
that is primarily responsible for hematotoxicity appears
well established. The major evidence supporting this asser-
tion includes the observation that benzene has been the
common denominator in many different occupational settings
in various parts of the world in which exposure to other
65

-------
chemicals has varied widely. In addition, the pancytopenic
effect of benzene in man can be readily duplicated in ex-
periments with a variety of animal species. This is not
true of the other commonly associated contaminants.
It is possible, however, that compounds inhaled along
with benzene alter the expression of benzene hematotoxicity.
In particular, it is conceivable that other aromatic hydro-
carbons may modify benzene metabolism in humans. This is an
area in which additional information on the effects of
benzene in man would be of great value.
PANCYTOPENIA
The term pancytopenia refers to a diminution of all
formed elements in the blood. In benzene exposure this
diminution is due primarily to an interference in the pro-
duction of red cells, white cells, and platelets in the bone
marrow, although some evidence suggests that survival of
blood cells within the circulation may be shortened also.
Many of the cases associated with benzene exposure are
described as "aplastic anemia." This term classically
denotes a condition in which the number of hematopoietic
precursor cells within the bone marrow diminishes markedly.
Aplastic anemia is usually associated with a high degree of
pancytopenia and is observed particularly in severe or fatal
66

-------
cases. However, a decrease in bone marrow cellularity is
not always noted in individuals with hematotoxicity asso-
ciated with benzene exposure nor in all animals treated with
benzene. This may be attributable to individual variation,
to ineffective response of poietic bone marrow to benzene,
or to the sampling error inherent in the presumption that a
local aspirate of bone marrow represents all hematopoietic
tissue. For purposes of discussion, cases of clear bone
marrow aplasia are considered under the heading of pancyto-
penia .
Numerous cases of individual cytopenias, e.g. anemia
without leukopenia or thrombocytopenia, also are reported in
benzene-exposed individuals. These have been observed
mostly in industries where severe benzene hematotoxicity in
some workers has led to evaluation of the entire work force.
It should be noted that there is wide variation in the
normal red cell, white cell, and platelet counts as well as
appreciable reserve production capability. A person in whom
one of the formed elements decreases to a level below the
accepted normal value may also experience an effect on the
production of other cell types that is not clinically ap-
parent. Accordingly, the individual cytopenias are also
discussed under the heading of pancytopenia.
67

-------
Exposure to benzene clearly produces hematological
toxicity in animals and man. Evidence of a pancytopenic
effect of benzene was first noted in 1897 by Santesson, who
reported four cases of fatal aplastic anemia occurring in
workers fabricating bicycle tires. Since that time numerous
case reports and surveys of occupationally exposed groups of
workers have documented this association, and many reviews
of these cases have appeared.24'39,77,84,87,91,143,!46,163,!70
The causal relationship of benzene to pancytopenia in man is
most clearly supported by studies, described below, of
groups of workers in whom the appearance of pancytopenia was
temporally related to the inception of benzene use and in
whom the outbreak of hematological effects was ended by
replacement of benzene with some other solvents.
Current concepts of the hematopoietic system stress the
role of a pluripotential myeloproliferative stem cell in
production of erythrocytes, granulocytes, platelets, and
perhaps fibroblasts and other monocytic cells. This stem
cell is believed to be able to differentiate into the
precursors of the various formed elements in response to
microenvironmental conditions. The production of pancyto-
penia suggests that benzene is toxic to this stem cell or to
early hematopoietic precursors of red cells, granulocytes,
and platelets. Some incomplete evidence also points to an
68

-------
even earlier pluripotential cell with capabilities of
differentiating into lymphocytic as well as myelocytic
precursors. This may be pertinent to benzene toxicity in
that lymphocytopenia is frequently observed in pancytopenic
individuals with a history of benzene exposure. In addi-
tion, chromosomal abnormalities in circulating lymphocytes
have been reported in association with benzene exposure (see
Section 2), and evidence suggests that such exposure in-
creases the risk of lymphocytic neoplasms.
The symptoms of pancytopenia in individuals exposed to
benzene are described by a number of authors who have
investigated occupationally exposed groups. In milder
cases, these tend to be such nonspecific complaints as
lassitude, tiredness, easy fatiguability, malaise, dizzi-
ness, headaches, palpitation, and shortness of breath. Such
symptoms tend to appear gradually, and presumably reflect
anemia. Occasionally, in more severe cases, hemorrhagic
manifestations due to thrombocythemia or decreased platelet
function are also observed. In severe cases of pancyto-
penia, death is often due to hemorrhage or to overwhelming
infection, the latter reflecting the decrease in granulo-
cytes. In addition, patients with pancytopenia may sub-
sequently develop fatal acute leukemia, sometimes with an
intervening period of apparent recovery.
69

-------
Among the earliest systematic evaluations of the
pancytopenia effects of benzene are those of Greenburg et
81	78	79
al, Goldwater, and Goldwater and Tewksbury. These
investigators studied workers in rotogravure printing plants
where benzene had been in use for a period of 3 to 5 years.
The benzene contents of the ink solvents and thinners ranged
from 10 to 80 percent. Forty-eight analyses of benzene
concentrations in the air of pressrooms of three plants
revealed levels ranging from 11 to 1060 ppm, with a median
concentration of 132 ppm. The most frequent clinical
complaints were fatigue, dryness of the mucus membranes,
lethargy, dizziness, headache, and shortness of breath.
Hematological studies were performed in 332 exposed male
workers and 81 controls. Various degress of hematological
toxicity were observed in 65 of these workers, 23 of whom
were considered to be severely affected. As a result of
these studies, six individuals were referred for hospital-
ization. The remaining workers continued on the job and,
following replacement of benzene with other solvents,
79
hematological recovery was demonstrated. In comparison
with the control group, the most frequently observed find-
ings were anemia, macrocytosis, and thrombocytopenia. An
absolute lymphocytopenia was more common than was neutro-
penia, which was rarely observed. Other relatively infre-
70

-------
quent findings were prolongation of the bleeding or coagu-
lation times, an increase in capillary fragility, and an
increase in serum bilirubin and reticulocytes. Osmotic
fragility was normal, as was the erythrocyte sedimentation
rate. No monocytosis, eosinophilia, or basophilia was ob-
served .
Several occupationally exposed groups were studied
193
during World War II. Wilson evaluated workers in an
American rubber factory in which peak benzene levels were
said to be 500 ppm, with an average of about 100 ppm. Of
1104 workers studied, 83 were observed to have mild hemato-
logical effects and 25 had more severe pancytopenia. Nine
of the latter group were hospitalized, and three died.
8 5
Hamilton-Patterson and Browning reported observations of
200 women in 13 aircraft factories in England as compared to
200 controls. These workers had been involved in rubber
manufacture and had been exposed to solvents and adhesives
containing 5 to 20 percent aromatic hydrocarbons. In con-
81	78
trast to the findings of Greenberg et al and Goldwater
these authors suggest that neutropenia is the earliest and
most consistent indicator of benzene toxicity. No signifi-
cant difference in red cell or lymphocyte counts was ob-
8 8
served. Helmer in Sweden evaluated 184 workers (169
women, 15 men) in a rubber raincoat factory where benzene
71

-------
levels of 137 to 218 ppm were measured and it was believed
that levels had been higher iri the past. Evidence of hema-
tological toxicity was observed in 60 individuals (58 women,
2 men). Reevaluation 16 months after cessation of benzene
use revealed that 46 recovered, 12 still had significant
effects, and 2 had died. The most frequent symptoms were
gradual development of headache and tiredness. The author
also notes the frequency of cutaneous hemorrhages and
92
stresses the finding of thrombocytopenia. Hutchings et al
studied 87 benzene-exposed individuals in Australian air
force workshops after the discovery of a fatal case of
aplastic anemia. The measured peak benzene concentrations
ranged from 10 to 1400 ppm, and were well above 100 ppm in
most areas. For most of the time, however, the concentra-
tions ranged from 10 to 35 ppm. Solvents contained up to 53
percent benzene. In addition, the worker exposure time was
studied in relation to atmospheric benzene concentrations.
Little difference was observed in the hematological measure-
ments of 87 benzene-exposed workers as compared to those of
500 workers exposed to other hydrocarbons and 300 unexposed
controls, although there was a tendency toward a lower
hemoglobin and platelet counts. The duration of benzene
exposure in these workers is unclear. Hutchings et al con-
clude that the worker who died of aplastic anemia may have
been unusually susceptible to the effects of benzene.
72

-------
Pagnotto et al report a study of benzene exposure in
a rubber coating plant where petroleum naphtha containing
1.5 to 9.3 percent benzene was in use. Atmospheric benzene
levels were generally less than 25 ppm but ranged up to 125
ppm. Correlation between urinary phenol levels and measured
concentrations of benzene in air was excellent. In one
plant, the hemoglobin levels of 5 of 32 men studied were
reduced. One of these men had the second highest urinary
phenol level measured in the plant (480 mg per liter,
equivalent to 58 ppm benzene exposure). In a second plant,
only 1 of 9 individuals studied was anemic, but this worker
was severely affected requiring hospitalization. In a third
plant, none of six workers studied was anemic, nor was
leukopenia observed. Some follow-up data suggesting mild
persistent anemia are presented by the National Institutes
143
of Occupational Safety and Health (NIOSH).
Several studies of groups occupationally exposed to
benzene have been reported in the past decade. Stewart et
172
al (published in abstract only) reported on ten persons
with mild anemia and macrocytosis who had recovered within 8
months following apparent cessation of benzene exposure.
103
Kliche et al reported studies of 18 roof tilers exposed
for an average of 17 years to what is stated to be 15 ppm
benzene. Six of these persons are described as having mild
73

-------
early disease and seven as having chronic benzene hemato-
toxicity. The author also presents indirect evidence of
qualitative abnormalities of platelets despite normal
platelet counts in these subjects.
A number of Eastern European studies of benzene-exposed
53
workers have been reported in recent years. Doskin
evaluated 365 workers employed for 3 years in an apparently
new chemical factory. Although detailed monitoring appar-
ently was performed, the measured concentrations are not
given. The author states that benzene levels exceeded the
maximum permissible concentration 2 to 8 fold in 64 percent
of the measurements in the first year, 37 percent in the
second year, and 3 percent in the third year. The ana-
lytical method and total number of measurements are not
stated. We believe, although we have not been able to con-
firm, that the maximum permissible concentration of benzene
in the Soviet Union at that time apparently was 5 ppm. Pre-
employment and serial hematological measurements were ob-
tained during the 3-year period and were compared with
values from a control group. Approximately 40 percent of
the workers exhibited mild hematological abnormalities
during the first year, and this percentage declined greatly
in subsequent years. The most common early sign of benzene
3
hematotoxicity was mild thrombocytopenia (96 to 155,000/mm )
74

-------
followed by anemia, which was normochromic with a tendency
toward subsequent development of hyperchromia (presumably
because of an increase in mean corpuscular volume) after 1
year of employment. An initial increase in WBC count was
followed in certain cases by leukopenia. In support of this
observation Doskin cites Soviet literature describing a
phasic response to benzene. Except for an early report by
90
Hunter suggesting increased erythropoiesis as an initial
response to benzene, Western investigators have not de-
scribed such a phasic response. Other somewhat different
findings in Doskin's study include observations of lympho-
cytosis rather than lymphocytopenia and relatively greater
effects in younger subjects, in contrast with Aksoy's find-
3
ing of no effect of age on response to benzene. Analysis
of bone marrow of 3 0 workers showed hypercellularity,
particularly in subjects with leukocytosis, a decrease of
megakaryocytes in subjects with thrombocytopenia, and an
increase in lymphocytes in younger subjects. In addition, a
decrease in the phagocytic ability of leukocytes was noted.
Although not substantiated, these findings suggest that
exposure of workers to concentrations of 10 to 40 ppm
benzene for less than 1 year produces mild hematological
effects. Information about the benzene monitoring system
would be of value in interpreting these findings.
75

-------
Kozlova and Volkova"^^ in the Soviet Union have also
used the phagocytic function of leukocytes as an indicator
of benzene hematotoxicity. They studied 252 workers exposed
to benzene during a 5-year period in which benzene con-
centrations initially ranged from 47 to 310 ppm then, as
control measures were improved, decreased to average con-
centrations of 25 to 47 ppm by the end of the period. The
workers were classified in three groups depending upon
exposure levels, the lowest being 24 to 39 ppm. All groups
showed a decrease in cell counts, and severity of the
changes was greatest with higher exposure levels. In the
higher-level exposure group, the extent of the hematological
effects correlated well with duration of exposure. The most
prominent findings were leukopenia, predominantly reflecting
neutropenia, and thrombocytopenia. Changes in red cells
tended to occur relatively late. In most cases a decrease
in phagocytic activity occurred earlier than other hema-
tological effects.
99
Another study from Eastern Europe reported that 16 of
27 workers experienced an increase in levels of red cell delta-
aminolevulinic acid, a precursor in the heme biosynthetic
pathway. In 12 of the 16, earlier benzene exposures are
stated to have ranged from 6.4 to 15.6 ppm and more recent
exposures were to 1.6 ppm. The remaining four workers were
76

-------
exposed only to 1.6 ppm benzene. Blood counts were normal.
The pertinence of these findings must await confirmation of
the authors' hypothesis concerning the effect of benzene on
9 8 99
porphyrin metabolism. ' Also of potential but unproven
importance is the finding by Smolik et al"*"66 of a decrease
in the mean serum complement of 34 workers as compared to
those of a control group. Benzene levels ranged from 3.4 to
6.8 ppm, and duration of exposure, from 3 months to 18
years. Other investigators also have suggested that benzene
alters the immune system in man, including the frequent
observation of lymphocytopenia and reports of altered serum
immuno-globulin levels, the presence of antibodies against
circulating blood cells, and morphologically altered lym-
109 110 147 149
phocytes and monocytes. ' ' '	Based on the immune
surveillance hypothesis, it could be conjectured that a
decrease in immune function plays a role in benzene leu-
kemogenesis. At present, however, the evidence does not
clearly delineate a primary effect of benzene on immune
function in man.
In a study performed in Korea, Chang evaluated hema-
tological toxicity in relation to measured benzene levels
and urinary phenol excretion in workers occupationally
41
exposed to benzene. The author presents data indicating
hematological effects in workers exposed to concentrations
77

-------
as low as 20 ppm; by extrapolation he concludes that "ben-
zene poisoning may occur when workers are exposed to as low
as 10.1 ppm benzene in the air." Unfortunately, many
details are missing in the published account of this study.
After eliminating workers with various disease states from
the study group, the author evaluated 119 individuals said
to be exposed to benzene in an unspecified industrial area.
Benzene levels were measured by a method based on ultra-
violet absorption in ethanol, but the number and duration of
measurements are not given. Urinary phenol was measured
after 4 hours of work by reaction with p-nitroaniline. The
number of such determinations for each worker is not spec-
ified. Of the 119 subjects, 28 showed hematological abnor-
malities: 21 with a normochromic or hyperchromic anemia, 2
with leukopenia, and 5 with both anemia and leukopenia. In
comparison with other workers, those with hematotoxicity
generally had been exposed for a shorter duration at higher
benzene levels. Extrapolation from a plot of benzene con-
centration (in ppm) against interval from start of work
until occurrence of hematotoxicity revealed an exponential
function described as y = (82.5) (0.77^"^X) + 10.1; where y
is ppm benzene and x is work duration in months (Figure 2).
No hematological toxicity was observed in the 18 subjects
exposed to 10 to 20 ppm benzene. The graph suggests that 7
78

-------
ppm
Figure 2. Effect of benzene in air and duration
of work on abnormal blood picture.
* Exponential line is for abnormal blood picture.
79

-------
of the 24 subjects exposed to 20 to 30 ppm benzene exhibited
hematological toxicity after 42 to 96 months of exposure.
At the higher benzene levels apparently 1 of 5 workers
exposed to 100 to 120 ppm benzene and 4 of 13 workers ex-
posed to 100 to 110 ppm benzene developed hematological
toxicity. Of interest is that the average duration of work
for all subjects was substantially less at the higher ben-
zene levels, perhaps implying an occupationally related
attrition. Studies of phenol excretion showed a direct
correlation with measured benzene levels (r = 0.469, p <
0.01). The levels of urinary phenol in this study were
about 2 to 3 times lower than those reported by Elkins for
equivalent benzene concentrations. The author suggests that
this disparity may be due in part to his use of a 4-hour
exposure period rather than the 8-hour exposure used by
Elkins. The urinary phenol findings suggest that the
author has not underestimated the atmospheric benzene con-
centration at the time of the study. This interesting study
presents the most detailed evaluation of benzene hematotoxic-
ity versus dose available in the literature. Unfortunately,
because the information is incomplete, particularly as to
characterization of the work force and exposures, it is
difficult to interpret the relevance of these findings to
benzene exposure in other populations.
80

-------
Numerous cases of benzene hematotoxicity in Italy have
been reported, particularly by Vigliani, Saita, Forni, and
.. •	58-62,153-160,187-192 m.	, . ,
their colleagues.	The accumulated case
reports in Milan and Pavia among shoe workers and other
occupationally exposed groups have been reviewed recent-
18 8 191
ly ' and are described below. Measured benzene con-
centrations have ranged from 25 to 1500 ppm and often have
been over 200 ppm.
Among the many French investigators who have reported
benzene hematotoxicity, Girard and his colleagues present an
extensive series of investigations.^^	Their studies
include the frequent observation of a decreased leukocyte
alkaline phosphatase in a group of 319 workers exposed to
6 6 6 7
benzene levels said to range from 10 to 25 ppm. ' They
have also noted a statistically significant occurrence of a
history of occupational benzene exposure in patients with
71 72
aplastic anemia as compared to a control group. '
Studies of Aksoy et al
Aksoy and his colleagues in Turkey have performed one
of the more extensive evaluations of individuals with ben-
3-15
zene hematotoxicity.	Most of the cases observed were in
shoe workers who between 1955 and 1960 began using an adhe-
sive containing high levels of benzene. Individual cases of
aplastic anemia were first noted in 1961. By 1977 Aksoy and
81

-------
his colleagues had studied 46 patients with this disorder,
of whom 14 had died of aplastic anemia, 5 had developed
acute leukemia, 1 developed myeloid metaplasia, 22 were in
complete remission, 2 were still under treatment, and 2 were
lost to follow-up."^ This compilation includes only pa-
tients personally observed by Aksoy et al. Thirty-five were
shoe workers and the remainder were also exposed to adhe-
sives -containing 9 to 88 percent benzene. Exposure was
generally in small shops with poor ventilation. Where
measured, the benzene concentrations ranged from 150 to 650
ppm.
Aksoy and his colleagues also present a thorough study
of hematological findings in 217 apparently healthy male
3
shoe workers in comparison with 100 control subjects.
Fifty-one of the 217 exposed were considered to have a
benzene-associated hematological abnormality. Forty-one
were afflicted with leukopenia and/or thrombocytopenia; the
other ten are included on the basis of various abnormalities
including eosinophilia, basophilia, giant platelets, lympho-
cytosis, and an acquired Pelger-Huet anomaly. The latter is
a morphological abnormality seen in leukemia and reported in
association with benzene exposure. The authors also report
that 33 percent of the total work group was anemic, but
because the hematocrit readings returned to normal following
82

-------
iron therapy in those treated, the anemia is not defini-
tively ascribed to benzene toxicity. Macrocytosis was not
observed, but the absence of this finding may be due to the
complicating iron deficiency anemia, which usually leads to
microcytosis. No difference in the relative incidence of
benzene hematotoxicity was noted in different age groups.
Of note was a tendency toward a higher incidence of hemato-
toxicity in workers exposed for less than 1 year as compared
with those exposed for longer periods. The authors suggest
that this could be due to all hematological changes occur-
ring or starting in the first year of exposure, or to a loss
of affected individuals from the work force.
Aksoy's group has also detailed the hematological
findings in 32 patients with clinically significant pancyto-
5
penia. All used adhesives containing benzene. Ambient
levels of benzene were said to range from "15 and 30 ppm
outside working hours and were recorded to reach 210 or,
rarely, 640 ppm when benzene containing adhesives were being
used." Cellularity of the bone marrow was decreased in 12
cases, normal in 12, and increased in 7. In the remaining
case, giant erythroid precursors were observed, perhaps
indicating preleukemia. The severity of disease and likeli-
hood of fatal outcome was greatest in the group with hypocel-
lular bone marrows, although the duration of exposure ap-
83

-------
peared unrelated to the bone marrow findings. Macrocytic
erythrocytes were observed in 14 patients, most of those
being patients with hyperplastic bone marrows. Nucleated
erythrocytes were noted in the peripheral blood smear of six
patients. Other red cell abnormalities included a mild to
moderate increase in the osmotic fragility, with or without
24-hour preincubation, in 13 of 20 subjects tested; an
increase in fetal hemoglobin in 20 of 24 patients; and a
decrease in Hb A^ in 3 of 24 patients. This group has
described additional studies suggesting an increase in
4 9
hemoglobin F in benzene hematotoxicity. ' Evidence of a
shortened red cell life span included a slight reticulo-
cytosis, mild hyperbilirubinemia, and increased urobilinogen
in some subjects, and a moderately shortened radioactive
chromium survival in the one subject in whom this was mea-
sured. Except for the latter finding, these observations
could also be due to ineffective erythropoiesis. White
cell abnormalities included an absolute lymphocytopenia in
24 subjects, a slight increase in monocytes in 4 subjects,
and the development of a Pelger-Huet anomaly in 1 subject
following recovery from pancytopenia. A low platelet count
(thrombocytopenia) was observed in 28 individuals, and the
peripheral blood smears showed giant or morphologically
abnormal platelets.
84

-------
Miscellaneous Abnormalities in Benzene-Induced Fancytopenia
There is evidence that in addition to producing a
decrease in number of circulating blood cells, benzene
exposure causes formation of abnormal red cells, platelets,
and white cells. The most common red cell alteration is
macrocytosis, a condition that also occurs in erythroleuke-
mia and in vitamin deficiency states associated with ab-
normal metabolism of nucleic acid. Megaloblastic and
macroerythrocytic precursors have been noted in the bone
marrow of patients with benzene toxicity. ^ ^Abnormal
red cell function is also suggested by the finding of
significant hemolysis in occasional cases of benzene hemato-
. . .. 19,55,57,91,119,137,149,156 .	. _ ,	.
toxicity,	by a report of abnormal
7	78
osmotic fragility in one but not another study, and by
indirect indications of altered heme synthesis, including
changes in levels of porphyrin in blood and urine and in
levels of delta aminolevulinic acid in red cells . ^ ^
A number of investigators report abnormal morphology and
function of granulocytes. Observations include a decrease
106
in phagocytic function, a loss in leukocyte alkaline
phosphatase, ^^ an altered osmotic fragility,"'"^"'" a change
105
in the fluorescence characteristics of leukocyte nuclei,
a -n i u 4-	i 3,157,164,195 „ ,
and a Pelger-Huet anomaly.	Many studies sug-
gest that benzene hematotoxicity results in altered platelet
85

-------
.. 29,44,56,96,103,136,158,159,162,169 , ,
function,	perhaps leading
to an increased susceptibility to bleeding. Morphological
abnormalities of circulating platelets and of megakaryocytes
have also been reported. ^	^ No direct evidence of
altered lymphocyte function is available, although this has
been suggested on the basis of apparent alterations in
. .. • ,	, •	, 109 ,110 , 147, 148 ,166
immune function m benzene-exposed individuals.
At present, however, it is unclear whether the reported
immunological findings are a primary effect of benzene, are
secondary to the formation of altered hematopoietic cells,
or reflect some unrelated situation. In addition, cyto-
genetic aberrations in circulating lymphocytes have been
observed (see Section 1). Abnormal monocytes in the blood
151
of workers exposed to benzene also have been reported.
Many of these abnormalities in circulating red cells,
white cells, and platelets are reported to occur as rela-
tively early manifestations of benzene hematotoxicity. For
this reason they have been proposed for use in screening
tests to detect early benzene effects. An absolute lympho-
cytopenia is noted as a relatively early indication of
benzene hematotoxicity by sonie^'^ but not other^'^ in-
vestigators. An increase in circulating eosinophils and
, . , . ,	. .	,	, 3,5 ,25,32,89,90 T
basophils is also sometimes observed.	In
addition, the serum levels of various enzymes^' are
86

-------
reported to be altered in early benzene hematotoxicity.
Further systematic study of possible indicators of benzene
hematotoxicity would be of value in screening exposed indi-"
viduals for the detection of early damage to bone marrow.
Although animal studies provide evidence of rapid
interference in erythropoiesis following administration of
benzene, the relatively long normal red-cell survival period
of 120 days causes any diminution in red cell production to
be manifest only slowly as a decrease in circulating red
cells. In contrast, the normal period of platelet survival
is 7 to 10 days, and survival of granulocytes in the cir-
culation is perhaps 2 4 hours. Although a red cell count
measures almost all mature red cells, about 30 percent of
platelets are sequestered in the spleen, and the measurable
circulating granulocytes represent less than 20 percent of
the total in the body. As a further complication, there is
a relatively large reserve of bone marrow, which is capable
of perhaps a sixfold greater output of mature cells in a
normal situation. Accordingly, counts of the circulating
cells do not give an adequate index of very early benzene
hematotoxicity.
Long-Term Evaluations
The prognosis in mild cases of pancytopenia is good if
benzene exposure is discontinued. In some individuals,
87

-------
however, pancytopenia has progressed even after presumed
cessation of exposure, and apparent hematological recovery
has sometimes been followed by acute leukemia occurring as
4 9 97
late as 15 and 27 years after the initial findings. '
Relatively few longitudinal assessments of occupational
exposure groups have been performed. One such study was
89
reported by Hernberg et al, who reevaluated a group of
individuals exposed to benzene in a shoe factory, originally
studied by Savilahti.Benzene had been in use for about
10 years, and levels close to 400 ppm were measured. The
original investigation disclosed abnormal blood counts, most
commonly thrombocytopenia, in 107 of 147 workers studied.
Ten of these individuals required hospitalization, one of
whom died with severe pancytopenia. Of note is that one of
the more severely affected individuals, whose peripheral
blood counts returned to normal in 2 years, subsequently
developed acute leukemia. Hernberg et al restudied 125 of
the original workers in comparison with a nonexposed control
group 9 years after cessation of benzene exposure. There
was definite improvement in mean platelet count, which
increased from 176,480 to 257,360. However, the platelet
count in the group exposed to benzene remained significantly
(p<0.01) lower than that of the control level of 293,000 per
mm . The mean erythrocyte count for male workers was stable
88

-------
during this 9-year period and was also significantly lower
than that of the control group (4.4 vs 4.7 million per ;
p<0.01). Improvement to control levels was observed in the
mean erythrocyte count of female workers and the mean
leukocyte count of the entire group exposed to benzene.
These findings suggest that recovery after benzene hemato-
toxicity may occur very slowly or may be incomplete. Of
interest is the observation that the reported mean platelet
and red cell counts in the exposed population were within
normal limits, and the observation of a persistent benzene
effect in this population required simultaneous study of a
normal control group. A possible inference from this find-
ing is that peripheral blood counts within the wide range of
normal do not preclude an effect of benzene on hematopoietic
89
tissue. A further observation by Hernberg et al is that
the severity of the original findings in 1955 did not cor-
relate with the extent of recovery in 1964. This analysis,
however, was not based on the blood counts of individual
workers but rather on a comparison of the initially severe
with the initially mild cases.
Another long-term follow-up of an occupationally ex-
8 3
posed group has been reported by Guberan and Kocher.
Their evaluation of 216 of 282 workers 10 years after ap-
parent cessation of benzene exposure revealed two indi-
89

-------
viduals with isolated thronbocytopenia, one with anemia and
thrombocytopenia, and one vzith mild pancytopenia. They note
that one worker had died of aplastic anemia 9 years after
cessation of benzene exposure. Few additional details are
provided. Further study of the long-term consequence of
occupational exposure to benzene in well-defined cohorts
would be of great interest.
LEUKEMIA
Leukemia can be defined as a neoplastic proliferation
and accumulation of white blood cells in blood and/or bone
marrow. Hematologists generally agree concerning char-
acterization of the four main types of leukemia: acute and
chronic myelogenous (also known as granulocytic) leukemia,
and acute and chronic lymphocytic leukemia. This clas-
sification represents diseases that differ in terms of
incidence, course, prognosis, and, presumably, etiologic
factors. In addition, there are other types of leukemia,
related to these four major types, about which there is some
disagreement concerning diagnostic criteria. Erythroleuke-
mia, acute promyelocytic leukemia, stem cell leukemia, and
acute myelomonocytic leukemia, all of which have been re-
ported in association with benzene exposure, are generally
considered to be variants of acute myelogenous leukemia.
90

-------
There is, however, some question as to whether the rela-
tively rare acute monocytic leukemia (Schilling type) is an
entity of its own or is related to acute myeloblastic
leukemia.
Chronic myelocytic leukemia has often been classified
under the heading of myeloproliferative syndrome, which
includes polycythemia vera, myelofibrosis and myeloid meta-
plasia, and essential thrombocythemia. These disorders have
in common a neoplastic proliferation of the pluripotential
stem cell responsible for the formation of granulocytes,
platelets, red blood cells, and perhaps, fibroblasts and
other monocytic cell types. They also share a potential for
the development of acute myeloblastic leukemia. Although
the myeloproliferative disorders may overlap clinically,
chronic myelogenous leukemia does appear to be distinct and
readily separable on the basis of a characteristic chro-
mosomal abnormality and other distinguishing features. The
cells identifiable in the blastic crises of chronic myelo-
genous leukemia may be more closely related to lymphoblasts
than to myeloblasts, again consistent with the pluripoten-
tial nature of the stem cell discussed above.
Investigators have associated exposure to benzene with
preleukemia. Preleukemia is very difficult to define, and
there is little agreement as to clinical criteria. In most
91

-------
cases, the diagnosis is made in retrospect after the patient
has developed a clearcut acute leukemia, usually of the
myeloblastic type. A characteristic case history would be a
benzene-exposed individual with pancytopenia whose bone
marrow demonstrates generalized hypoplasia, but with a
slight increase in somewhat atypical blast cells. Over a
period of weeks to years, repetitive bone marrow measure-
ments show a gradual increase in these blast forms until
frank acute myeloblastic leukemia is indicated by almost
complete blastic replacement of the bone marrow and by the
presence of these cells in the peripheral blood. Preleu-
kemia is defined as beginning at the point at which there
is a reasonable expectation that the patient would develop
acute leukemia and ending at the point at which acute leu-
kemia is diagnosed. Despite much current interest in the
subject, no criteria are agreed upon to define these two
points. Accordingly, it is difficult to evaluate literature
concerning preleukemia in benzene-exposed individuals beyond
• J- 4.-	4- U	4-"	*4-	4- ^ 6,10,71,115,173
indicating that such a continuum is often reported.
Relationship of Benzene to Leukemia
The evidence concerning the relationship of benzene and
acute leukemia has been reviewed by a number of authors and
panels in recent years. Most of these individuals and
groups accept causative role for benzene in human acute
92

-------
myelogenous leukemia and it is so described in the routine
hematological literature. The following quotations are from
three standard American textbooks on hematology:
"A variety of chemicals and drugs have been sug-
gested as possible leukemogenic agents in human leuke-
mia, but only benzol can be unequivocally implicated."^^
"Any chemical capable of producing myelotoxicity
must be regarded as a potential leukemogen, if the
findings in radiation-induced leukemia apply which
indicate that cell damage with depression of marrow
function may produce alterations leading to the trans-
formation of damaged cells into neoplastic ones. The
only chemical which has been clearly identified as one
which increases the incidence of myeloid leukemias in
man is benzene in rather heavy occupational exposure.
The development of stem cell, erythroblastic and
myeloblastic leukemias and persistent chromosome ab-
normalities in exposed individuals known to have had
1 c 9
neutropenia is especially significant. "J--,z
"Sporadic cases of acute leukemia have occurred
after exposure to chemical agents such as benzene,
phenylbutazone, and chloramphenicol. Aplastic anemia
and acute myelogenous leukemia have followed prolonged
exposure to benzene. About 15% of the patients who
develop hematologic abnormalities due to benzene ex-
posure also develop acute leukemia. Leukemia is
usually preceded by a period of bone marrow aplasia."3 3
Other assessments of this relationship, however, leave
some room for doubt. Thus, a recent document from the
197
National Cancer Institute concluded:
"At best benzene must be considered as a suspect
leukemogen."
In evaluation of the causal relationship of benzene to
acute leukemia, three types of information are pertinent:
the nonleukemic effects of benzene, case reports of benzene-
associated leukemia, and results of epidemiological studies.
93

-------
Nonleukemic biomedical effects of benzene - Benzene
clearly damages hematopoietic tissue in man. That this
damage can be due to benzene alone is supported by observa-
tion of pancytopenia in diverse exposures in many different
countries, as well as by animal studies in which administra-
tion of benzene in the absence of any other solvent produces
pancytopenia (see Section 3). A somewhat negative point is
the absence of conclusive evidence of leukemogenic action of
benzene in animals. Arsenic may be the only other compound
for which there is good epidemiological evidence of carcin-
ogenesis in man in the absence of an analogous animal model.
The fact that patients with idiopathic aplastic anemia
or aplastic anemia developing from other agents, most
notably chloramphenicol and phenylbutazone, may also prog-
23 131
ress to acute leukemia ' provides inferential evidence
that benzene-induced damage to hematopoietic stem cells
could lead to acute leukemia. The mutagenic effects of
benzene, particularly in relation to chromosomal abnor-
malities in man, may also be considered inferential evidence
supporting a causal role of benzene in leukemia (see Section
2) .
Case reports of leukemia associated with benzene - Well
over 100 cases of leukemia in benzene-exposed individuals
have been described in the literature since the original
94

-------
report in 1928. However, enumeration of case reports does
not, by itself, provide definitive evidence that benzene is
a causative factor in acute leukemia. There is a finite
probability that each case represents a chance association
with benzene exposure in an individual whose leukemia is due
to some other cause. Reports of single cases and small
series of cases generally lack information concerning the
size of the population at risk that is required for firm
conclusions. On the other hand, a number of factors would
tend to lead to an underreporting of benzene-associated
leukemia. The general acceptance of this relationship in
the medical literature for some time would tend to hinder
the preparation or publication of manuscripts describing one
further case. A long period often folloys benzene exposure
before the onset of acute leukemia; this delay could lead to
the relationship being overlooked by the patien- or physi-
6,49,82,97,114,148,160,164,191 _ ..	,
cian.	Furthermore, because
benzene is a ubiquitous component of our chemical age,
exposure may be unrecognized. Accordingly, simple quanti-
tation of case reports in the medical literature cannot
provide definitive information concerning the causal rela-
tionship of benzene to acute leukemia.
These case reports, however, do provide presumptive
evidence of a leukemogenic effect of benzene. Of particular
95

-------
note is the frequent description of persons suffering from
benzene-associated pancytopenia in whom evolution to acute
leukemia was observed. Idiopathic aplastic anemia is an
uncommon disorder, reported far less frequently than acute
myelogenous leukemia. The relatively frequent documentation
of benzene-associated pancytopenia progressing to acute
leukemia, which is in keeping with that observed in other
causes of aplastic anemia, further supports the possibility
that exposure to benzene increases the risk of developing
acute leukemia. Similarly, the case reports are also per-
tinent for the frequency with which evolution of the hema-
tological findings progress to or through an erythroleukemic
10,20,51,52,59,63,104,108,137,173 m. ,
stage.	The relative oc-
currence of the erythroleukemia variant in these case re-
ports appears to be more common than that observed in cases
of acute myeloblastic leukemia and its variants in the
general population. Unless this represents some unexpected
bias in the reporting process, the frequency with which
erythroleukemia is observed in benzene-exposed patients also
supports a specific leukemogenic action of benzene.
Epidemiological Studies - Evidence supporting a role for
benzene in leukemia has been obtained in a number of epi-
demiological studies. Three general approaches have been
used. First, the discovery of a relatively large number of
96

-------
individuals with leukemia having a history of occupational
benzene exposure has led to assessment of the likelihood of
this occurrence. This has been done by estimating the
number of workers in the occupational group and computing
the incidence of leukemia in comparison with that expected
in the general population. A second approach has been to
obtain occupational histories of persons with leukemia and
compare the incidence of potential benzene exposure in these
leukemics with that of control populations. The third
approach has been to evaluate the mortality characteristics,
including leukemia deaths, of relatively large populations
working in an industry involving known benzene exposure.
Any high incidence of leukemia can be further analyzed in
terms of occupational subgroup, duration of employment, and
other factors.
Each of these approaches entails advantages and disad-
vantages. For example, the third approach, representing a
standard epidemiological technique, is by far the most
thorough in its characterization of an entire work force;
however, the validity of information on leukemia incidence
depends on the vagaries of death certifications rather than
on direct observation. In contrast, enumeration of the
population at risk in the first approach represents a crude
estimate, and little is known about the population char-
97

-------
acteristics of the occupational group. This approach does
offer a high level of confidence in the diagnosis of leu-
kemia, in that the subjects have been observed by the in-
vestigators .
The relatively recent studies of Aksoy and his col-
leagues in Turkey strongly support the causal relationship
* U	4. 1 ,	6-, 10,11,13,14,16
of benzene exposure to acute leukemia.
Their evidence includes individual case reports of workers
with aplastic anemia which progresses through a preleukemic
phase to frank acute myeloblastic leukemia or erythroleuke-
mia; an accumulation of cases resulting in a statistically
significant higher incidence of acute leukemia among shoe-
workers; and an outbreak of leukemia in this population that
appears temporally related to the onset of benzene use and
that has subsided following replacement of benzene as a
solvent for adhesives. From 1967 to 1973 Aksoy et al ob-
served 26 patients with acute leukemia among shoe workers.^'
Analysis in selected work areas revealed that "the con-
centration of benzene was found to reach a maximum of 210-
650 parts per million when adhesives containing benzene were
in use." There were 14 cases of acute myeloblastic leukemia,
4 of preleukemia, 3 of acute erythroleukenia, 3 of acute
lymphoblastic leukemia, and 1 each of acute promyelocytic
and acute monocytic leukemia. Duration of benzene exposure
ranged from 1 to 15 years. The authors state that there
98

-------
were 28,500 shoe workers at risk in Istanbul and derived a
leukemia incidence of 13 per 100,000 in this population.
This incidence is statistically significantly higher (p<0.02)
than the risk of 6 per 100,000 assumed for the general
population. Because this latter number apparently is de-
rived from the incidence of leukemia in developed nations,
rather than being specific to Istanbul, there is some degree
of uncertainty. This appears to be more than counterbal-
anced, however, by a number of factors. Foremost is that
the distribution of cases reported by Aksoy et al strongly
differs from that of leukemia in the general population. If
the relative incidence were computed solely for acute
myeloblastic leukemia and its variants, a magnification of
the risk in benzene-exposed shoe workers would be observed.
Secondly, Aksoy et al apparently have not age-adjusted their
findings. In their series, the average age at the time of
diagnosis was 34.2 years. This is a relatively low-risk age
period for leukemia, with a reported death rate about half
43
of the overall incidence. Recalculation of their data
with an age factor would presumably increase the statistical
significance of the findings. In addition, Aksoy has re-
cently stated that his studies underestimate the relative
risk of leukemia because the incidence of leukemia in the
general population of Turkey is 2.5 to 3.0 per 100,000 and
99

-------
because shoeworkers with acute leukemia probably were
admitted to other Istanbul hospitals without his knowledge.
Recently, Aksoy presented his observations of acute
16
leukemia in shoe workers during the period 1967 to 1976.
As shown in Table 1, the peak incidence of leukemia in shoe
workers occurred between 1971 and 1973. This follows by a
few years the appearance of a notable incidence of aplastic
anemia in this occupational group. The decline in cases
since 1973 is temporally related to a decrease in use of
benzene as an adhesive solvent, which began gradually in
1969. Aksoy also reports that pancytopenia was present in
27.5 percent of the cases before the onset of acute leu-
kemia, which occurred 6 months to 6 years later. The hema-
tological findings often indicated a period of recovery
before the onset of leukemia, a phenomenon also noted by
other investigators. Aksoy states that during- this period,
over 100 cases of aplastic anemia were observed that were
either idiopathic or associated with an agent other than
benzene, and in none of these cases did acute leukemia
develop. He also states the opinion that no blood dyscrasia
is required before the onset of leukemia and provides an
example of a 2 3-year-old shoeworker who was hematologically
normal when studied 4 years before the onset of acute
erythroleukemia. As with other cases of leukemia associated
100

-------
Table 1. ANNUAL NUMBER OF LEUKEMIA CASES AMONG
SHOEWORKERS WITH CHRONIC EXPOSURE TO BENZENE
IN ISTANBUL BETWEEN 1967 AND 1975 '16)
Number of
leukemic
Years	shoeworkers
1967	1
1968	1
1969	3
1970	4
1971	6
Number of
leukemic
Years	shoeworkers
1972	5
1973	7
1974	4
1975	3
1976	0
Four other leukemic workers with different jobs and two leukemic
individuals outside of Istanbul are not included in this series.
101

-------
with benzene exposure in which the patient had no detectable
pancytopenia beforehand, the interval between hematological
observations is too long to ensure that pancytopenia did not
in fact occur.
A relationship between benzene exposure and leukemia is
stressed in a series of reports by Vigliani and his col-
188 191
leagues in Northern Italy. '	In recent review arti-
cles, Vigliani and Vigliani and Forni summarized their
187 188
experience from 1942 to 1975. '	They observed 66 cases
of significant benzene hematotoxicity in Milan, of which 11
were cases of acute myeloblastic leukemia or its variants.
All 11 leukemia patients died, and 7 subjects died of
aplastic anemia. In Pavia between 1959 and 1974 there were
135 patients with benzene hematotoxicity; 13 died of acute
myeloblastic leukemia or its variants, and three of aplastic
anemia. All of the patients in Pavia were workers in the
shoe industry. Measured benzene concentrations in the
breathing zone of workers using glue usually were 200 to 500
ppm, and ranged from 25 to 60 0 ppm. Many factory shoe
workers also worked at home, where presumably further ben-
zene exposure occurred. Shoe workers were also the largest
single occupational group with hematotoxicity in Milan,
although many other working groups were also affected. Of
interest is an apparent outbreak of benzene hematotoxicity,
102

-------
including eight cases of severe pancytopenia and two cases
of myelogenous leukemia, in the rotogravure; industry. This
outbreak was temporally related to the use of inks and
solvents containing large amounts of benzene. Ambient
benzene levels were calculated to be between 200 and 400
ppm, with peaks up to 1500 ppm. In an earlier study,
191
Vigliani and Saita estimated the number of workers ex-
posed to benzene in Pavia and Milan and, based on the
incidence of acute leukemia in the general population of
Milan, calculated a 20-fold higher risk of acute leukemia in
these workers.
An interaction between benzene and radiation in leuke-
94
mogenesis is suggested by the study of Ishimaru et al.
These authors performed a retrospective study of survivors
of the two atomic bombings in Japan, evaluating the effect
of occupation on the incidence of leukemia. Controls were
matched with patients by age, sex, residence, and distance
from the atom bomb explosion. The occupational history of
leukemic patients was obtained from relatives. Of 492 cases
of leukemia through 1967 in this population, adequate
histories and appropriate controls were obtained for 413, of
whom 303 were adults. Comparison of the controls and
leukemic patients revealed that the risk of leukemia was 2.5
times higher in those with an occupational history poten-
103

-------
tially related to benzene or X-rays. The risk was signifi-
cantly higher in those with 5 or more years of potential
exposure but not in those who had been employed in such
occupations for less than 5 years. The relative risks were
similar in Hiroshima and Nagasaki and were higher for acute
leukemia (2.9) than for chronic leukemia (1.8). The major
limitation of this study is the possible inaccuracy of the
occupational history of leukemic subjects as compared with
that of the living controls.
71 72
Girard et al ' have evaluated the frequency of a
positive history of benzene exposure in 401 patients hospi-
talized patients with serious hematological disorders as
compared with 124 patients hospitalized for nonhematological
problems. A statistically significant increase (p < 0.05;
2
X > 3.84) in history of benzene exposure was noted for
2
patients with aplastic anemia (10 of 48, x = 12.2), acute
2
leukemia (17 of 140; x = 5.6), and chronic lymphocytic
2
leukemia (9 of 61; x = 6.7) as compared with the control
group (5 of 12 4).
The rubber industry has long been characterized by
exposure to various solvents including benzene. Pancyto-
penia associated with benzene exposure ha? been noted in
workers in a number of countries (see above). In the United
193
States, Wilson reported a large cohort of individuals
104

-------
exposed to benzene during the expansion of rubber production
early in World War II. Some degree of cytopenia was noted
in 83 of 1104 individuals studied. Leukemia was not re-
ported. Ascertaining the present vital status and causes of
death in this cohort, particularly in relation to leukemia
incidence, could be of great value.
Other studies of workers in the rubber industry include
a report of the U.S. Department of Health, Education, and
183
Welfare.	This report cites a 54 percent increase in the
death rate for cancers of the lymphatic and hematopoietic
system in male rubber-industry workers dying in 1950 as
compared with workers in all manufacturing industries.
There were also a 30 percent increase in death rate due to
large bowel cancer, a 19 percent increase due to cancers of
the respiratory tract, and an 8 percent higher overall
cancer death rate.
117 118
Mancuso and his colleagues ' performed a series of
studies of occupational cancer rates, with particular
emphasis on the rubber industry, including evaluation of a
cohort of 1977 workers. No apparent increase in hemato-
poietic cancers was observed. There was, however, a higher
incidence of tumors of the gallbladder, bile duct, and
salivary gland in occupational subgroups not associated with
high solvent exposure.
105

-------
17 1ft 194—197 1 ft f)
A comprehensive series of studies ' ' "	' of
the health status of rubber-industry workers has been per-
formed by the Occupational Health Studies Group of the
University of North Carolina. They have evaluated the 10-
year mortality experience of a large cohort of male workers
(5106 deaths) at four tire manufacturing plants. The
subjects were in the work force or were retired in 1964.
18 0
This series of studies was reviewed recently by Tyroler.
The mortality due to all cancers (1014) was normal or
slightly elevated, depending on the data base used for
comparison. Deaths due to cancer of the lymphatic and
hematopoietic system (total of 109) were 31 percent higher
than expected and were increased in cohorts of each of the
four companies. In the category of lymphosarcoma and
Hodgkin's disease, the standard mortality ratio (SMR) was
129 and an increase in the expected number of deaths was
observed in two of the four company cohorts. Similarly, for
deaths due to all forms of leukemia the SMR was 130 and the
increase was observed in three of the four cohorts. When
this latter category was further subdivided, the overall SMR
for lymphatic leukemia was found to be 158 and the expected
death rate was elevated in two of the four company cohorts.
Of particular note is that the SMR for deaths attributed to
lymphatic leukemia was 291 in the age group 40 to 64.
106

-------
Several approaches were followed in further study of
the increased incidence of lymphatic leukemia. Contrasting
the work history of 17 patients with lymphatic leukemia with
those of three matched controls for each case revealed that
solvent exposure increased the overall risk by a factor of
3.25. Further classifying the groups according to high,
low, and medium solvent exposure, yielded a 5.5 factor for
the high-exposure group. In those patients first subjected
to high exposure between 1940 and 1960, the factor for the
relative risk of lymphatic leukemia was 9.0. The relation-
ship of solvent exposure to lymphatic leukemia was statis-
tically significant at p < 0.025. The study also showed an
increase in the mean difference in years of work history
between lymphatic leukemia and the case controls. This was
inversely proportional to the extent of solvent exposure.
Tyroler has cited several epidemiologic reasons why these
18 0
findings must be interpreted with caution.	These include
the dependence on death certificates for diagnostic informa-
tion, which apparently is now being validated with clinical
and pathological findings. Tyroler notes good agreement
between the death certificate and medical information con-
cerning deaths stated to be due to leukemia. It would be
pertinent to learn whether the deaths certified as caused by
lymphatic leukemia represent cases of chronic lymphatic
107

-------
leukemia, a relatively common disorder in this age group, or
of acute lymphoblastic leukemia. The latter is rare in
adults and is sometimes difficult to distinguish from the
acute myeloblastic and stem cell leukemias that are clas-
sically associated with benzene exposure. Other conceivable
limitations to the studies by the University of North
Carolina group include the possibility that the association
between work history and nonfatal cases in this cohort might
be systematically different from that observed in fatal
cases, and that individuals with leukemia and a work history
of solvent exposure may also have had uncontrolled exposures
to radiation, drugs, or other environmental agents that
might cause the observed increase in leukemia rates. A
further limitation is the lack of historical data concerning
the benzene exposure or the concentrations of other sol-
vents. These studies do, however, strongly support the
possibility that long-term exposure to benzene in the U.S.
rubber industry leads to an increased risk of lymphatic
leukemia.
Monson and Nakano also have recently studied mortality
of workers in the rubber industry.	Evaluation of a
cohort of 13,571 white males in one plant in Akron, Ohio,
revealed an SMR of 82 for all causes of death and an SMR of
94 for all malignant neoplasms in comparison with U.S.
108

-------
134
mortality statistics.	When the causes of death were;
further subdivided by types of neoplasm, the highest SMR
(128) was observed for leukemia (43 expected, 55 observed) .
This high incidence was observed particularly in workers in
the tire (SMR 150) and processing (SMR 240) divisions, but
was also observed in other worker subgroups. The excess in
leukemia was notable in those who were less than 25 years of
age at the time of employment and in those who began work
before 1935. However, reevaluating the data for the 7374
employees who had worked at least 25 years gave no evidence
of a decrease in risk for those who began working after
1935. The SMR for nonleukemic hematological neoplasms
(lymphatic tumors and myeloma) was 101 in the overall cohort
and 160 in the workers employed in the tire building area.
Studies were also performed in three additional cohorts from
this plant; female workers, black male workers, and salaried
135
white males.	Data on these groups showed a tendency
toward an increase in deaths due to lymphatic tumors and
myeloma (SMRs 111,125, and 115, respectively). Deaths due
to leukemia were not greater than expected, although gen-
erally higher than the SMR for all causes of death. Evalua-
tion of 574 deaths in white males working in other factory
locations at the same rubber company revealed 14 deaths due
to all lymphatic and hematopoietic tumors as opposed to 11.8
expected.
109

-------
Other epidemiological studies of" .individuals poten-
tially exposed to benzene have failed to reveal an increased
incidence of acute leukemia. Thorpe reported an extensive
study of leukemia mortality rates for the period 1962-71 in
36,000 employees and annuitants of eight European affiliates
176
of a major petroleum company.	The report is noteworthy
for the extensive discussion of methodological problems
inherent in such retrospective studies. These problems
include the low incidence of leukemia in the general popula-
tion; the general inability to quantitatively define benzene
exposure in the selected population; the difficulty of
obtaining both accurate occupational histories and complete
follow-up of the workers; and the problems of verifying the
diagnoses of leukemia and of the subtype of leukemia.
In Thorpe's study four of the affiliates reported 18
cases of leukemia, whereas the other four reported none. In
only 6 of the 18 cases was the subtype of leukemia known.
Benzene exposure was believed to have occurred in 8 of the
18 individuals although the author points out the difficulty
of clearly distinguishing among the job categories relative
to benzene exposure. The data revealed no statistically
significant increase in the leukemia rate over that expected
in the population. There was, however, a tendency toward
higher leukemia rate in the benzene-exposed as compared to
110

-------
the unexposed work groups. This study has been criticized
38	175
by Brown and defended by Thorpe.	The major criticism
concerns factors possibly leading to an underreporting of
leukemia incidence in this population.
Although benzene is associated with coke oven opera-
tions, study of cancer mortality in coke plant workers has
revealed no statistically significant increase in leukemia
14 4
incidence.	This is also true for coke by-product workers,
the occupational subgroup expected to be subjected to ben-
zene exposure. These workers, however, constitute a rela-
145
tively small group. Redmond et al studied 34 5 indi-
viduals working 5 or more years in a coke by-product area.
They found only 10 fatal neoplasms, of which 2 were of the
lymph and hematopoietic tissue category (1.1 expected).
Accordingly, no firm conclusions can be reached concerning
benzene leukemogenesis in coke oven workers.
A risk of benzene exposure might also be expected in
other occupations that are not well characterized epide-
miologically. Adelstein examined the standardized propor-
tional mortality ratios for various cancers in British
males, 1959-1963, listing them according to 27 occupational
groups."'" The highest level of leukemia deaths was in the
group described as "professional, technical workers, art-
ists" (p < 0.01). This presumably would include chemistry
111

-------
professors, laboratory technicians, and other professional
groups with relatively frequent exposure to benzene. Art-
ists may be particularly at risk because they use unknown
solvent mixtures in poorly ventilated and unregulated work
areas. As with other studies, however, an exposure to
benzene cannot be defined and it is possible that other
factors, such as radiation, may be responsible for the
increased incidence of leukemia in this occupational group.
In the United States, Li et al"*"^ evaluated the cause
of death of 3637 members of the American Chemical Society
who died between 1948 and 1967. Among deaths occurring
between ages 20 and 64, an increase in all malignant neo-
plasms was noted (444 observed, 354 expected, p<0.001) as
compared with a control group derived from U.S. professional
men. The increase in tumors of the lymphatic and hemato-
poietic system (94 observed, 50 expected) was the most
highly statistically significant (p < 0.001) of all tumor
types. Further subdivision revealed that the major increase
was in nonleukemic hematopoietic and lymphatic tumors (61
observed, 34 expected, p < 0.001). The difference in number
of leukemia cases (33 observed, 25 expected) was not statis-
tically significant. Among deaths occuring after age 64, in
comparison with U.S. white males, the highest level of
statistical significance again was observed for all tumors
112

-------
of the lymphatic and hematopoietic system (p < 0.001) . In
this analysis, the increased incidence of both leukemic (16
observed, 9 expected) and nonleukemic (17 observed, 9 ex-
pected) hematological neoplasms was statistically signifi-
cant .
A recent and relatively thorough epidemiological eval-
uation of benzene-exposed workers was conducted by Infante
93
et al of NIOSH. These investigators evaluated the mor-
tality of 7 48 white male workers exposed to benzene in
manufacture of a rubber product. The cohort consisted of
those employed at any time during the period 1940 through
1949, and the period of risk for death was January 1, 1950,
through June 30, 19 75. The two control groups were the
general population of U.S. white males and a group of 1447
white males employed in the same state and time period for
at least 5 years in manufacture of a fibrous glass product.
The vital status of this second control group was determined
as of June 1, 1972. In the benzene-exposed group, data
concerning the vital status as of June 30, 1975, were avail-
able for approximately 75 percent of the workers. The
remaining 25 percent were assumed to be living. This
assumption leads to an underestimation of the determined
risk of death due to leukemia or other causes within this
group and presumably accounts for the observation of only
113

-------
140 deaths as opposed to 187 expected. Among the 140
deaths, 9 were due to all lymphatic and hematopoietic cancer
as opposed to 3.45 based on expectations for U.S. white
males (p < 0.05) and 5.10 expected in the fibrous glass
industry control group (not statistically significant). Of
particular pertinence is that 7 of the 9 deaths in the
benzene-exposed group were of leukemia, as opposed to 1.3 8
and 1.48 expected on the basis of the respective control
groups (p < 0.02). Four of the leukemias were reported as
acute myelocytic, two as monocytic, and one as chronic
myelocytic. The latter case was unusual in that the pa-
tient's age at death was 29 years, relatively young for
chronic myelocytic leukemia, and there was only a 2-year
period between initial exposure and death. In the other six
individuals, this period ranged from 10 to 21 years. As
pointed out above, the description of "monocytic" leukemia
in two cases most likely represents the myelomonocytic
variant of acute myelocytic leukemia, but this is uncertain
in the absence of review of the clinical material. Infante
et al also cite an eighth case who was not part of the
cohort, having started employment in 1950. This person died
at age 28 of myelocytic leukemia 3 years after the initial
exposure.
The authors also use their data to determine the rela-
tive risk for myeloid and monocytic leukemias, as opposed to
114

-------
total leukemia, and conclude that the population studied had
a tenfold increased risk of death from these forms of leuke-
mia. This factor could be an overestimate due to lack of
clinical confirmation of the cause of death. Furthermore,
the authors state that the death data were recalculated on
the basis of a 50.37 percent incidence of myelorronocytic
types of leukemia in this age group. Because chronic
lymphatic leukemia is common in this age group and is a
relatively mild disease in which death commonly occurs from
some other cause, the percentage of deaths caused by myelo-
monocytic types of leukemia may well be much higher than the
50.37 percent value used in the analysis. Counterbalancing
this is the potential underestimation due to the assumption
that 25 percent of the cohort, for whom vital status is
unavailable, is living. The authors also discuss the
presumed dose, which is described below. In general, this
epidemiological study provides excellent confirmatory evi-
dence of the causal relationship of benzene exposure to
acute myelocytic leukemia.
Relationship of Benzene Exposure Level to Leukemogenesis
The medical literature provides little information
concerning the dose of benzene inhaled by individuals who
subsequently developed acute leukemia. In those few reports
where benzene levels are cited, the duration of monitoring
115

-------
has been clearly inadequate for estimation of individual
exposure levels. This is particularly true for assessment
of the leukemogenic effects of benzene as opposed to its
effects in pancytopenia. Concerning the latter, there are
a number of studies in which a large percentage of the work
force has developed benzene hematotoxicity. In these
instances the available monitoring information might rea-
sonably be used to estimate the average exposure. In con-
trast, because leukemia occurs in only a very small per-
centage of the work force exposed to benzene, it is dif-
ficult to be certain whether the occasional individual who
does develop leukemia might have undergone some high ex-
posure, perhaps due to a specific job or to faulty work
habits, that is not typified by an area-wide benzene moni-
toring system.
In the few studies of acute leukemia where benzene
levels are reported, the concentrations generally have been
above 100 ppm. These include the studies of Aksoy et al and
Vigliani et al discussed earlier. In other reports, benzene
102
levels associated with acute leukemia have been 2 20 ppm
164
and 63 to 517 ppm.	There is some indication that the
duration of benzene exposure is longer for acute myelogenous
leukemia than for pancytopenia. A recent tabulation showed
¦ 3,5,7,100 .u - .
far more cases of pancytopenia	than of acute
116

-------
myelogenous leukemia developing after less than 2 years of
exposure. This interpretation is complicated by the ap-
parent lag period between cessation of benzene exposure and
the development of acute leukemia, an interval reported to
49	97
be as long as 15 and 27 years. Accordingly, the ap-
parently longer period of benzene exposure in cases of acute
myelogenous leukemia than in pancytopenia may not be a
function of dose but rather of the biological lag period
between initiation of leukemogenesis and its eventual
clinical appearance. For instance, the two case reports of
77
leukemia tabulated by Goldstein as having less than 2
years exposure to benzene were those of Sellyei and Kelemanl,
in which an 18-month exposure led to pancytopenia, develop-
ment of a Pelger-Huet anomaly, and, 7 years later, acute
myelogenous leukemia. Similarly, the case reported by
102
Kinoshita et al had an onset as aplastic anemia after 6
months of benzene exposure, with subsequent development of
acute leukemia. Also perhaps pertinent are the findings of
Aksoy et al in Turkey. In 1972, these authors reported 32
cases of benzene-associated pancytopenia and only 4 cases of
5 6
acute myelogenous leukemia. ' In 197 6 they reported 20
cases of acute myelogenous leukemia or erythroleukemia,^
whereas the total of pancytopenia cases had increased only
16
to 46 by mid-1977. of the 44 patients with pancytopenia
117

-------
who were available for fol]ow-up, 5 had subsequently de-
veloped leukemia."*"^ Accordingly, it is not certain whether
the generally longer exposure period associated with leu-
kemia as opposed to pancytopenia is a reflection of benzene
dose.
The study of Infante et al" provides the most informa-
tion about dose. As described in detail above, in a deter-
mination of vital status in 1975, these authors noted a
statistically significant increased incidence of leukemia
(total of 7) in a cohort of workers identified as having
been employed any time between 1940 and 1949 in a factory
where benzene was used. The exhaust ventilation was said to
be excellent. Data concerning benzene levels include a
report in 1946 stating that "Tests were made with benzol
detectors and the results indicate that concentrations have
been reduced to a safe level and in most instances range
from 0 to 10 or 15 parts per million" (emphasis added). It
is also stated that 112 surveys conducted between 1963 and
1974 "indicated that employees' benzene exposure was gen-
erally below the recommended concentration in effect at the
time of each survey" (emphasis added). This frequency of
survey is less than once per month in the stated time
period, and thus represents grossly inadequate information
for determining the benzene exposure levels of the seven
118

-------
individuals who developed leukemia. These individuals may
in fact have been exposed to concentrations of benzene well
below the acceptable industrial hygiene limits, which ranged
from 100 ppm maximum allowable concentration in 1941 to 10
ppm as a time-weighted average in 1971. It is also possible
that these individuals were exposed to concentrations well
above the permissible concentrations, which were not de-
tected in the relatively infrequent monitoring surveys.
8 6
This possibility is supported by recent testimony of Harris ,
who cited a survey performed in 1973-74 of the pliofilm
manufacturing plant studied by Infante et al, with the
following observation:
"The spreader and drying units are entered on an
intermittent basis to remove damaged film and rethread
rollers and to repair mechanical failures. It was
reported to the survey team that workers can spend up
to 30 minutes inside these units. Periodic inspections
of shorter duration are also conducted inside the
spreader. Samples inside these units indicate a highly
dangerous level of benzene, ranging from 200 to 350
ppm. Chemical cartridge respirators are required but
are often not worn."
Short-term exposure levels of up to 30 ppm benzene were also
noted in other areas of this plant.
In summary, the available literature concerning the
benzene levels associated with the development of acute
leukemia is inadequate for generation of dose-response
curves or for an analysis of risk related to dose.
119

-------
Possible Mechanisms of Leukemogenesis
The mechanisms by which benzene produces leukemia are
unknown. Of particular interest is whether leukemogenesis
is independent of the pancytopenic effect of benzene or is
directly or indirectly related to overt damage to stem
cells. The former possibility would put benzene into the
category of other carcinogens for which the regulatory
approach is to assume that no safe level exists. Alterna-
tively, if benzene leukemogenesis requires preexisting
damage to hematopoietic tissue, perhaps associated with an
error in repair, then it is possible that a true threshold
for leukemia could be determined. The prerequisite for bone
marrow toxicity might be sufficient damage to cause an overt
decrease in circulating blood cells. If this is true, then
the public could be protected against leukemia by a standard
that precludes such overt toxicity. Furthermore, this would
allow extrapolation, with suitable care, from animal inhala-
tion dose-response studies of benzene hematotoxicity.
It is also possible, however, that leukemia results
from bone marrow damage that is too slight to produce
clinically recognizable cytopenias. This would not nec-
essarily preclude use of an animal model if the parameters
of bone marrow toxicity are sensitive enough. Both Vigliani
120

-------
and Aksoy"1"^' describe patients with apparent benzene-
associated acute leukemia in whom prior study revealed
normal blood counts. The frequency of such blocd counts,
however, was not sufficient to rule out an undetected
pancytopenic phase. Furthermore, it is conceivable that
these were the individuals in the population whcse leukemia
was unrelated to benzene exposure. The available informa-
tion does not clearly support any of the pathways of leuke-
mogenesis discussed here. Obviously an animal model of
benzene leukemogenesis would be useful.
It has been suggested that benzene may act as a cocar-
cinogen, which would in part explain why only certain ben-
zene-exposed individuals develop leukemia. An effect of
benzene as a cocarcinogen or initiator might also explain
cases in which initial hematotoxicity led to cessation of
exposure, followed by a long delay before eventual develop-
. ^ . -| -i - 6,49,82,97,114,148,160,164,191
ment of acute leukemia.	' ' ' '
Evidence on this subject is not firm.
A genetic predisposition has also been invoked to
explain the response to benzene. Erf and Rhoads note pos-
55
sible hematotoxicity in brothers. Aksoy and his col-
leagues present a number of cases of benzene hematotoxicity
developing in families. These include two thalassemic
121

-------
brothers with pancytopenia, two cousins with pancytopenia, a
nephew and paternal uncle with acute lymphoblastic and acute
myeloblastic leukemia, a man with preleukemia whose father
had died of a possible acute leukemia, and two maternal
cousins, one of whom had preleukemia and one an acute
myelomonocytic leukemia. With one exception, a painter,
these persons all worked in the shoe industry. Although
these cases represent suggestive evidence of a genetic
predisposition to benzene hematotoxicity, the degree to
which family groups participate in the shoe industry has not
been specified. If, as in many similar crafts, it is likely
that most of the male family members will follow the same
occupation, then the cases reported by Aksoy et al may not
be greater than what is expected by chance. This would be
true particularly if family groups were large and if the
members tended to work in the same shops with the same
benzene-containing adhesives. Accordingly, a genetic pre-
disposition to benzene toxicity is not yet proven. Genetic
factors also could be operative in determining the rate or
extent to which benzene is metabolized to its presumed
hematotoxic intermediate. Also of interest is an increased
incidence of leukemia in a number of disorders that have in
129
common an increased predisposition to chromosomal damage.
The question of other individual host factors in
suspectibility to benzene has been raised by a number of
122

-------
81
authors. Greenburg et al suggested that obesity pre-
disposes to hematotoxicity, presumably reflecting the solu-
bility of benzene in fat. Although several investigators
suggest that younger individuals and females are more
susceptible to benzene toxicity , ^	this has not been
confirmed in other studies. There is also conflicting
evidence as to whether persons with beta-thalassemia minor,
an inherited disorder of hemoglobin synthesis, may be at
4 65 69 155
increased risk for benzene hematotoxicity. ' ' '	If
greater susceptibility is confirmed, this might indicate
that more rapidly proliferating bone marrow, which is part
of the thalassemia syndromes, increases the risk of hemato-
poietic cell damage due to benzene. Environmental factors,
107
such as high ambient temperatures, may also affect ben-
zene toxicity. In addition, the ingestion or inhalation of
food, drugs, or chemicals might modify the metabolism of
benzene and hence its toxicity. At present the information
is too meager for clear identification of individual host
factors in development of pancytopenia or leukemia resulting
from benzene exposure. Thus, it is unclear whether the
marked variability in individual benzene hematotoxicity
observed in occupational settings primarily Reflects varia-
tions in benzene dose or operation of unknown host factors.
123

-------
In summary, the principal mechanisms by which benzene
could act as a leukemogen are considered to be 1) overt
pancytopenia, 2) inapparent damage to bone marrow, 3) cocar-
cinogenic action, 4) response based on genetic predisposi-
tion, and 5) operation of coincident host factors.
OTHER BENZENE-ASSOCIATED DISORDERS
In addition to pancytopenia and acute myelogenous
leukemia, which are the primary forms of toxicity associated
with benzene exposure, various other hematological disorders
also have been associated with benzene exposure, including
some of the leukemia variants discussed earlier. We now
consider some of the evidence relating benzene exposures to
acute and chronic lymphatic leukemia, other lymphoprolifera-
tive disorders including Hodgkin's disease and multiple
myeloma, chronic myelogenous leukemia, acute monocytic
leukemia, myelofibrosis and myeloid metaplasia, thrombo-
cythemia, paroxysmal nocturnal hemoglobinuria, and various
disorders involving organ systems other than hematopoietic
tissue.
Acute lymphoblastic leukemia is the usual form of
childhood leukemia but is relatively rare in adults, in whom
acute myelogenous leukemia is more common. It is sometimes
difficult to distinguish acute lymphoblastic leukemia from
124

-------
other forms of acute leukemia by standard morphological
techniques. Aksoy et al noted four cases of acute lympho-
blastic leukemia in a group of 34 leukemic individuals."^
The control population consisted of 50 leukemic individuals
with no history of benzene exposure, of whom 2 6 had acute
lymphoblastic leukemia. Goguel et al noted 2 cases of acute
lymphoblastic leukemia in 50 cases of leukemia in benzene-
7 6
exposed individuals. Individual case reports of acute
lymphoblastic leukemia associated with benzene exposure are
,	. , 50,89
also reported.
More substantial evidence is available concerning an
association of benzene exposure with chronic lymphatic
leukemia. As described in detail earlier, McMichael et
127
al noted an association of chronic lymphatic leukemia
with solvent exposure in a study of the cause of death of
more than 6000 men employed in the rubber industry. (They
observed no increase in acute leukemia.) In addition to a
V. t 4-u ¦ A-	1	34, 37,54,68,101,140
number of other individual case reports
and a collection of three cases by Tareef et al in the
173
Soviet Union, Girard and his colleagues in France have
noted a statistically significant increase in history of
benzene exposure in patients with chronic lymphatic leukemia
71 7 4
as compared to control subjects without this disease.
In contrast to these findings, however, Aksoy et al"'"® in
125

-------
187 191
Turkey and Vigliani et al ' in Italy have not reported
any cases of chronic lymphatic leukemia in their relatively
large series. This apparent discrepancy in the observations
of the French investigators and those in Italy and Turkey is
puzzling. Chronic lymphatic leukemia is a disease of the
elderly. Because substantial occupational exposure to
benzene may have been discontinued in France at an earlier
time period, cases of chronic lymphatic leukemia may even-
tually appear in the Italian and Turkish groups having more
recent benzene exposure. Another possible explanation is
that the risk of developing chronic lymphatic leukemia due
to benzene exposure is modified by the presence of other
solvents, which may have differed in France, Italy, and
Turkey. The evidence of a relationship between benzene
exposure and chronic lymphatic leukemia is suggestive but
not conclusive.
35,40	, ¦	8,116 . . ,
Lymphosarcoma,	Hodgkin's disease,	reticulum
139	173 178
cell sarcoma, and multiple myeloma ' also are re-
ported in association with benzene exposure. In none of
these situations have there been enough case reports to
imply other than a chance relation to benzene. Similarly,
acute monocytic leukemia of the Schilling type is rarely
related to benzene exposure.
Various myeloproliferative disorders are reported in
association with benzene. A recent review tabulated 27
126

-------
cases of chronic myelogenous leukemia. Thirteen were from
7 6
a series by Goguel et al evaluating leukemia observed in
benzene-exposed individuals from 1950 to 1965 in the Paris
area. The authors note no clinical differences from cases
of chronic myelogenous leukemia observed in the absence of
benzene exposure. Many of the other case reports are also
36 112
from France. '	In addition, Tareef et al in the Soviet
Union reported five cases of chronic myelogenous leukemia
173
associated with benzene exposure.	One of these indi-
viduals eventually developed acute myeloblastic leukemia, a
not infrequent outcome of chronic myelogenous leukemia.
Myelofibrosis and myeloid metaplasia is a relatively
rare myeloproliferative disorder in which fibrosis of the'
bone marrow is accompanied by extramedullary hematopoiesis,
particularly in the spleen and liver. An increased inci-
dence of this disorder in atom bomb survivors is reported.
Several case reports associate this disorder with benzene
64,123,142,194 „ .	,12
exposure.	Most recently, Aksoy et al
described a 35-year-old shoeworker with evidence of myelo-
fibrosis and myeloid metaplasia including a diagnostic
splenic aspiration. Nine years earlier, the patient had
been diagnosed as having a benzene-induced pancytopenia due
to an 8-year occupational exposure. Although these findings
are suggestive, there are too few case reports to document a
127

-------
causal relationship between benzene and myelofibrosis and
myeloid metaplasia.
g
Aksoy et al recently reported another myeloprolifera-
tive variant, thrombocythemia, as a transient phenomenon in
a shoemaker whose benzene-associated pancytopenia evolved
into a disorder characterized by a high platelet count
3
(540,000 to 1,600,000 per mm ) and then into acute leukemia.
Polycythemia vera, another myeloproliferative disorder,
apparently has not been described in association with ben-
zene exposure.
Paroxysmal nocturnal hemoglobinuria is a hemolytic
disorder noted infrequently in benzene-exposed individu-
als."^ '^ This is an extremely rare condition that has been
linked to both aplastic anemia and acute myelogenous leuke-
mia. The disease is paraneoplastic, being defined by a
population of circulating red cells that are abnormally
sensitive to the hemolytic effect of complement. It often
occurs in the setting of an aplastic anemia, may evolve into
acute myelogenous leukemia or particularly erythroleukemia,
and is sometimes noted as a relatively transient phenomenon
during cases of acute myelogenous leukemia. Accordingly,
its observation in workers exposed to benzene is not un-
expected .
128

-------
In addition to hematopoietic tissue, several other
organ systems are said to be affected by benzene. Effects
on the central nervous system following acute exposure to
benzene have been clearly documented in man. These have
196	39
been reviewed by Gerarde and by Browning, the latter
noting 13 fatal cases in Great Britain following inhalation
of high levels of benzene in enclosed areas. Reported
effects following acute exposure include headache, nausea,
staggering gait, paralysis, convulsions, and eventual uncon-
sciousness and death. Recent reports of fatalities have
174	21
been presented by Tauber and by Bass. Giddiness and
euphoria also have been noted. These acute effects are
usually observed only at relatively high levels of benzene,
well above concentrations believed to be responsible for
hematological effects following chronic exposure. Eastern
European investigators have suggested recently, however,
that effects on the central nervous system, including
changes recorded by electroencephalograph and alterations in
132
cerebral circulation, may occur in occupational settings.
There are also unconfirmed suggestions in the literature
39 133 177
that the human cardiovascular ' ' and gastrointestinal
systems^may be affected by benzene.
129

-------
SUMMARY
Benzene exposure has been clearly demonstrated to
produce hematotoxicity in man. The most commonly reported
effect is a decrease in one or more of the formed elements
of the blood. In more severe cases, this takes the form of
pancytopenia, often with aplastic bone marrow. Evaluation
of occupationally exposed groups reveals a wide spectrum of
disease ranging from fatal aplastic anemia to individual
cytopenias and, in some studies, qualitative abnormalities
of blood cells in the presence of normal peripheral blood
counts. The evidence that benzene is causally related to
pancytopenia includes the observation that benzene is the
common denominator in outbreaks of pancytopenia observed in
many different occupational exposure settings throughout the
world, that detection of pancytopenia in a work force is
often temporally related to the use of benzene, and that
similar effects are observed in animals treated with ben-
zene .
The causal relationship of benzene exposure to leukemia
is more controversial, particularly because an animal model
of this effect has not been clearly demonstrated. Recent
studies, however, have provided strong confirmatory evidence
of a causal relationship, which now appears to be beyond
reasonable doubt for acute myelogenous leukemia and its
130

-------
variants. The evidence includes the many individual case
reports of benzene-induced pancytopenia that proceed to
acute myelogenous leukemia and erythroleukemia, the accu-
mulated case reports of leukemia associated with benzene
exposures, and, most importantly, the epidemiological
evidence indicating a greater risk for leukemia among ben-
zene-exposed individuals.
It would appear that the evidence indicating an in-
creased risk of leukemia on exposure to benzene for various
periods of time and at various concentrations is overwhelm-
ing. Unfortunately, the data are not adequate for deriving
a scientifically valid dose-response curve. Such a curve
may be estimated on the basis of various assumptions; these
assumptions, however, usually represent hypotheses that,
although they may be valid, are not yet proven. Hence the
estimation of a dose-response curve is not appropriate in
this report, which deals with the currently available
scientific knowledge of health effects, but is undertaken in
the benzene Risk Assessment Document.
131

-------
REFERENCES: Benzene Toxicity In Man
1.	Adelstein, A.M. Occupational mortality: cancer.
Ann. Occupat. Hyg. 15:53-57, 1972.
2.	Akman, N., Domanic, N., Muftuoglu, A.U. Paroxysmal
nocturnal hemoglobinuria: report of a case with
features of intravascular consumption coagulopathy.
Hacettepe Bull. Med. Surg. 4:159, 1971.
3.	Aksoy, M., Dincol, K., Akgun, T., Erdem, S., Dincol, G.
Haematological effects of chronic benzene poisoning in
217 workers. Brit. Journal Industrial Med. 28:296-
302, 1971.
4.	Aksoy, M., Erdem, S., Dincol, G. The reaction of
normal and thalassaemic individuals to benzene poisoning:
the diagnostic significance of such studies. IN:
Abnormal Haemoglobins and Thalassaemia Diagnostic
Aspects, Academic Press, Inc., New York, 1975.
5.	Aksoy, M., Dincol, K., Erdem, S., Akgun, T., Dincol, G.
Details of blood changes in 32 patients with pancytopenia
associated with long-term exposure to benzene. Brit.
J. Industr. Med. 29:56-64, 1972.
6.	Aksoy, M., Dincol, K., Erdem, S., Dincol, G. Acute
leukemia due to chronic exposure to benzene. Am.
Journal Med. 52:160-166, 1972.
7.	Aksoy, M., Erdem, S., Akgun, T., Okur, 0., Dincol, K.
Osmotic fragility studies in three patients with a
aplastic anemia due to chronic benzene poisoning. Blut
13:85-90, 1966.
8.	Aksoy, M., Erdem, S., Dincol, K., Hepyuksel, T., Dincol,
G. Chronic exposure to benzene as a possible contributary
etiologic factor in Hodgkin's disease. Blut 38:293-
298, 1974.
9.	Aksoy, M., Erdem, S. Ann. N.Y. Acad. Sci. 165(30)
:15, 1969.
10. Aksoy, M., Erdem, S., Dincol, G. Types of leukemia in
chronic benzene poisoning. A study in thirty-four
patients. Acta Haematologica 55:65-72, 1976.
132

-------
11.	Aksoy, M., Erdem, S., Dincol, G. Leukemia in shoe-
workers exposed chronically to benzene. Blood 44(6):
837-841, 1974.
12.	Aksoy, M., Erdem, S., Dincol, G. Two rare complications
of chronic benzene poisoning: Myeloid metaplasia
and paroxymal nocturnal hemoglobinuria. Report of two
cases. Blut 30:255-260, 1975.
13.	Aksoy, M., Erdem, S., Erdogan, G., Dincol, G. Acute
leukaemia in two generations following chronic exposure
to benzene. Human Heredity 24:70-74, 1974.
14.	Aksoy, M., Erdem, S., Erodogan, G., Dincol, G. Combination
of genetic factors and chronic exposure to benzene in
the aetiology of leukaemia.' Hum. Hered. 26:149-153,
1976.
15.	Aksoy, M., Erdem, S. Some problems of hemoglobin
patterns in different thalassemic syndromes showing the
heterogeneity of beta-thalassemic genes. Annals
N.Y. Acad. Sci., 165:13-24, 1969.
16.	Aksoy, M. Testimony before Occupational Safety and
Health Adminstration, U.S. Dept. of Labor, July 13,
1977.
17.	Andjelkovic, D., Taulbee, J., Symons, M. Mortality
experience of a cohort of rubber workers, 1964-1973.
J. of Occup. Med., 18:387-394, 1976.
18.	Andjelkovich, D., Taulbee, J., Symons, M., Williams, T.
Mortality of rubber workers with reference to work
experience. J. of Occup. Med., 19;397-405, 1977.
19.	Andre, R., Dreyfus, B. Anemie hemolytique grave associee
a un prupura hemorragique thrombopenique. Role probable
due bensol. Transfusion massive. Splenectomie, guerison.
Sangre 22:57-65, 1951.
20.	Appuhn, E., Goldeck, H. Fruh- und spatschacen der
blutbildung durch benzol und seine homologen. Arch.
Gewerbepath u Gewerbehygiene 15:399-428, 1957.
21.	Bass, M. Sudden sniffing death. JAMA 212(12) :2075-
2079, 1970.
133

-------
22
23
24
25
26
27
28
29
30
31
32
33
Bazenova, R.V. The functional condition of: the stomach,
pancreas and liver of workers employed in undertakings
using benzene. Gigiena Truda I Professional'nye Zabolevanija
6 (5 ) : 35-39, 1962"!
Bean, R.H.D. Phenylbutazone and leukaemia: a possible
association. Brit. Med. J. 2:1552, 1960.
Benzene: Uses, Toxic Effects, Substitutes. Occupational
Safety and Health Series. International Labour Office,
Geneva, 1968.
Bernard, J., Basset, A. Results of an inquiry on
benzolism in the aeronautic industry. Sang 17: 120,
1946.
Bernard, J., Braier, L. Les leucoses benzeniques.
Proceedings of the Third International Congress of the
International Society of Hematology, Cambridge, England,
Aug. 21-25, 1950. (New York, Grune & Stratton, pp.
251-263, 1951).
Bernard, J. La lymphocytose benzenique. Sangre 15:
501-505, 1942.
Bickel, L. Ueber beziehungen zwischen akuter aplastischer
anamie, akuter aleukamischer lymphadenose und agranulozytose.
Wien Klin Wochenschr 42:1186, 1929.
Binet, L., Conte, M., Bourliere, F. Intoxication
benzolique mortelle chex une femme vendant des sacs en
cuir synthetique. Presence de benzene dans le sang.
Bull. Mem. Soc. Med. Hop. Paris 61:118, 1945.
Biscaldi, G.P. Acute panmyelophthisis due to benzene.
Description of a case with favourable outcome. Med.
Lavoro 64:363-374, 1973.
Biscaldi, G.P., Robustelli Delia Cuna, G., Pollini, G.
Segni di evoluzione leucemica dell1emopatia benzolica.
Haematologica 54:579-589, 1969.
Blaney, L. Early detection of benzene toxicity. Ind.
Med. 19:227-228, 1950.
Bodey, G.P., Freireich, E.J. Acute leukemia, Chapter
13 IN: Hemotology Principals and Practice. Charles E.
Mengel, Emil Frei, 111, Ralph Nachman, eds. Year Book
Medical Publishers, Inc., Chicago, 1972. pp. 385.
134

-------
34.	Bogetti, B., Vassallo, M. Sul rischio di benzolismo
nelle falegnamerie per I'uso di collanti sintetici.
Lavoro e Medicina 19 (2): 33 — 38, 1965.
35.	Bousser, J'., Neyde, R. , Fabre, A. Un cas d'hemopathie
benzolique tres retardee a type de lymphosarcoma.
Arch. Mai. Prof. 9:130, 1948.
36.	Bousser, J., Tara, S. Apropos de trois cas de leucemie
myeloide chronique provoques par le benzol. Arch.
Mai. Prof. 12:399-404, 1951.
37.	Bowditch, M., Elkins, H.B. Chronic exposure to benzene
(benzol). I. The industrial aspects. The Journal
of Industrial Hygiene and Toxicology 21 (8) :321-330 ,
1939.
38.	Brown, S.M. Letters to the Editor. Leukemia and
potential benzene exposure. J. Occup. Med. 17(1):5-6,
1975.
39.	Browning, E. Benzene. IN: Toxicity and Metabolism of
Industrial Solvents. Elsevier, Publisher. Amsterdam,
1965.
40.	Caprotti, M., Colombi, R., Corsico, R. Ulcers in the
mucous membrane of the colon in benzene poisoning -
Clinical and radiological study. Lavoro Umano 14 (9): 445-
452, 1962.
41.	Chang, Im Won. Study on the threshold limit value of
benzene and early diagnosis of benzene poisoning.
Journal Cath. Med. Coll. 23:429-434, 1972.
42.	Clinical Hematology, Seventh Edition. M,M. Wintrobe,
CT Richard Lee, Dane R. Boggs, Thomas C. Bithell, John
W. Athens, John Foerster, eds. Lea and Febiger,
Philadelphia, PA, pp. 1458, 1974.
43.	Cooke, J.V. The occurrence of leukemia. Blood,
9:340, 1954.
44.	Corisco, R., Biscaldi, G.P., Lalli, M. Benzene-induced
changes in the size of haemopoietic and haematic
cells. Lavoro Umano 19(l):16-27, 1967.
135

-------
45.	Craveri, A. Fibrinolysis, blood platelets, fibrinogen
and other blood-coagulation tests in clinical benzene
poisoning. Medicina del Lavoro 53 (11) : 722-727 ,1962.
46.	Croizat, P., Guichard, A., Revol, L., Creyssel, R.,
meunier. Etude anatomo-clinque et chimique d'un cas
de maladie de Marchiafava-micheli. Sang 19:218, 1948.
47.	Curletto, R., Ciconali, M. Haematological disorders in
benzene poisoning. Medicina del Lavoro 53(8-9): 505-
546, 1962.
48.	Damashek, W. What do aplastic anemia, PNH and hypoplastic
leukemia have in common? Blood 30:251, 1967.
49.	DeGowin, R.L. Benzene exposure and aplastic anemia
followed by leukemia 15 years later. JAMA 185: 748-
751, 1963.
50.	Delore, P., Borgomano. Leucemie aique au cours de
1'intoxication benzenique. Sur l'origine toxique de
certaines leucemies aigues et leurs relations avec les
anemies graves. J. Med. Lyon 9:227-233, 1928.
51.	Di Guglielmo, G., Iannaccone, A. Inhibition of mitosis
and regressive changes of erythroblasts in acute
erythropathy caused by occupational benzene poisoning.
Acta. Haemat. 19:144-147, 1958.
52.	Di Guglielmo, R., Ricci, M. Prima descrizione di un
caso di mielosi eritremica da benzolo nella sua varieta
cronica. Settimana Med. 46:365-368, 1958.
53.	Doskin, T.A. Effect of age on the reaction to a
combination of hydrocarbons. Hygiene and Sanitation
36:379- 384, 1971.
54.	Drouet, P.L., Pierquin, L., Herbeuval, R. Lymphomatose
chronique chez une benzolique. Sang. 18:246-249,
1947.
55.	Erf, L.A., Rhoads, C.P. The hematological effects of
benzene (benzol) poisoning. Journal of Industrial
Hygiene and Toxicology 20 (8) : 421-435, 1939.
56.	Favre-Gilly, M., Bruel, Mile. Syndrome hemorragique
benzolique avec simple alteration morphologique des
plaquettes. De l'utilite d'examen systematique des
plaquettes sur lame chez les ouvriers exposes au
benzol. Arch. Mai. Prof. 91:274-277, 1948.
136

-------
57.	Ferrara, A., Balbo, W. Malattia emolitica in soggetto
esposto a rischio professionale di tipo benzolico.
Riv. Infort. Mai. Prof. 44:713, 1957.
58.	Forni, A.M., Cappellini, A., Pacifico, E., Vigliani,
E.C. Chromosome changes and their evolution in subjects
with past exposure to benzene. Arch. Environ. Health
23:385-391, 1971.
59.	Forni, A., Moreo, L., Chromosome studies in a case of
benzene-induced erythroleukaemia. Europ. Journal
Cancer 5:459-463, 1969.
60.	Forni, A., Moreo, L. Cytogenetic studies 	
in benzene leukaemia. Europ. Journal Cancer 3:251-
255, 1967.
61.	Forni, A., Pacifico, D., Limonta, A. Chromosome
studies in workers exposed to benzene or toluene or
both. Arch. Environ. Health 22:373-378, 1971.
62.	Forni, A., Vigliani, E.C. Chemical leukemogenesis in
man. Ser. Haemat. 7:210-223, 1974.
63.	Galavotti, B., Troisi, F.M. Erythro-leukaemic myelosis
in benzene poisoning. British Journal of Industrial
Medicine 7:79-81, 1950.
64.	Gall, E.A. Benzene poisoning with bizarre extra-
medullary hematopoiesis. Arch. Pathol. 25:315-
326, 1938.
65.	Gaultier, M. , Gervais, P., de Traverse, P.-M., Fournier,
P.-E., Coquelet, N.-L., Loygue, A.-M., Housset, H.
Genetic variations in the hemoblogin caused by the
professional environment. Arch. Mai. Prof. 29: 197-
203, 1970.
66.	Girard, R., Mallein, M.L., Bertholon, J., Coeur, P.,
Tolot, F. Leucocyte alkaline phosphatase and benzene
exposure. Med. Lavoro 61:502-508, 1970.
67.	Girard, R., Mallein, M.L., Bertholon, J., Coeur, P.,
CI. Evreux, J. Etude de la phosphatase alcaline
leucodytaire et du caryotype des ouvriers exposes au
benzene. Arch. Mai. Prof. 31 (1-2): 31-38, 1970.
68.	Girard, R., Mallein, Mme., Fourel, R., Tolot, F.
Lymphose et intoxication benzolique professionnelle
chronique. Arch. Mai. Prof. 27:781-786, 1966.
137

-------
69.	Girard, R., Mallein, M.L., Jouvenceau, A., Tolot, F.,
Revol, L., Bourret, J. Etude de la sensibilite aux
toxiques industriels des porteurs de trait thalassemique.
(32 sujets soumis a surveillance hematologique prolongee)
Le Journal de Medecine de Lyon 48:1113-1126, 1967.
70.	Girard, R., Prost, G., Tolot, F. Comments on indemni-
fication for benzene induced leukemia and aplasia.
Arch. Mai. Prof. 32 (9): 581-583, 1971.
71.	Girard, R., Revol, L. La frequence d'une exposition
benzenique au cours des hemopathies graves. Nouv
Revue Fr. Hemat. 10;477-484, 1970.
72.	Girard, R., Rigaut, P., Bertholon, J., Tolot, F.,
Bourret, J. Les expositions benzeniques meconnues.
Leur recherche systematique au cours des hemopathies
graves. Enquetes chez 200 hemopathiques hospitalises.
Arch. Mai. Prof. 29:723-726, 1968.
73.	Girard, R., Tolot, F., Bourret, J. Malignant hemopathies
and benzene poisoning. Medicina del Lavoro 62:71-
76, 1971.
74.	Girard, R. , Tolot, F., Bourret, J. Hydrocarbures
benzeniques et hemopathies graves. Arch. Mai. Prof.
31 (12) :625-636, 1970.
75.	Goguel, A., Cavigneaux, A., BErnard, J. Benzene
leukaemia. Institut National de la Sante et de la
Recherche Medicale 22 (3): 421-441, 1967.
76.	Goguel, A., Cavigneaux, A., Bernard, J. Benzene
leukemias in the Paris region from 1950 to 1965.
Nouv. Rev. Franc. Hemat. 7:465-480, 1967.
77.	Goldstein, B.D. Hemototoxicity in man, Chapter 7 IN:
Benzene Toxicity, A Critical Evaluation. Sidney
Laskin and Bernard Goldstein, eds. In Press, 1977.
78.	Goldwater, L.J. Disturbances in the blood following
exposure to benzol. J. Lab. Clin. Med. 26:957-973,
1941.	•
79.	Goldwater, L.J., Tewksbury, M.P. Recovery following
exposure to benzene (benzol). J. Indust. Hyg. 23:217-
231, 1941.
138

-------
80.
81
82
83
84
85
86
87
88
89
90
91
Gorini, P., Colombi, R., Pecorari, D. Investigation on
the origin of the macrocytosis in the cytoplastic
anemia from chronic benzene poisoning. Lav. Umano
11:121-133, 1959.
Benzene poisoning in rotogravure printing. Journal
of Industrial Hygiene and Toxicology 21 (8): 395-
420, 1939.
Guasch, J., Pelayo, E., Vallespin, J. Anemia benzolica
complicada con leucemia aguda a los seis anos de
evolucion. Sangre 4:129-146, 1959.
Guberan, E., Kocher, P. Pronostic lointain de 11 intoxication
benzolique chronique: controle d'une population 10 ans
apres 11 exposition. Schweiz. Med. Wachr. 101: 1789-
1790, 1971.
Hamilton, A. General Review. Benzene (benzol) poisoning.
Arch. Path. 11:434-454, 1931.
Hamilton-Paterson, J.L., Browning, E. Toxic effects in
women exposed to industrial rubber solutions. Brit. Med. J.
1:349-352, 1944.
Harris, R.L. Testimony before occupational Safety and
Health Administration. U.S. Dept. of Labor, August 8,
1977.
Health Effects of Benzene - A Review. The National
Research Council, Committee on Toxicology, National
Academy of Sciences, Washington, D.C. 1976.
Helmer, K.J. Accumulated cases of chronic benzene
poisoning in the rubber industry. Acta Medica Scand.
118: 354-375, 1944.
Hernberg, S., Savilahti, M., Ahlman, K., Asp, S.
Prognostic aspects of benzene poisoning. Brit. Journal
Industr. Med. 23:204-209, 1966.
Hunter, F.T. Chronic exposure to benzene (benzol).
II. The clinical effects. Journal of Industrial
Hygiene and Toxicology 21(8):331-354, 1939.
Hunter, D. Industrial toxicology. Quarterly Journal
of Medicine pp. 185-250, 1944.
139

-------
92.	Hutchings, M., Drescher, S., McGovern, F.B., Coombs,
F.A. Investigation of benzol and toluol poisoning in
Royal Australian Air Force workshops. The Medical Journal
of Australia 2 (23):681-693, 1947.
93.	Infante, P.F., Rinsky, R.A., Wagoner, J.K., Young, R.J.
Leukaemia in benzene workers. Lancet, 2:76- 78, 1977.
94.	Ishimaru, T., Okada, H., Tomiyasu, T.f Tsuchimoto, T.,
Hoschino, T., Ichimaru, M. Occupational factors in the
epidemiology of leukemia in Hiroshima and Nagasaki.
Amer. J. Epidem. 93:157-165, 1971.
95.	Ito, T. Study on the sex difference in benzene poisoning.
Report 1. On the obstacles in benzene workers.
Showa Igakukai Zasshi 22:268-272, 1962.
96.	Inceman, S., Tangun, Y. Impaired platelet-collangen
reaction in a case of acute myeloblastic leukemia due
to chronic benzene intoxication. Turk. Tip. Cemig.
Mecm. 35;417-424, 1969.
97.	Justin-Besancon, L., Pequignot, H., Barillon, A. Leucose
aigue survenue 27 ans apres exposition aux vapeurs
benzoliques. La Semaine des Hopitaux 35:186-188,
1959.
98.	Kahn, H., Muzyka, V. The chronic effect of benzene on
prophyrin metabolism. Work-Environm. Health 10-
140-143, 1973.
99.	Kahn, KH.A., Muzyka, V.I. The effect of benzene on the
A-aminolevulic acid and porphyrin content in the
cerebral cortex and in the blood. Industrial Hygiene
and Profession Associated Disorders 3:59-60, 1970.
100.	Kauppila, 0., Setala, A. Chronic benzene poisoning.
Report of five cases. Annuals Medicinae Internae Fenniae
45:49-51, 1956.
101.	Kiec, E., Kunski, H. A case of chronic lymphatic
leukemia due to prolonged exposure to benzene. Med. Pracy
16;362-365, 1965.
102.	Kinoshita, Y., Terada, H., Saito, H., et al. A case of
myelogenous leukemia. Journal of Japan Haematological
Society pp. 85-96, 1965.
140

-------
103.	Kliche, N., Meubrink, H., Wahl, F. Chronische benzolschaden
bei dachdeckern. Z. Ges Hyg. 15:310-316, 1969.
104.	Kohli, P., Brunner, H.E., Siegenthaler, W. Erythroleukamie
nach chronischer benzolintoxikation. Untersuchung der
ferro- und erythrocytenkinetik mit radioaktivem eisen
und chrom. Schweiz. Med. Wschr. 97:368-373, 1967.
105.	Kolesar, D., Ballog, 0. Studies by fluorescent microscopy
of the effect of occupational benzene exposure on
leucocyte nuclei changes. Bratislavske Lekarske Listy
4511 (4):212-219, 1965.
106.	Koslova, T.A., Volkova, A.P. Blood picture and phagocytic
activity of leucocytes in workers having contact with
benzene. Gig. Sanit. 25:29-34, 1960.
107.	Koslova, T.A. The effect of benzene on the organism at
high air temperature. Abstract. Air Pollution Translations,
Vol. 1, F-7 308, AP-56, HEW, 1969.
108.	Kuhlraann, D., Von. Mitteilung einer todlichen benzolver-
giftung aus der schuhindustrie. Zbl. Arbeitsschutz
9:62-64, 1959.
109.	Lange, A., Smolik, R., Zatonski, W. , Szymanska, J.
Serum innumoglobulin levels in workers exposed to
benzene, toluene and xylene. Int. Arch. Arbeitsmed.
31:37-44, 1973.
110.	Lange, A., Smolik, R., Zatonski, W., Glazman, H.
Leukocyte agglutinins in workers exposed to benzene,
toluene and xylene. Int. Arch. Arbeitsmed. 31:45-50,
1973.
Ill	mortality
among chemists. J. Nat. Cancer Inst., 43:1159-1164,
1969.
112.	Liaudet, J., Combaz, M. Leucemie myeloide chronique
chez un chimiste petrolier de 35 ans, manipulant du
benzene depuis 1'age de 18 ans. Journal Europ. Toxicol.
6 (6) :309-313, 1973.
113.	Lob, M. L'action du benzene sur les thrombocytes et
sur certaines activities enzymatiques. Archives des
Maladies Professionnelles 24 (4-5):371-374, 1963.
141

-------
114. Ludwig, Von H., Werthemann, A. Benzol-Mvelopathien.
Schweiz Med. Wschr. 92:378-384, 1962.
115.	Mallein, M.L., Bryon, P.A., Fiere, D. , Girard, R.
Cryptoleucose aigue ou "anemie refractaire" benzenique.
(Deux nouveaux cas). Arch. Mai. Prof. 32:577-579,
1971.
116.	Mallory, T.B., Gall, E.A., Brickley, W.J. Chronic
exposure to benzene (benzol). III. The pathologic
results. Journal of Industrial Hygiene and Toxicology
21 (8):355-377, 1939.
117.	Mancuso, T.F., Brennan, J.J. Epidemiological considerations
of cancer of the gallbladder, bile ducts and salivary
glands in the rubber industry. J. Occup. Med., 12(9) :
333-341, 1970.
118.	Mancuso, T.F., Ciocco, A., El-Attar, A.A. An epidemiological
approach to the rubber industry. J. Occup. Med.,
10(5):213-232, 1968.
119.	Marchal, G., Duhamel, G. L1anemie hemolytique dans les
leucemies. Sangre 21:254-261, 1950.
120.	Marchand, M.M. Un cas mortel de leucoso benzolique.
Arch. Mai. Prof. 21:576-477, 1960.
121.	Maugeri, S., Colombi, R., Pollini, G., Strosselli, E.,
Corsico, R. Evolution and characteristics of benzol
blood dyscrasia. Med. Lavoro 56:544-560, 1965.
122.	Mazzella Di Bosco, M. Considerations on some cases of
benzene leucosis in shoe-factory workers. Lavoro Umano
16 (3): 105-121, 1964 .
123.	McLean, J.A. Blood dyscrasia after contact with petrol
containing benzol. The Medical Journal of Australia
47 (II) :845-849, 1960.
124.	McMichael, A.J., Andjelkovic, D.A., Tyroler, H.A.
Cancer mortality amond rubber workers: an epidemiological
study. Annals N.Y. Aca. Sci., 271:125-137, 1976.
125.	McMichael, A.J., Spirtas, R., Gamble, J.F., Tousey,
P.M. Mortality among rubber workers: relationship to
specific jobs. J. of Occup. Med., 18:178-185, 1976.
126.	McMichael, A.J., Spirtas, R., Kupper, L.L. An Epidemiologic
study of mortality within a cohort of rubber workers.
J. of Occup. Med. 15:458-464, 1974.
142

-------
127.	McMichael, A.J., Spirtas, R., Kupper, L.L., Gamble,
J.F. Solvent exposure and leukemia among rubber workers:
an epidemiologic study. J. Occup. Med. 17(4):234-
239, 1975.
128.	Merklen, Pr., Israel, L. Intoxication par le benzol:
aleucie hemorragique, autrement dit "hypoleucie hemorragique
avec anemie". Sangre 8:700-709, 1934.
129.	Miller, R.W. Etiology of childhood leukemia. Pediat.
Clin. North America 13:267, 1966.
130.	Mina, L.R. Changes in the carbonic anhydrase of the
blood and the endophenoloxydases of the leucocytes in
chronic benzene and monochlorobenzene poisoning.
Igiena 8(3):231-240, 1959.
131.	Modan, B., Segal, S., Shani, M., Sheba, C. Aplastic
anemia in Israel: evaluation of the etiological role
of chloramphenicol on a community-wide basis. Am. J.
of the Med. Sci., 270 (3) :441-445, 1975.
132.	Monaenkova, A.m., Snegova, G.V. Cerebral circulation
in some occupational poisonings: lead, carbon disulfide,
benzene. Gig. Tr. Prof. Zabol. 17 (4):33 — 37, 1973.
133.	Monaenkova, A.M., Zorina, L.A. Hemodynamics and heart
muscle changes in chronic benzene poisoning. Gig.
Truda Prof. Zabolevaniya 4:30-34, 1975.
134.	Monson, R.R., Nakano, K.K. Mortality among rubber
workers. I. White male union employees in Akron, Ohio.
Am. J. of Epidemiology 103(3):297-303, 1976.
135.	Monson, R.R., Nakano, K.K. Mortality among rubber
workers. II. Other employees. Am. J. of Epidemiology,
103 (3) : 297-303, 1976.
136.	Monteverde, A., Grazioli, C., Fumagalli, E. The effect
of fibrinogen and platelets on the thromboelastogram in
benzene poisoning. Medicina del Lavoro 54(2):95-102,
1963.
137.	Nissen, N.I., Ohlsen, A.S. Eryuromyelosis. Oversigt
og et tilfaelde hos en benzolarbejder. Ugeskr. Laeg.
114:737-742, 1952.
138.	Pagnotto, L.D., Elkins, H.B., Brugsch, H.G., Walkley,
E.J. Industrial benzene exposure from petroleum naphtha--
I. Rubber coating industry. Am. Ind. Hyg. Assoc. J.,
22:417-21, 1961.
143

-------
139.	Paterni, L., Sarnari, V. Involutional myelopathy due
to benzene poisoning, with the appearance1, of a mediastinal
reticulosarcoma at an advanced stage. Securitas
50 (10) :55-59, 1965.
140.	Poinso, MM.R., Bondil, J.-N., Ruf. G. Leucose lymphoide
benzenique. Bull Soc. Med. Hosp. Paris 70:13-16,
1954.
141.	Pollini, G., Colombi, R. Changes in the osmotic
resistance of the leucocytes in persons exposed to
benzene. Lavoro Umano 16(4):177-184, 1964.
142.	Rawson, R., Parker, F., Jackson, H. Industrial solvents
as possible etiologic agents in myeloid metaplasia.
Science 6:541-542, 1941.
143.	Recommended Standard for Occupational Exposure to
Benzene. U.S. Department of Health, Education, and
Welfare. NIOSH 74-137, 1974.
144.	Redmond, C.K., Ciocco, A., Lloyd, J.W., Rush, H.W.
Long-term mortality study of steelworkers. VI. Mortality
from malignant neoplasms among coke oven workers.
J. Occup. Med., 14:621-629, 1972.
145.	Redmond, C.K., Strobino, B.R., Cypress, R.H. Cancer
experience among coke by-product workers. Annals
N.Y. Acad. Sci., 271:102-115, 1976.
146.	Review of the Health Effects of Benzene. By the
Committee on Toxicology of the National Academy of
Sciences. National Research Council, Washington, D.C.
December, 1975.
147.	Revnova, N.V. Concerning auto-immunity shifts in
chronic occupational benzol poisoning. Gig. Tr. Prof. Zabol.
7:38-42, 1962.
148.	Robustellie della Cuna, G., Favino, A., Biscaldi, G.P.,
Pollini, G. Trasformazione in leucemia acuta di un
casodi mielopatia involutiva benzolica. Haematologica
57:65-89, 1972.
149.	Roth, L., Dinu, I.V., Turcanu, P., Moise, G. Cytologic
and immunochemical features of benzene induced reticuloses.
Timisoara Med. 17:29-38, 1972.
144

-------
150.
151.
152
153
154
155
156
157
158
159
160
161
162
Roth, L., Turcanu, P., Dinu, I., Moise, G. Monocytosis
in those who work with benzene and chronic benzene
poisoning. Folia Haematol. 100:213-224, 1973.
Roth, L., Turcanu, P., Servan, V. , Moise, G. Qualitative
veranderungen der lymphozyten nach benzolabsorption.
Zschr. Inn. Med. 20:932-935, 1970.
Rundles, R.W. Chronic granulocytic leukemia, Chapter
80 IN: Hemotology. McGraw-Hill Book Co., New York,
1972. pp. 681.
Saite, G., Dompe, M. Sul rischil benzolico nei principali
stabilimenti rotocalcograifci di Milano. Med. Lavoro
38:269-285, 1947.
Saita, G. Mielosi aplastica e successiva mielosi
leucemica leucopenica, provocate da benzolo. Med. Lavoro
36:143-159, 1945.
Saita, G., Moreo, L. Talassemia ed emopatie professional!.
Nota I - Talassemia e benzolismo cronico. Medicina
del Lavoro 50(l):25-44, 1959.
Saita, G., Moreo, L. Haemolytic attack following
inhalation of a single massive doze of benzol.
Medicina del Lavoro 52 (11):713-716 , 1961.
Saita, G., Moreo, L. A case of chronic benzene poisoning
with a Pelger-Huet type leucocyte anomaly. Medicina del
Lavoro 57 (5):331-335, 1966.
Saita, G., Sbertoli, C., Farina, G.F. Thromboelastographic
investigations in benzene haemopathy. Medicina del Lavoro
55 (11) :655-664, 1964.
Saita, G., Sbertoli, G. L1agglutinogramma nell'intossicazioue
cronica da benzolo. Med. Lavoro 45:250-253, 1962.
Saita, G., Vigliani, E.C. Th action of benzene in
inducing leukemia. Med. Lavoro 53 (10) :581-586, 1962.
Savilahti, M. Over 100 cases of benzene poisoning in a
shoe factory. Archiv fur Gewerbcpathologie und
Gewerbehygiene 15:147-157, 1956 /
Selling, L. A preliminary report of some cases of
purpura haemorrhagica due to benzol poisoning.
Bull. John Hopkins Hosp. 21:33-37, 1910.
145

-------
163.	Selling, L., Osgood, E.E. Chronic benzene poisoning.
International Clinics 3:52-63, 1935.
164.	Sellyei, M., Kelemen, E. Chromosome study in a case of
granulocytic leukaemia with "Pelgerisation" 7 years
after benzene pancy topenia. Europ. J. Cancer 7:83-
85, 1971.
165.	Smolik, 	
Rappaport, H., Toth, B., Raha, C.R., Tomatis, L. ,
Feldman, R., Ramahi, H. Studies on the toxicity
of petroleum waxes. Toxicol. & Applied Pharmacol.
4:1-62, 1962.
166.	Smolik, R., Grzybek-Hryncewicz, K., Lange, A., Zatonski,
W. Serum complement level in workers exposed to benzene,
toluene and zylene. Int. Arch. Arbeitsmed. 31:243-
247, 1973.
167.	Sobczyk, W. , Siedlecka, B., Gajewska, Z. EEG recordings
in workers exposed to benzene compounds. Med. Pracy
24:273-283, 1973.
168.	Solov'eva, E.A. Morphological changes of blood platelets
and of megakarycocytes in chronic benzene poisoning.
Gig. Tr. Prof. Zabol. 7(ll):57-60, 1963.
169.	Sroczynski, J., Kossmann, S., Wegiel, A. Coagulation
system in experimental poisoning with benzene vapors.
Bui. Slezby Sanit. Epidemiol Wojewodztwa Katowickiego
15:131-134, 1971.
170.	Snyder, R., Kocsis, J.J. Current concepts of chronic
benzene toxicity. CRC Critical Reviews in Toxicology
Pages 265-288, June, 1975.
171.	Snyder, R. Relation of benzene metabolism to benzene
toxicity. IN: symposium on Toxicology of Benzene and
Alkylbenzenes, Braun, D., Editor. Industrial Health
Foundation, Pittsburgh, pp. 44-53, 1974.
172.	Stewart, R.D., Dodd, H.C., Baretta, E.D., Schaeffer,
A.W., Mutchler, J.E. Chronic overexposure to benzene
vapor. Toxicol. Applied Pharmacol. i0:381, 1967.
(Abstract)
173.	Tareeff, E.M., Kontchalovskaya, N.M., Zorina, L.A.
Benzene leukemias. Acta. Unio. Int. Contra. Can Crum
19: 751-755, 1963.	"
146

-------
174.	Tauber, J. Instant benzol death. Journal Occup. Med.
12:91-92, 1970.
175.	Thorpe, J.J. Author response. J. Occup. Med., 17:6,
1975.
176.	Thorpe, J.J. Epidemiologic survey of leukemia in
persons potentially exposed to benzene. Journal
of Occupational Medicine 16 (6):375-382 , 1974.
177.	Thienes, C.H., Haley, T.J. Cardiac Poisons. IN;
Clinical Toxicology, Chapter 5, 5th edition, Lee and
Febrger, Philadelphia, 1972.
178.	Torres, A., Giralt, M., Raichs, A. Coexistencia de
antecadentex benzolicos cronicos y plasmocitoma multiple.
Presentacion de dos casos. Sangre 15:275-279, 1970.
179.	Tough, I.M., Smith, P.G., Court Brown, W.M., Harnden,
D.G. Chromosome studies on workers exposed to atmospheric
benzene. Europ. Journal Cancer 6:49-55, 1970.
180.	Tyroler, H.A. Testimony before occupational Safety and
Health Administration. U.S. Dept. of Labor, August 8,
1977.
181.	Tzanck, A., Dreyfuss, A., Jais, M. Hemopathie postbenzolique
at leucoblastose medullaire. Sangre 11:550-557, 1937.
182.	Undritz, E. Un cas d'intoxication professionnelle par
des vapeurs de benzine, collophane et huile de lin.
Le Progres Medical 69(16):569-570, 1938.
183.	U.S. Dept. of HEW. Mortality	in 1950 by occupation and
industry. Vital Statistics -	Special Reports. 53(1—
5), 1961-1963.
184.	Vannucchi, V. Mielosi acute da benzolismo cronico con
aspetto iniziale di aplasia e	aspetto successivo di
leucemia o di eritroleucemia.	Riv. Crit. Clin. Med.
57:51, 1957.		
185.	Van Ravesteyn, A.H. Chronische benzolvergiftiging en
leucaemie. Nederl. Tijdschr. Geneesk. 85:4 038-4044,
1941.
186.	Van Schoonhoven Van Beurden, A.J.R.E. Benzolleucaemie.
Nederl. Tijdschr. Geneesk. 93:2584-2592, 1949.
147

-------
187.	Vigliani, E.C. Leukemia associated with benzene poisoning.
Presented at: Conference on Occupational Carcinogenesis,
Abstract 38, March 24-27, 1975.
188.	Vigliani, E.C., Forni, A. Benzene and leukemia.
Env. Res. 11:122-127, 1976.
189.	Vigliani, E.C., Forni, A. Benzene, Chromosome changes
and leukemia. Letter to the Editor, Journal Occup. Med.
11:148-149, 1969.
190.	Vigliani, E.C., Forni, A. Leucemogenesi professionale.
Minerva Med. 57:3952-3955, 1966.
191.	Vigliani, E.C., Saita, G. Benzene and leukemia. The
New England Journal of Medicine 271(17):872-876, 1964.
192.	Vigliani, E.C., Saita, G. Leucemia emocitoblastica da
benzolo. Med. Lavoro 34:182-191, 1943.
193.	Wilson, R.H. Benzene poisoning in industry. Journal Lab.
Clin. Med. 27:1517-1521, 1942.
194.	Wurster-Hill, D.H., Cornwell, G.G., Mclntyre, O.R.
Chromosomal aberrations and neoplasm - a family study.
Cancer 33(1):72-81, 1974.
195.	Zini, C., Alessandri, M. Anomalia leucocitaria pseudo-.
pelgeriana in -un caso di emopatia benzolica con leucosi
acuta terminale. Haematologica 52:258-266, 1967.
196.	Gerarde, H.W. Toxicology and Biochemistry of Aromatic
Hydrocarbons. pp. 97-108, Elsevier Publishing Company,
New York, 1960.
197.	Kraybill, H.F. A survey of NCI viewpoints on carcinogenic
risk of benzene exposure. National Cancer Institute.
148

-------