EPA
TOXICOLOGICAL REVIEW
of
CUMENE
(CAS No. 98-82-8)
In Support of Summary Information on the
Integrated Risk Information System
June 1997
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, DC
-------
TABLE OF CONTENTS
Author and Reviewers iii
Foreword v
1.0 Introduction 1
2.0 Chemical and Physical Information Relevant to Assessments 2
3.0 Toxicokinetics Relevant to Assessments 2
4.0 Hazard Identification 3
4.1 Studies in Humans 3
4.2 Prechronic and Chronic Studies and Cancer Bioassays in Animals 4
4.3 Reproductive/Developmental Studies 8
4.4 Other Studies 10
4.5 Synthesis and Evaluation of Major Noncancer Effects and Mode of Action 11
4.6 Weight of Evidence Evaluation and Cancer Classification 13
4.7 Other Hazard Identification Issues 14
4.7.1 Possible Childhood Susceptibility 14
4.7.2 Possible Gender Differences 14
5.0 Dose-Response Assessments 14
5.1 Oral Reference Dose 14
5.1.1 Choice of Principal Study and Critical Effect 14
5.1.2 Methods of Analysis 15
5.1.3 Oral Reference Dose Derivation 15
5.2 Inhalation Reference Concentration 16
5.2.1 Choice of Principal Study and Critical Effect 16
5.2.2 Methods of Analysis 16
5.2.3 Inhalation Reference Concentration Derivation 17
5.3 Cancer Assessment 18
6.0 Major Conclusions in Characterization of Hazard Identification and
Dose-Response Assessments 18
6.1 Hazard Identification 18
6.2 Dose Response 19
7.0 References 20
8.0 Appendixes 25
Appendix A: Benchmark Concentration Analyses of Data from
Cushman et al. (1995) 25
Appendix B: Summary of and Response to External Peer Review Comments 26
-------
AUTHOR AND REVIEWERS
Chemical Manager/Author
Gary L. Foureman, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
Reviewers
This document and summary information on IRIS have received peer review both by EPA
scientists and by independent scientists external to EPA (U.S. EPA, 1994a). Subsequent to
external review and incorporation of comments, this assessment has undergone an Agency-wide
review process whereby the IRIS Program Manager has achieved a consensus approval among
the Office of Air and Radiation; Office of Policy, Planning, and Evaluation; Office of Prevention,
Pesticides, and Toxic Substances; Office of Research and Development; Office of Solid Waste
and Emergency Response; Office of Water; and the Regional Offices.
Internal EPA Reviewers
Larry D. Anderson, Ph.D.
Toxicologist
Office of Prevention, Pesticides, and Toxic Substances
U.S. Environmental Protection Agency
Washington, DC
John W. Allis, Ph.D.
Health Scientist
National Health and Environmental Effects Research Laboratory
Office of Research Development
U.S. Environmental Protection Agency
Research Triangle Park, NC
Mary C. Henry, Ph.D.
Pharmacologist
Office of Prevention, Pesticides, and Toxic Substances
U.S. Environmental Protection Agency
Washington, DC
in
-------
Samuel Rotenberg
Hazardous Waste Management Division
U.S. Environmental Protection Agency, Region III
Philadelphia, PA
Vanessa Vu
Director, Risk Assessment Division
Office of Prevention, Pesticides, and Toxic Substances
U.S. Environmental Protection Agency
Washington, DC
External Peer Reviewers
Richard H. Bruner, DVM
Division Director
Pathology Associates International
P.O. Box 26, 3900 NCTR Road
Jefferson, AR 72079
Ronald D. Hood, Ph.D.
Professor and Interim Department Chair
Department of Biological Sciences
University of Alabama
Tuscaloosa, AL 35487-0344
Norbert P. Page, Ph.D.
Page Associates
17601 Stoneridge Court
Gaithersburg, MD 20878
Summaries of the external peer reviewers' comments and the disposition of their
recommendations are in Appendix B.
IV
-------
FOREWORD
The purpose of this review is to provide scientific support and rationale for the hazard
identification and dose-response assessments for both cancer and noncancer effects (the oral
reference dose and the inhalation reference concentration) from chronic exposure to cumene.
It is not intended to be a comprehensive treatise on the chemical or toxicological nature of
cumene.
In Section 6, EPA has characterized its overall confidence in the quantitative and
qualitative aspects of hazard and dose-response (U.S. EPA, 1995a). Matters considered in this
characterization include knowledge gaps, uncertainties, quality of data, and scientific
controversies. This characterization is presented in an effort to make apparent the limitations of
the individual assessments and to aid and guide the risk assessor in the ensuing steps of the risk
assessment process. For other general information about this assessment or other questions
relating to the Integrated Risk Information System (IRIS), the reader is referred to EPA's Risk
Information Hotline at (513)569-7254.
-------
1.0 INTRODUCTION
This document presents the derivation of the noncancer dose-response assessments for oral
exposure (the oral reference dose or RfD) and for inhalation exposure (the inhalation reference
concentration or RfC) and the cancer hazard and dose-response assessments.
The RfD and RfC are meant to provide information on long-term toxic effects other than
carcinogenicity. The RfD is based on the assumption that thresholds exist for certain toxic
effects such as cellular necrosis but may not exist for other toxic effects, such as some
carcinogenic responses. The RfD is expressed in units of milligrams per kilogram per day.
In general, the RfD is an estimate (with uncertainty spanning perhaps an order of magnitude) of a
daily exposure to the human population (including sensitive subgroups) that is likely to be
without an appreciable risk of deleterious effects during a lifetime. The inhalation RfC is
analogous to the oral RfD. The inhalation RfC considers toxic effects for both the respiratory
system (portal-of-entry) and for effects peripheral to the respiratory system (extrarespiratory or
systemic effects). It is expressed in units of mg/m3.
The carcinogenicity assessment is meant to provide information on three aspects of the
carcinogenic risk assessment for the agent in question: (1) the U.S. Environmental Protection
Agency (EPA) classification and (2) quantitative estimates of risk from oral exposure and
(3) inhalation exposure. The classification reflects a weight-of-evidence judgment of the
likelihood that the agent is a human carcinogen and the conditions under which the carcinogenic
effects may be expressed. Quantitative risk estimates are presented in three ways. The slope
factor is the result of application of a low-dose extrapolation procedure and is presented as the
risk per mg/kg/day. The unit risk is the quantitative estimate in terms of either risk per |ig/L of
drinking water or risk per |ig/m3 of air breathed. The third form in which risk is presented is
drinking water or air concentration providing cancer risks of 1 in 10,000, 1 in 100,000, or 1 in
1,000,000.
Development of these hazard identifications and dose-response assessments for cumene has
followed the general guidelines for risk assessments as set forth by the National Research
Council (1983). Other EPA guidelines that were used in the development of this assessment
include Risk Assessment Guidelines of 1986 (U.S. EPA, 1987a), (proposed) Guidelines for
Carcinogen Risk Assessment, 1996 (U.S. EPA, 1996a), Guidelines for Developmental Toxicity
Risk Assessment (U.S. EPA, 1991c), (proposed) Interim Policy for Particle Size and Limit
Concentration Issues in Inhalation Toxicity (U.S. EPA, 1994b), (proposed) Guidelines for
Neurotoxicity Risk Assessment (U.S. EPA, 1995b), Methods for Derivation of Inhalation
Reference Concentrations and Application of Inhalation Dosimetry (U.S. EPA, 1994c),
Guidelines for Reproductive Toxicity Risk Assessment (U.S. EPA, 1996b) Recommendations
for and Documentation of Biological Values for Use in Risk Assessment (U.S. EPA, 1988), and
Use of the Benchmark Dose Approach in Health Risk Assessment (U.S. EPA, 1995c).
Literature search strategy employed for this compound were based on the Chemical
Abstract Service Registry Number (CASRN) and at least one common name. As a minimum, the
following databases were searched: RTECS, HSDB, TSCATS, CCRIS, GENETOX, EMIC,
-------
EMICBACK, DART, ETICBACK, TOXLINE, CANCERLINE, and MEDLINE and MEDLINE
backfiles.
Any pertinent information submitted by the public to the Integrated Risk Information
System (IRIS) submission desk also was considered in the development of this document.
2.0 CHEMICAL AND PHYSICAL INFORMATION RELEVANT TO ASSESSMENTS
Cumene also is known as isopropyl benzene, (l-methylethyl)-benzene, and
2-phenylpropane. Some relevant physical and chemical properties of cumene are listed below
(Mackay and Shui, 1981; Hansch and Leo, 1985).
CASRN: 98-82-8
Empirical formula: C9H12
Molecular weight: 120.2
Vapor pressure: 4.6 mm Hg at 25 °C
Water solubility: 50 mg/L at 25 °C
LogKow: 3.66
Conversion factor: 1 ppm = 4.9 mg/m3, 1.0 mg/m3 = 0.2 ppm
Points to be made from these properties include that, at room temperature, cumene is a
volatile liquid, that airborne concentrations of over 6,000 ppm (29,400 mg/m3) may be attained,
and that cumene is nearly insoluble in water. Structurally, cumene is a member of the alkyl
aromatic family of hydrocarbons, which also includes toluene (methylbenzene) and ethylbenzene.
3.0 TOXICOKINETICS RELEVANT TO ASSESSMENTS
Inhalation tests conducted in humans indicate that cumene is absorbed readily via the
inhalation route, that it is metabolized efficiently to water soluble metabolites within the body,
and that these metabolites are excreted efficiently into the urine with no evidence of long-term
retention within the body. These results concur with the results of animals studies. The
combined findings indicate that neither cumene nor its metabolites are likely to accumulate
within the body.
Human volunteers (five men and five women) were exposed head-only for 8-h periods to
cumene vapors (Seiiczuk and Litewka, 1976). Every 10 days, each subject was exposed to one of
three different concentrations of cumene, 240, 480, or 720 mg/m3. Samples of exhaled air were
collected (method not clear from text) during exposures for estimation of respiratory tract
retention, and urine was collected from each subject during exposure and for 40 h thereafter.
The mean respiratory tract retention was reported to be 50% (range, 45 to 64%), even at the
highest concentration, although no data is given to support derivation of these values. Excretion
of cumene, estimated from urinary amounts of 2-phenyl-2-propanol, was maximal after 6 to 8 h
of exposure and approached zero at 40 h postexposure. The plot of time against urinary
excretion of this metabolite revealed a rapid early phase (t ,/2 ~ 2 h) and a slower later phase
-------
(t,/2 ~ 10 h). Approximately 35% of the calculated absorbed dose was excreted as 2-phenyl-
2-propanol during the 8 h of exposure and 40 h postexposure.
Groups of Fischer 344 rats (minimum 4/sex/group) were studied after being exposed to
radiolabeled cumene (>98% purity) either by single intravenous dose (35 mg/kg); single oral
gavage doses (33 or 1,350 mg/kg); single 6-h nose-only inhalation (100, 500, or 1,500 ppm); or
eight daily oral gavage doses (33 mg/kg), with the eighth dose being radiolabeled (Research
Triangle Institute, 1989). The inhalation studies indicated rapid absorption, with detectable
levels of cumene appearing in the blood within 5 min of the beginning of exposure. The gavage
studies showed that cumene was absorbed readily via this route, with maximum blood levels
occurring at the earliest time point sampled (4 h) for the lower dose and at 8 to 16 h for the
higher dose. Elimination of cumene from the blood appeared as monoexponential with a half-
life calculated between 9 to 16 h for the gavage doses. The pattern of cumene disappearance
from the blood in the inhalation studies also appeared to be monoexponential with the half-lives
increasing with dose, from 3.9 h at 100 ppm, to 4.6 h at 500 ppm, to 6.6 h at 1,200 ppm.
Analysis of tissues (presumably immediately after exposure) indicated that several tissues,
including adipose, liver, and kidney, all had elevated tissue/blood ratios of cumene, regardless of
the route of cumene administration, indicating thorough distribution of cumene throughout the
body independent of administration route. In general, very similar rates of elimination were
observed across routes and exposure concentrations, with urine being the major route of
elimination (>70%) at any dose administered by any route. Total body clearance was rapid and
complete, less than 1% of the absorbed fraction being present in the body 72 h after the highest
exposure regime examined, 1,200 ppm for 6 h. Metabolism of cumene by cytochrome P-450 is
extensive and takes place within hepatic and extrahepatic tissues, including lung (Sato and
Nakajima, 1987), with the secondary alcohol 2-phenyl-2-propanol being a principal metabolite.
Over all doses and routes examined in the Research Triangle Institute study (1989), >50% of
urinary excretion in rats was accounted for by 2-phenyl-2-propanol and its glucuronide or sulfate
conjugates. The balance of excretion in the urine of these exposed rats was accounted for by
conjugates of 2-phenyl-l,2-propanediol and an unknown metabolite, possibly a dicarboxylic acid
metabolite of cumene.
4.0 HAZARD IDENTIFICATION
4.1 Studies in Humans—Epidemiology, Case Reports, and Clinical Controls
No such studies were located for this compound.
-------
4.2 Prechronic and Chronic Studies and Cancer Bioassays in Animals—Oral and
Inhalation
Cushman, J.R., J.C. Norris, D.E. Dodd, K.I. Darmer, and C.R. Morris. 1995. Subchronic
inhalation toxicity assessment of cumene in Fischer 344 rats. J. Am. Coll. Toxicol.
14(2): 129-147.
Two successive subchronic inhalation toxicity studies were conducted with cumene vapors
(>99.9% pure) on Fischer 344 rats. In the first study, groups (21/sex) were exposed to 0, 100,
496, or 1,202 ppm (0, 492, 2,438, or 5,909 mg/m3) cumene vapor for 6 h/day, 5 days/week, for
13 weeks (duration adjusted for continuous exposure to 0, 88, 435, and 1,055 mg/m3). The
second study was a repeat of the first, except that the group size was decreased to 15/sex, and an
additional group (50 ppm, duration adjusted to 44 mg/m3) and a 4 week postexposure period
were added. Animals were sacrificed a few days after the last exposure in the first study and
after the 4-week postexposure period in the second study. Parameters monitored included
clinical signs of toxicity; body weight; food and water consumption; hematology and serum
chemistry; organ weights; and gross pathology and histopathology, including examination of all
respiratory tract tissues (three sections of the lungs and four sections of the nasal turbinates).
In both studies, evaluations of neurological function (functional observation battery [FOB] and
motor activity) were conducted. In the first study, an FOB was performed on 10 rats/sex/group,
and motor activity tests were conducted on 15 rats/sex/group. In the second study, motor activity
tests only were performed on 15 rats/sex/group. The FOBs were performed prior to the exposure
and on the weekends following Weeks 1, 2, 4, 9, and 13 of exposure; motor activity was
determined prior to exposure and on the weekend following Weeks 4, 9, and 13 of exposure.
The same animals were examined at each evaluation. Also in the first study, 6 rats/sex/group
were perfusion-fixed for analysis of the nervous system tissues. Because cataracts were detected
in the first study, a more thorough protocol was used in the second study. In the first study, the
eyes were examined once by a single ophthalmologist during the last week of exposure. In the
second study, eyes were examined independently by two ophthalmologists preexposure and at
Weeks 4, 9, and 13, and at Week 4 postexposure, and any cataracts detected were confirmed
histopathologically. In the first study, sperm from epididymides (taken from 15 male rats/group)
and the left testis from each male were evaluated for sperm count and sperm morphology, and
cross-sections of testes were examined for evaluation of the stages of spermatogenesis in an
effort to judge the potential of cumene to cause reproductive toxicity. Auditory brain stem
responses were measured at 4, 8, 16, and 30 kHz during Postexposure Week 1 of the second
study.
Transient, reversible cage-side observations during exposure periods included hypoactivity,
blepharospasm, and a delayed or absent startle reflex at the highest concentration. Rats exposed
to 496 ppm were reported as being hypoactive during exposure, although no further specifics
were given. An increased incidence of cataracts was observed in males at all exposure
concentrations in the first study. These results were not observed in the second study, in which
incidence of cataractous changes were not different from historical controls nor confirmed by
more extensive histopathological analysis. In the first study, statistically significant (p < 0.05)
exposure-related decreases in motor activity (ambulatory and total activity) were observed in
male rats exposed to the two highest concentrations of cumene, but these results were not
-------
reproduced in the second study. There were no exposure-related changes in the FOBs in either
study. No effects were observed in the neurohistopathological examinations. Evaluation of the
auditory brain stem responses revealed no changes in the auditory function of the exposed
animals, although 3/10 female rats in the highest exposure group were noted to have variability
in their waveforms recorded at 4 kHz. These results were judged by the authors not to be
indications of ototoxicity because of the variability of the responses. It also is noted that the
ototoxicity known to occur with toluene, a structural analog of cumene, is evident only at
frequencies of 8 kHz and higher. The only gross histopathology noted was periocular swelling,
which occurred in animals at the two highest concentrations (and for which neither incidence nor
severity was reported).
In the first study, both absolute and relative weights were increased significantly (>10%,
p < 0.05) in the kidneys, adrenal glands, and livers of both sexes at the highest concentration.
In females, mean kidney weights were increased 11% (absolute) and 16% (relative), adrenal
weights 19% (absolute) and 26% (relative), and liver weights 34% (absolute) and 40% (relative).
In males, mean renal weights were increased 12% (absolute) and 10% (relative), adrenal weights
20% (absolute) and 27% (relative), and liver weights 33% (absolute) and 30% (relative). These
changes also were noted in the liver at the next lower concentration (500 ppm) for both females
(7% absolute, 11% relative) and males (20% absolute, 17% relative). The results of the second
study, with a 4-week postexposure period, indicated limited reversibility of these alterations
because significant mean weight increases still were present in female liver (13% absolute, 11%
relative) and female adrenals (12% absolute, 8% relative) of the highest exposure group. Only
male relative kidney weights (6%) and absolute liver weights (11%) remained increased
significantly. These alterations in weight are considered lexicologically significant and adverse
because such persistence indicates limited reversibility and uncertainty about the progression and
fate of these alterations under chronic exposure. There were no cumene-related differences in
weights of lungs, testes, ovaries, or brain at any exposure level in either study. At the end of the
exposure in the first study, water consumption was increased by as much as 40% in male rats at
the two highest exposure concentrations. Alterations were noted in a number of hematological
parameters, including a concentration-related increase in leukocytes (which is consistent with the
study of Jenkins et al, 1970, below) and platelets in males and females, as well as in
lymphocytes in males at 496 and 1,202 ppm, and significant (p < 0.05) decreases in erythrocyte
parameters (erythrocyte count, hemoglobin, hematocrit, mean corpuscular hemoglobin, and mean
corpuscular hemoglobin concentration) in male rats at these concentrations. All of these
alterations (except for the platelet count, which was increased by around 20% over controls, in
males exposed to the highest concentration) were within normal ranges (Mitruka and Rawnsley,
1981), with no accompanying indications of hematological toxicity, and therefore are considered
of minimal toxicological importance. Morphological evaluation of epididymal and testicular
sperm showed no cumene-related differences in either count, morphology, or stages of
spermatogenesis, although one high-dose rat did have diffuse testicular atrophy.
The only microscopic effect associated with these organ weight changes was an increased
incidence of kidney lesions in male rats at the two highest exposure concentrations. The
incidence of hypertrophy and hyperplasia of proximal tubular epithelial cells and interstitial
nephritis were increased significantly at 496 ppm (12/15 and 13/15, respectively) and 1,202 ppm
(14/15 and 13/15, respectively) compared to controls (1/15 for each effect). There was also an
-------
increase in severity with dose in exposed renal tissues, including hyaline droplet formation,
where dose-related increases in the incidence of moderate and marked severities were 2/15, 3/15,
14/15, and 14/14 in controls and dose groups in ascending order of exposure concentration.
The relevance of these renal effects to human toxicity is questionable because the lesion
described relates closely to the male rat specific nephropathy. The EPA has established scientific
policy and several criteria for assignation of male specific renal nephropathy caused by chemicals
that induce excessive accumulation of a2u-globulin (U.S. EPA, 1991a; Hard et al, 1993). The
renal histopathology reported in this study fulfills several of these criteria: lesions were limited
to males; hyaline droplet formation (as confirmed by the Mallory-Heidenheim method) was
noted and increased in severity in a dose-related fashion; lesions associated with the pathologic
sequence of a2u-globulin nephropathy were noted, including tubular proteinosis (presumably
from exfoliation of epithelial cells into the proximal tubular lumen) and tubular epithelial cell
hyperplasia and hypertrophy (presumed to be regenerative from tubular necrosis). Although a
major criterion is not met in the study, positive identification of the accumulating protein in the
hyaline droplets as a2u-globulin, the pattern described strongly suggests male rat specific
nephropathy. Chronic progressive nephropathy, which also occurs predominately in male rats,
also is characterized by tubular hyperplasia and proteinosis (Montgomery and Seely, 1990), and
this also may be contributory to these renal lesions. The weight alterations in the adrenals and
female kidney are considered potentially adverse. The increased water consumption noted also
may indicate potential for renal effects, although this effect was present at dose levels at which
renal weights were not altered. Although the progression of these weight alterations from
continued exposure cannot be ascertained from this subchronic study, data from the second
(postexposure) study indicate limited reversibility to the adrenals, at least in females. The liver
weight alterations are not viewed as adverse because increase in liver weight without
accompanying pathology is a trait of common microsomal-inducing agents (Sipes and Gandolfi,
1991). Based on the lowest dose at which both relative and absolute weight alterations are
statistically (p < 0.05) and biologically (>10%) significant, 1,202 ppm is a lowest-observed-
adverse-effect level (LOAEL) based on weight alterations observed in the first study in the
adrenal tissues of both sexes and the kidneys in females. The next lower dose, 496 ppm, is a
no-observed-adverse-effect level (NOAEL).
Fabre, R., R. Truhaut, J. Bernuchon, and F. Loisillier. 1955. Toxicologic studies of solvents to
replace benzene. III. Study of isopropyl benzene or cumene. Arch. Mai. Prof. 16(4): 285-299.
In an inhalation study, Wistar rats were exposed to 2,500 mg/m3 cumene vapor for 8 h/day,
6 days/week, for up to 180 days (duration adjusted to 714 mg/m3), and rabbits were exposed to
6,500 mg/m3 using the same exposure regimen (duration-adjusted concentration is 1,857 mg/m3).
Clinical signs of toxicity, body weight, blood and bone marrow parameters, and histopathological
effects (brain, cerebellum, heart, stomach, liver, pituitary, intestine, spinal cord, bone marrow,
ovary, pancreas, parathyroid, lung, spleen, kidney, adrenals, testicle, thymus, thyroid, and
bladder) were monitored. In the rat, the number of red blood cells decreased slightly, but no test
for statistical significance was performed. Histological effects reported were "passive
congestion" in the lungs, liver, spleen, kidney, and adrenals and the presence of hemorrhagic
zones in the lung, hemosiderosis in the spleen, and lesions from epithelial nephritis "in some
-------
cases". It was not clear if these effects occurred in both species. Both of these exposure levels
induced adverse effects.
Jenkins, L.J., Jr., R.A. Jones, and J. Siegel. 1970. Long-term inhalation screening studies of
benzene, toluene, o-xylene, and cumene on experimental animals. Toxicol. Appl. Pharmacol.
16: 818-823.
In an inhalation exposure study, groups of Sprague-Dawley or Long-Evans rats (n = 15),
Princeton-derived guinea pigs (n = 15), beagle dogs (n = 2), and squirrel monkeys (n = 2) were
exposed to cumene at concentrations of 18 or 147 mg/m3 continuously for 90 days. Initial and
terminal body weight, hematologic and clinical chemistry parameters, and histopathologic data
were collected. The only effect noted was a slight degree of leukocytosis in rats at both
concentrations, which is consistent with the results of Cushman et al. (1995). The same effect
occurred in a similar group of rats exposed to cumene at 1,200 mg/m3 for 8 h/day, 5 days/week
for 30 exposures, although none were indicated as statistically significant. No other
lexicologically significant effects were noted in either guinea pigs, dogs, or monkeys. This
single concentration defines a LOAEL for this study.
Monsanto Company. 1986. One-month study of cumene vapor administered to male and female
Sprague-Dawley rats by inhalation. U.S. EPA/OTS Public Files, 8D submission. Microfiche
No. OTS0513229.
Male and female Sprague-Dawley rats (10/sex/group) were exposed to cumene vapor
concentrations of 0, 105, 300, or 599 ppm (0, 517, 1,475, or 2,946 mg/m3) for 6 h/day,
5 days/week, for approximately 4 weeks (minimum exposure, 20 days). Urinalysis, hematology,
and clinical biochemistry on serum (including BUN, SGOT, LDH, and total bilirubin) were
performed. Animals were observed daily for signs of toxicity. Necropsy was performed on
5 rats/sex/group at the end of the exposure. No deaths occurred during the study. Cage-side
observations included hypoactivity in the high-concentration animals on some days during the
exposure period. Signs of toxicity observed during the pre- or postexposure checks included
concentration-related increases in side-to-side head movements in both males and females in all
dose groups (combined total incidence during exposure of 0, 14, 21, and 48 for controls and the
three dose groups), head tilt (total incidence in all dose groups of 0, 5, 4, and 8 for controls and
the three dose groups), and arched back in one female in the high-dose group. Other less
significant observations included dried, reddish discharge around the nose and near the eyes in
nearly all dose groups and controls. Other effects observed in males include alopecia (mid-dose
group during Weeks 2, 3, and 4 of exposure) and swollen conjunctiva during Week 4 in the
high-dose group. Increases (p < 0.05) in mean absolute left and right kidney weights were
observed high-dose males, as were increases in left kidney in low and mid-dose males. In high-
dose females, the mean absolute weight of left kidneys was greater (p < 0.05) than in controls.
No compound-related pathological changes were detected during gross or microscopic
examination. Assuming that the renal changes among the males were associated with male rat
specific nephropathy (see above), the cage-side observations of head tilt and head movements
become the critical effects for this short-term study with a LOAEL of 105 ppm. This study
confirms that renal weight changes occur in females, thereby corroborating similar effects
reported in the study of Cushman et al. (1995). It should be noted that the effects associated with
-------
central nervous system perturbation (i.e., head movements) were not noted in several other longer
term studies, including that of Cushman et al. (1995) where neurotoxicity was specifically
assessed.
Wolf, M.A., V.K. Rowe, D.D. McCollister, R.L. Hollingsworth, and F. Oyen. 1956.
Toxicological studies of certain alkylated benzenes and benzenes. Arch. Ind. Health
14: 387-398.
Groups of 10 female Wistar rats were administered 139 doses of cumene by gavage in olive
oil at 154, 462, or 769 mg/kg/day over a 194-day (6- to 7-mo) period (duration adjusted dose x
139/194 = 110, 331, or 551 mg/kg/day). Rats given olive oil served as controls (n = 20). Body
weights, food consumption and mortality were noted throughout the study, although no results
are shown. Hematological evaluations were conducted after doses 20, 40, 80, and 130, and blood
urea nitrogen determinations and gross and histological examinations (lungs, heart, liver,
kidneys, spleen, adrenals, pancreas, and femoral bone marrow) were conducted at the end of the
study. Effects were not observed at 154 mg/kg/day. An increase in average kidney weight was
noted as a "slight effect" at 462 mg/kg/day. A more pronounced weight increase in average
kidney weight, noted as a "moderate effect", occurred at 769 mg/kg/day, although no
quantitative data is presented. The LOAEL is considered to be 462 mg/kg/day, and the NOAEL
is 154 mg/kg-day.
4.3 Reproductive/Developmental Studies—Oral and Inhalation
Bushy Run Research Center. 1989b. Developmental toxicity study of inhaled cumene vapor in
CD (Sprague-Dawley) rats. Final project report 52-621. TSCATS/0522881; EPA/OTS Doc.
No. 40-8992172.
Sprague-Dawley rats (25/group) were exposed to 0, 99, 488, or 1,211 ppm (0, 487, 2,399,
or 5,953 mg/m3) cumene for 6 h/day on Days 6 through 15 of gestation. Dams were observed for
clinical signs of toxicity, body weight, gravid uterine weight, liver weight, abnormalities of the
respiratory tract, numbers of corpora lutea, implantation sites, resorptions, and living and dead
fetuses. Fetuses were examined for external, visceral, and skeletal malformations and variations.
At the two highest concentrations, perioral wetness and encrustation, hypoactivity and
blepharospasm, and significantly (p < 0.05) decreased food consumption were observed in the
dams. At the highest concentration, there was a significant (p < 0.01) decrease in body weight
gain on Gestation Days 6 through 9 (accompanied by a significant decrease in food consumption)
and a slight increase (7.7%) in relative liver weight. There were no statistically significant
adverse effects on reproductive parameters or fetal development. For this study, 1,211 ppm is a
developmental NOAEL, and 488 ppm (2,399 mg/m3) is a maternal NOAEL.
Bushy Run Research Center. 1989c. Developmental toxicity study of inhaled cumene vapor in
New Zealand White rabbits. Final project report 52-622. TSCATS/0522881; EPA/OTS Doc.
No. 40-8992172.
New Zealand White rabbits (15/group) were exposed to 0, 492, 1,206, or 2,297 ppm
(0, 2,418, 5,928, or 11,292 mg/m3) cumene for 6 h/day on Days 6 through 18 of gestation.
Dams were observed for clinical signs of toxicity, body weight, gravid uterine weight, liver
-------
weight, abnormalities of the respiratory tract, numbers of corpora lutea, implantation sites,
resorptions, and living and dead fetuses. Fetuses were examined for external, visceral, and
skeletal malformations and variations. Two does died at the highest exposure concentration.
There were significant (p < 0.01) reductions in body weight gain (178.11 g lost compared to
31.55 g gained in the control group) and food consumption at the highest exposure level.
Significantly (p < 0.05) reduced food consumption also was observed in the 492- and 1,206-ppm
exposure groups. Clinical signs of toxicity observed in the does included significant (p < 0.01)
increases in perioral and perinasal wetness and blepharospasm at the high concentration. At
necropsy, there were color changes in the lungs of 33% of the does exposed to 2,297 ppm.
Relative liver weight was significantly (p < 0.01) elevated (16.8% of control weight) at the
highest exposure level. There were no statistically significant effects on gestation parameters;
however, there were nonsignificant increases in nonviable implants, and early resorptions and a
nonsignificant decrease in the percent of live fetuses at 2,297 ppm. The only variation observed
was an increase in ecchymosis (hemorrhagic areas of the skin) of the head that occurred in all
exposed animals (0, 5.4, 3.7, and 4.9% of the fetuses and 0, 35.7, 28.6, and 25.0% of the litters at
0, 492, 1,206, and 2,297 ppm, respectively), which was not concentration-related. On further
analysis, EPA (1991b) determined that the rates of ecchymosis in this study were within the
ranges observed for the historical controls of this test facility. No other malformations or
variations differed from control values. Although the alterations observed in gestational
parameters were not significant, they were consistent in indicating possible developmental
effects. Based on this consistency, the highest exposure level is considered a LOAEL. The next
lower level, 1,206 ppm, is considered a NOAEL for both developmental and maternal effects.
No multigeneration reproductive study exists for this compound by either oral or inhalation
route. Neither are there any data concerning cumene exposure prior to mating, from conception
to implantation, or during late gestation, parturition, or lactation. The principal study (Cushman
et al, 1995), however, conducted morphological evaluation of epididymal and testicular sperm in
rats exposed for 13 weeks to cumene vapors. No cumene-related differences in count,
morphology, or stages of spermatogenesis were noted, although one high-dose rat did have
diffuse testicular atrophy. The IRIS entry for the structurally related compound toluene
(methylbenzene) reports occurrence of a significant decrease (p < 0.05) in weight relative to
controls in the offspring in a one-generation reproductive study at a NOAEL of 1,885 mg/m3
(U.S. EPA, 1997).
Cumene was a minor component of aromatic naphtha vapors that were tested in a
inhalation reproductive toxicity study in rats and a developmental toxicity study in mice (McKee
et al., 1990). These studies were read as part of this assessment but were not considered further
because the concentration of cumene was less than 3% (about 2 to 40 ppm maximum
concentration) of the vapors tested.
-------
4.4 Other Studies
Neurotoxicology
Cumene appears to be similar to many solvents that produce a profile of acute effects
similar to those of known central nervous system (CNS) depressants such as alcohol. The
occurrence of neurological effects from inhalation exposure to cumene has been confirmed in
several studies, some of which are described below. These studies are acute exposures that show
neurotoxicological effects only at quite high concentrations (>500 ppm). Neurotoxicological
effects were not observed, however, in the longer term inhalation study by Cushman et al. (1995;
Section 4.2), which included complete batteries of functional and motor activity tests and
neurohistopathology.
Cumene was one of six alkylbenzenes tested at 0, 2,000, 4,000, or 8,000 ppm that all
produced a short-lived profile of neurobehavioral effects in mice, indicating CNS depressant
activity (Tegeris and Balster, 1994). Effects noted from brief (20-min) exposures to cumene
included those on CNS activity (decreased arousal and rearing at >2,000 ppm) muscle
tone/equilibrium (changes in grip strength and mobility >4,000 ppm), and sensorimotor activity
(including decreased tail pinch and touch response >4,000 ppm).
In an acute experiment accompanying the subchronic exposures, Cushman et al. (1995)
exposed Fischer 344 rats once to 0, 100, 500, or 1,202 ppm for 6 h and conducted functional
observations 1 h postexposure. Gait abnormalities and decreased rectal temperatures were noted
for both sexes at the highest exposure level only. Decreased activity levels were noted for both
sexes at the highest levels and for females only at the next highest level (500 ppm) of exposure.
Males, but not females, from the highest exposure group had decreased response to toe pinch at
6 h postexposure.
In a 5-day inhalation study, Fischer 344 rats exposed to 2,000 or 5,000 ppm (9,832 or
24,580 mg/m3) cumene vapor for 6 h/day showed toxic effects from exposure (Gulf Oil Corp.,
1985). All rats in the high-exposure group died after 2 days. At the lower dose, females
demonstrated CNS effects (hypothermia and staggering). Similar, but more severe, symptoms
were observed in the high-exposure animals before they died.
Fischer 344 rats (10/sex/group) were exposed to cumene at 0, 251, 547, 1,047, or
1,290 ppm (0, 1,234, 2,689, 5,147, or 6,342 mg/m3) for 6 h/day, 5 days/week for 2 weeks
(Chemical Manufacturers Association, 1989). Initial exposures to 2,000 ppm (9,832 mg/m3)
for 1 to 2 days resulted in such severe neurological and respiratory effects that the concentration
levels were reduced to those given above. During the remainder of the 2-week exposure period,
clinical observations (ocular discharge, decreased motor activity or hyperactivity, and ataxia)
were noted sporadically at all levels except 251 ppm. For females in the two highest dose
groups, the average relative kidney weight and relative and absolute adrenal weights were
increased significantly over control values. These data provide corroboration for these same
effects reported in the study of Cushman et al. (1995).
Respiratory Irritation
10
-------
The concentration of cumene causing a 50% reduction in the respiratory rate in mice was
determined to be 2,058 ppm (10,117 mg/m3) (Kristiansen et al, 1986). This concentration is
quite high and in the range where repeated exposure caused death and morbidity in rats (Gulf Oil
Corp., 1989; Chemical Manufacturer's Association, 1989) and rabbits (Bushy Run Research
Center, 1989c).
Genotoxicity
Cumene was tested at concentrations up to 2,000 //g/plate in a S. Typhimurium reverse
mutation assay (modified Ames test); negative results were observed with and without metabolic
activation (Lawlor and Wagner, 1987). Cumene was negative in an Ames assay at
concentrations up to 3,606 //g/plate (Florin et al., 1980). Cumene also tested negative, with and
without metabolic activation, in a set of HGPRT assays (using Chinese hamster ovary cells) at
concentrations up to 225 //g/mL (Yang, 1987; Gulf Life Sciences Center, 1985a).
A micronucleus assay performed in mice given up to 1 g/kg cumene by gavage was negative
(Gulf Life Sciences Center, 1985b).
A recent micronucleus assay done in Fisher 344 rats, however, gave values that were
weakly positive, although little dose response was seen, and deaths occurred at the highest dose
(5/10 animals at 2.5 g/kg ip, an extraordinarily high dose; NTP, 1996). In the first of two
duplicate NTP experiments, the average number of micronuclei per thousand polychromatic
erythrocytes at 72 h was 0.5 for controls, 1.2 at 78 mg/kg, 1.2 at 156 mg/kg, 1.3 at 313 mg/kg,
0.8 at 625 mg/kg, 2.6 at 1,250 mg/kg, and 1.3 at 2,500 mg/kg cumene and 17.3 in the positive
control (25 mg/kg cyclophosphamide). A similar lack of dose-response was noted in the second
experiment.
Cumene failed to induce significant rates of transformation in B ALB/3 T3 cells (without
activation) at concentrations up to 500 //g/mL (Putnam, 1987) but tested positive in an earlier
cell transformation test also using B ALB/3 T3 cells, in which an increase in transformations was
observed 60 //g/mL (Gulf Oil, 1984a). One test for unscheduled deoxyribonucleic acid (DNA)
synthesis (UDS) in rat primary hepatocytes, using exposures of up to 24 //g/mL cumene (without
activation), was negative (Curren, 1992), whereas results from an earlier test indicated UDS at
doses of 16 and 32 //g/mL cumene (Gulf Oil, 1984b). Those tests indicating positive mutagenic
potential (Gulf Oil, 1984a,b) were considered equivocal because they were not reproducible.
4.5 Synthesis and Evaluation of Major Noncancer Effects and Mode of Action
(If Known)—Oral and Inhalation
The overall hazard profile presented by cumene is one of low toxicity. Short-term acute
exposures of animals to high concentrations (>1,000 ppm) demonstrate that cumene, like other
solvents, can induce transient reversible neurotoxic effects. However, neither neurotoxicity,
portal-of-entry effects, developmental effects, nor markedly adverse systemic toxicity are
observed after long-term repeated dose studies conducted in animals at lower concentrations
(< 500 ppm).
11
-------
The increased renal weights in female rats reported by Cushman et al. (1995) to occur at the
highest concentrations tested are considered lexicologically significant under the conditions of
less than lifetime exposure because the fate and progression of such effects with longer exposure
are not known. Increased renal weights also have been reported in female rats in the 2-week
inhalation study of the Chemical Manufacturer's Association (1989), the 4-week inhalation study
of Monsanto (1986), and the oral gavage study of Wolf et al. (1956), although none observed or
reported renal histopathology.
Renal histopathology that included hyaline droplet formation and an increase in the
incidence of proximal tubular hypertrophy was observed in males only by Cushman et al. (1995).
These findings, along with others documented in this study (see Section 4.2) are among criteria
used to identify chemically induced male rat a2u-globulin-specific nephropathy (U.S. EPA,
1991a; Hard et al., 1993), which EPA does not consider an appropriate endpoint to determine
noncancer toxicity. Although it is not shown conclusively that the renal effects in the male rats
are attributable to an a2u-globulin mechanism, the available evidence strongly suggests that such
a mechanism is operable with this compound.
Renal weight changes also were noted in the male rats by Cushman et al. (1995). However,
the extent of association of the renal weight increase in males with the a2u-type histopathology is
not clear. The increase may either precede or be independent of renal histopathology.
Nevertheless, these weight changes noted in kidneys of male rats may be confounded by
indications of an a2u-globulin mechanism or exacerbation of rat chronic progressive nephropathy
(Montgomery and Seely, 1990); therefore, they are not used in this assessment.
The study of Cushman et al. (1995) with inhaled cumene showed that, in addition to
increases in kidney weights, liver weights also were increased in both sexes of rats in a
concentration-dependent manner. Increased liver weight is also an effect observed in rats
exposed to toluene (methylbenzene) a structural analog of cumene (NTP, 1990). Liver weight
increases without accompanying histopathology often are considered to result from both
hyperplastic and hypertrophic parenchymal changes associated with metabolism of the toxicant,
with the increases usually being reversible on discontinuance of the toxicant (Sipes and Gandolfi,
1991). Cushman et al. (1995) observed no hepatic histopathology. In addition, the 4-week
recovery period incorporated in the second subchronic study by Cushman demonstrated that the
liver changes were reversible. In male rats exposed to the highest concentration of cumene, the
33% increase in absolute liver weight relative to controls observed at the end of the first study
was decreased to only 11% at the end of the second study. In female rats, the results were
similar, with a 34% increase at the end of the first study, as compared to a 13% increase at the
end of the second study. Thus, although Cushman et al. (1995) did not document actual
increases in hepatic metabolism, other characteristics of the hepatic response indicate that the
liver responses were highly likely to be adaptive in nature and nonadverse.
Ototoxicity also is an effect observed in rats exposed to toluene but that was not observed
in the study of Cushman et al. (1995) at cumene concentrations as high as 1,200 ppm.
Neurotoxicological effects from long-term exposure to cumene warranted examination.
After short-term exposures to high concentrations (20 min at 2,000 to 8,000 ppm), cumene, along
12
-------
with many other solvents, has been shown to produce transient symptoms typical of CNS
perturbation typical of many other solvents (Tegeris and Balster, 1994), such as those reported in
the principal study (appearance of hypoactivity, blepharospasm, and delayed startle reflex) and in
the study by Monsanto (1986), in which head movements and hypoactivity were noted. Longer
term exposures to lesser concentrations do not appear to result in detectable effects because the
extensive examinations conducted in the Cushman et al. (1995) study produced no objective
reproducible indications of neurotoxicological adversity in rats that had undergone repeated
exposures to cumene for 13 weeks at concentrations as high as 1,202 ppm.
Cumene has a superficial similarity (an aromatic ring) to benzene. Nevertheless, blood
toxicity (a known effect of benzene) has been a focus of both short- and long-term studies.
Although clinical blood parameters were monitored in several long-term studies of several
species exposed to cumene (Fabre et al., 1955; Jenkins et al., 1970; Wolf et al., 1956), only
Cushman et al. (1995) detected any significant hematological perturbations. Due to the relatively
small alterations and the wide-ranging normal values for a number of these parameters, these
alterations were considered to be of minor toxicological significance.
4.6 Weight of Evidence Evaluation and Cancer Classification—Synthesis of Human,
Animal, and Other Supporting Evidence; Conclusions About Human Carcinogenicity;
and Likely Mode of Action
Under the proposed Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1996), it is
concluded that the carcinogenic potential of cumene cannot be determined because no adequate
data, such as well-conducted long-term animal studies or reliable human epidemiological studies,
are available to perform any assessment. Under the current Risk Assessment Guidelines (U.S.
EPA, 1987a), cumene is assigned carcinogen category D (not classifiable), indicating inadequate
or no human or animal data.
The metabolic pathways of this compound are, by and large, known and do not appear to
involve any suspect reactive species. One in vivo mutagenicity test (micronucleus) did give a
weakly positive result with a dose that produced mortality, although cumene gave negative
results in a relatively complete battery of in vitro and in vivo mutagenicity tests, including gene
mutation, chromosomal aberration, and primary DNA damage. Trends in structure-activity
relationships are unclear as neither toluene (methylbenzene) or ethylbenzene has been classified
by EPA with respect to carcinogenicity. It is clear, with respect to metabolism, however, that
cumene is more analogous to methylbenzene (toluene) than to ethylbenzene, and toluene showed
no evidence of carcinogenic activity in rats or mice in a 2-year inhalation study (NTP, 1990). At
present, there is no likely genotoxic mode-of-action to consider for carcinogenic activity by
cumene. In summary, there is not much suspicion that cumene would pose a significant
carcinogenic hazard.
4.7 Other Hazard Identification Issues
4.7.1 Possible Childhood Susceptibility
13
-------
A number of factors may differentially affect childrens' responses to toxicants. The only
toxicity information on cumene of possible relevance to this issue is that from developmental
studies, one study in rats (Bushy Run Research Center, 1989a) and another in rabbits (Bushy
Run Research Center, 1989b), in which no adverse fetal effects were observed. There is too little
information to make any further statements about how children may be differentially affected by
cumene, as there are no data regarding cumene exposure prior to mating, from conception
through implantation, or during late gestation, parturition, or lactation.
4.7.2 Possible Gender Differences
The only gender-related difference observed in the current data on cumene is the
occurrence of renal histopathology in male rats only. However, this phenomenon is more than
likely related to or confounded by the male specific nephropathy (U.S. EPA, 199la; Hard et al,
1993) and has no relevance to humans.
5.0 DOSE RESPONSE ASSESSMENTS
5.1 Oral Reference Dose
5.1.1 Choice of Principal Study and Critical Effect—with Rationale and Justification
The study of Wolf et al. (1956) is a repeated dose study (6 to 7 mo) of cumene via the oral
route. The study suffers from several deficiencies, including small group sizes and the lack of
any quantitative data reporting. The only significant effect observed in this study is a description
of dose-related increases in average renal weights observed in the animals exposed to the middle
and high dosages (462 and 769 mg/kg/day). Too, the observations of Wolf et al. were in female
rats (the only sex tested) so that the renal effects are not likely to be confounded as are those
reported for males in the study of Cushman et al. (1995). Similar weight alterations have been
reported in other less-than-lifetime exposures to cumene (Cushman et al., 1995), in which they
have been shown to have limited reversibility. These alterations are considered lexicologically
significant and adverse because such persistence indicates limited reversibility and uncertainty
about the progression and fate of these alterations under true chronic exposure. The lack of any
such effect at the lowest dose tested (154 mg/kg/day; duration adjusted, 110 mg/kg/day) defines
the NOAEL of this study.
An alternative possibility for the principal study would be to use the results of the Cushman
et al. (1995) subchronic inhalation study after performing a route-to-route extrapolation. Limited
interroute kinetic information (blood levels of total metabolites only) is available in rats, from
which comparable blood levels and tissue levels possibly could be calculated for oral versus
inhalation exposures (Research Triangle Institute, 1989). However, the Wolf et al. (1956) study,
although limited in quality, is via the oral route and is of longer exposure duration than the
inhalation study (6 versus 3 mo). Based on these facts, it is judged to be more appropriate to use
the study of Wolf et al. (1956).
5.1.2 Methods of Analysis—No-Observed-Adverse-Effect Level and
14
-------
Lowest-Observed-Adverse-Effect Level
The increase in renal weight in female rats observed at the middle dose (462 mg/kg/day) of
the Wolf et al. (1956) study is considered a LOAEL, and the low dose in this study
(154 mg/kg/day), at which no adverse effects were noted in any systems examined, was
designated the NOAEL. Benchmark dose analysis was not attempted for this endpoint because
no quantitative data are presented.
5.1.3 Oral Reference Dose Derivation—Including Application of Uncertainty Factors and
Modifying Factors
The NOAEL for increased kidney weight in the Wolf et al. (1956) study is 154 mg/kg/day,
and the NOAEL(ADJ), based on adjustment for the stated dosing schedule of 139 doses/
194 days, equals 110 mg/kg/day.
Uncertainty factors (UFs) are applied to account for recognized uncertainties in
extrapolation from experimental conditions to the assumed human scenario (i.e., chronic
exposure over a lifetime). Historically, UFs are applied as values of 10 in a multiplicative
fashion (Dourson and Stara, 1983). Recent EPA practice, however, also includes use of a partial
UF of 101/2 (3.333; U.S. EPA, 1994b) on the assumption that the actual values for the UFs are
lognormally distributed. Application of these factors in the assessments is that, when a single
partial UF is applied, the factor is rounded to 3, such that the total factor for a UF of 3 and 10, for
example, would be 30 (3 x 10). When two partial UFs are evoked, however, they are not
rounded, such that a UF of 3, 3, and 10 would result in a total uncertainty of 100 (actually
1Ql/2 x 1Ql/2 x jQl)
Uncertainty factors and the justification for their use are as follows. A factor of 10 is used
for extrapolation of intraspecies differences in response (human variability) as a means of
protecting potentially sensitive human subpopulations. A factor of 10 is applied for
consideration of interspecies variation. A full factor is considered necessary for this variation as
no human toxicity information currently exists. Partial UFs also are applied for 6 mo to chronic
duration extrapolation and for database deficiencies. The partial database deficiency is evoked
because of the lack of a full-scale multigeneration reproductive study. Cushman et al. (1995)
provides evidence for a lack of concern that cumene may be a reproductive toxicant. However,
these data are limited in that they can not provide a complete scientific argument that would
definitively exonerate cumene as a reproductive toxicant. For example, there are no data
regarding cumene exposure to mating, from conception through implantation, or during late
gestation, parturition, or lactation. The wide tissue distribution demonstrated after inhalation of
cumene, which included the reproductive organs (Research Triangle Institute, 1989),
demonstrates that these tissues would be as highly exposed as the remainder of the body. The
IRIS entry for the structurally related compound toluene (methylbenzene) reports occurrence of a
significant decrease (p < 0.05) in weight relative to controls in the offspring in a one-generation
reproductive study at a NOAEL of 1,885 mg/m3 (U.S. EPA, 1997). The total UF = 10 x 10 x
101/2 x 101/2 = 1,000. No modifying factor (MF) is considered necessary.
RfD = 110 mg/kg/day - 1,000 = IE - 1 mg/kg/day
15
-------
5.2 Inhalation Reference Concentration
5.2.1 Choice of Principal Study and Critical Effect—with Rationale and Justification
The pair of 3-mo subchronic inhalation studies reported by Cushman et al. (1995) are
designated together as the principal study for derivation of the RfC. Although the inhalation
study of Fabre et al. (1955) had a longer exposure regime (around 6 mo), only a single exposure
concentration was employed versus the four in the principal study. The study of Jenkins et al.
(1970) used more species than the principal study and attained nearly continuous exposure for
90 days. In comparison, the principal study used larger groups of animals and conducted
thorough and extensive cage-side observations, neurotoxicological examinations, and auditory
function tests. Also, neither Fabre et al. (1955) nor Jenkins et al. (1970) reported any significant
adverse effects, unlike the principal study. The choice of Cushman et al. (1995) as the principal
study is considered justified because of these methodological and analytical attributes.
The critical effects are the increases (p < 0.05, changes >10% relative to controls) in both
absolute and relative mean weights in the adrenal glands of both sexes and in the kidneys of
female rats at the highest concentration tested. Although both absolute and relative liver weights
also were increased, this effect was not considered adverse because an increase in liver weight,
without accompanying pathology, is a trait of common microsomal-inducing agents (Sipes and
Gandolfi, 1991). The next lower concentration is designated as the NOAEL, although some
other effects were described somewhat subjectively and generally at this concentration
(hypoactivity and some periorbital swelling); these are not deemed sufficient to warrant
consideration of this concentration a LOAEL, primarily because of their occurrence in controls.
5.2.2 Methods of Analysis—No-Observed-Adverse-Effect Level and
Lowest-Observed-Adverse-Effect Level
The highest concentration tested, 1,202 ppm, is designated the LOAEL. The next lower
dose, 496 ppm, is designated the NOAEL.
Analyses for benchmark concentrations (BMCs) were performed on the absolute weight
alterations in male and female adrenal and female renal weights (Appendix A). An overview of
the benchmark dose approach for health risk assessments is given in U.S. EPA (1995c). The
only data set of the three that could be modeled to a level of statistical significance (F < 0.01)
was male adrenal weights. The BMC10 (the lower 95% confidence bound on the concentration
from the maximum likelihood estimate of a 10% relative change) values obtained for these data
were identical to one another for the two models, 484 ppm.
The critical effect that was the most corroborated by the cumene database, however, was
the increase in female kidney weight, which was not modeled successfully. Rather than rely on
unsuccessful modeling results or on results from a possibly inappropriate endpoint, the NOAEL
of 496 ppm is used for all further quantitative analysis. It should be noted that the BMC10 of
484 ppm obtained for the only data set that was successfully modeled, male adrenal weight gain,
is nearly the same as the NOAEL.
16
-------
Calculation of the human equivalent concentration (HEC) from the NOAEL of 496 ppm is
shown below.
The NOAEL first is converted to milligrams per cubic meter and duration adjusted; then,
assuming 25 °C and 760 mm Hg,
NOAEL (milligrams per cubic meter) = 496 ppm x MW/24.45 = 2,438 mg/m3.
• This converted value then is duration adjusted to continuous exposure, which equals the
NOAEL(ADJ).
2,438 mg/m3 x 6h/day x 5 days/7 days = 435 mg/m3 = NOAEL(ADJ)
The scenario for this effect was a gas causing a systemic or extrarespiratory effect that
assumed attainment of periodicity for the blood/air (b/a) cumene concentrations. Because no
b/a lambda (i.e., partition coefficient) values for cumene are known for either animals or
humans, a default value of one is used for this ratio, which indicates that there exist no
differences between animals and humans in blood concentrations attained for the same air
concentration of cumene.
• Therefore, NOAEL(HEC) = NOAEL(ADJ) x [b:a lambda(a) / b:a lambda (h)] = 435 mg/m3 x
1 =435 mg/m3.
5.2.3 Inhalation Reference Concentration Derivation—Including Application of Uncertainty
Factors and Modifying Factors
The NOAEL(HEC) for increased kidney and adrenal weights in the Cushman et al. (1995)
study is 435 mg/m3.
Uncertainty factors are applied to account for recognized uncertainties in extrapolation
from experimental conditions to the assumed human scenario (i.e., chronic exposure over a
lifetime). Historically, UFs are applied as values of 10 in a multiplicative fashion (Dourson and
Stara, 1983). Recent EPA practice, however, also includes use of a partial UF of 101/2 (3.333;
U.S. EPA, 1994b) on the assumption that the actual values for the UFs are lognormally
distributed. Application of these factors in the assessments is that, when a single partial UF is
applied, the factor is rounded to 3, such that the total factor for a UF of 3 and 10, for example,
would be 30 (3 x 10). When two partial UFs are evoked, however, they are not rounded, such
that a UF of 3, 3, and 10 would result in a total uncertainty of 100 (actually 101/2 x 101/2 x 1Q1).
The UFs applied and the justification for their use are as follows. A factor of 10 is used for
extrapolation of intraspecies differences in response (human variability) as a means of protecting
potentially sensitive human subpopulations. A factor of 10 is applied for subchronic to chronic
extrapolation as the progression or fate of observed effects in kidney and adrenals resultant from
true chronic administration is uncertain. A partial (101/2) UF is applied for consideration of
interspecies extrapolation, which already has been addressed partially through the calculation of
an HEC. A partial UF also is used for database deficiencies, principally because of the lack of a
17
-------
full-scale multigeneration reproductive study (as discussed above in the section on UF for the
RfD). The total UF = 10 x 10 x 101/2 x 101/2 = 1,000.
No MF is proposed for this assessment.
RfC = 435 mg/m3 - 1,000 = 4E - 1 mg/m3.
5.3 Cancer Assessment
As discussed above (Section 4.5), there are no epidemiological, occupational, or long-term
in vivo animal studies addressing the issue of cancer. No data exist to support any quantitative
cancer assessment for this compound.
6.0 MAJOR CONCLUSIONS IN CHARACTERIZATION OF HAZARD
IDENTIFICATION AND DOSE-RESPONSE ASSESSMENTS
6.1 Hazard Identification
Cumene is a water insoluble petrochemical used in the manufacture of several chemicals,
including phenol and acetone. It is metabolized primarily to the secondary alcohol, 2-phenyl-
2-propanol, in both animals and humans. This alcohol and conjugates thereof are excreted
readily by both rodents and humans.
No human toxicity data exists for cumene. Increases in organ weights (most notably
kidney) are the most prominent effects observed in rodents exposed repeatedly to cumene by
either the oral (Wolf et al, 1956) or the inhalation (Cushman et al, 1995) routes. No adverse
effects were observed in rat or rabbit fetuses whose mothers had been exposed to aerosolized
cumene during development.
The sparsity of long-term repeated dose toxicity data and the absence of any human toxicity
data both constitute areas of scientific uncertainty in this assessment. The longest repeated-dose
study is the oral study of Wolf et al. (1956), at about 7 mo, followed by the 3-mo subchronic
inhalation study of Cushman et al. (1995). Neither of these studies is sufficient in duration to
reveal the fate of the observed alterations in organ weights. Although there exists no
multigeneration reproductive study for cumene, its rapid metabolism and excretion, coupled with
the information on sperm morphology reported by Cushman et al. (1995), indicate that cumene
has low potential for reproductive toxicity.
The potential human hazard for carcinogenicity of cumene has not been determined,
although there is some evidence that suggests this compound may not be likely to produce a
carcinogenic response (i.e., numerous genotoxic tests, including gene mutation, chromosomal
aberration, and primary DNA damage tests, all but one of which were negative or not
reproducible, were conducted). No highly reactive chemical species are known to be generated
during the metabolism of cumene. Although structure-activity relationships to cumene are
problematic, it is clear that cumene, with respect to metabolism, is more analogous to
18
-------
methylbenzene (toluene) than to ethylbenzene. Toluene has been tested in a 2-year inhalation
protocol and showed no evidence of carcinogenic activity in either rats or mice (NTP, 1990).
No dose-response assessment was performed on this compound because no data are available.
6.2 Dose Response
The quantitative estimates of human risk as a result of low-level chronic exposure to
cumene are based on animal experiments because no human data exist.
The human dose that is likely to be without an appreciable risk of deleterious noncancer
effects during a lifetime (the RfD) is 0.1 mg/kg-day. This amount is 1/1000 of the dose, adjusted
for the stated schedule, at which no adverse effects were noted in female rats dosed orally with
cumene over a period of about 7 mo (Wolf et al, 1956).
The overall confidence in the RfD assessment is low to medium. The confidence in the
principal study is low. For purposes of quantitative assessment, the quality of the principal study
(Wolf et al., 1956) is marginal because the group sizes are minimal and comprise females only,
and little quantitative information is presented. The confidence in the database, judged here as
medium to low, is improved from the earlier version on IRIS, principally because of the
availability of inhalation developmental studies; some reproductive measures; corroboration of
the critical effect by other studies, including those using oral dosing; and kinetic information.
Kinetic information on oral and inhalation routes of exposure (Research Triangle Institute, 1989)
justifies utilization of inhalation developmental studies performed in two species, rats and
rabbits, in which no adverse results were noted. However, no 2-year chronic study is available
via the oral or inhalation route. No multigeneration studies are available for this compound.
Results on some male reproductive parameters were, however, documented in Cushman et al.
(1995), the principal study for the RfC. The rapid metabolism and excretion of cumene in both
animals and humans, coupled with the information on sperm morphology reported by Cushman
et al. (1995), also indicate cumene to have a low potential for reproductive toxicity. The critical
effect, altered tissue weights, was the same across routes of exposure (this was also the critical
effect for the RfC) and was observed in several studies giving confidence in the consistency of
this effect.
Justification for the use of a partial uncertainty factor for subchronic to chronic
extrapolation was twofold: (1) the duration of the principal study (6 to 7 mo) was intermediate,
between subchronic (3mo) and chronic (24 mo) duration, and (2) toxicokinetic data (Section 3)
indicate that inhaled cumene and its metabolites are cleared quickly from both humans and rats,
which also could indicate low potential for cumulative damage.
The daily exposure to the human population that is likely to be without an appreciable risk
of deleterious effects during a lifetime (the RfC) is 4E-1 mg/m3. This concentration is 1/1000 of
the adjusted no-effect level for significant increases (>10%) in renal and adrenal weights in rats
exposed to cumene in the subchronic inhalation study of Cushman et al. (1995).
The overall confidence in the RfC assessment is medium. The RfC is based on rat
subchronic inhalation studies performed with relatively large group sizes in which thorough
19
-------
histopathological analyses and ancillary studies of neurotoxicity and ocular pathology were
performed. The scientific quality of this evidence is high. The confidence in the database for the
cumene RfC is rated as medium. Acceptable developmental studies were carried out (via
inhalation route) in two species, rats and rabbits, with no adverse results noted; however, no
2-year chronic studies are available. As with the RfD database, full-scale multigeneration
reproductive studies are lacking. The critical effect, altered tissue weights, is consistent across
routes of exposure (altered kidney weight was also a critical effect for the RfD).
The use of a partial uncertainty factor for interspecies extrapolation is justified because
species-to-species dosimetric adjustments were made and an HEC was calculated.
An area of scientific uncertainty and controversy in this assessment concerns the renal
lesions in the male rats observed in the principal study. The descriptions of these lesions strongly
suggest the male-specific rat nephropathic response elicited by compounds such as J-limonene
and decalin (U.S. EPA, 199la). This assessment has discounted these histopathological lesions
in establishing an effect level for derivation of the RfC because EPA does not consider such
lesions to be an appropriate endpoint for determining noncancer toxicity. If the male rat renal
effects had not been discounted, then the RfD would have been approximately fivefold lower,
because the NOAEL would be 100 ppm versus 496 ppm. What has been accepted as
lexicologically relevant from the profile of renal toxicity in the principal study is the increase in
female renal weight. Other repeated-dose studies with cumene also have reported increased renal
weights among female rats (Wolf et al, 1956; Monsanto, 1986; Chemical Manufacturer's
Association, 1989). These independent observations, coupled with the uncertainty about the
progression and outcomes of these alterations (because of the absence of any true lifetime
studies) further justifies considering these weight alterations as lexicologically significant.
7.0 REFERENCES
Bushy Run Research Center. 1989a. Cumene fourteen-week vapor inhalation study in rats with
neurotoxicity evaluation (part 1-2) with attached studies and cover letter dated December 7,
1989. TSCATS/0522881; EPA/OTS Doc. No. 40-8992172.
Bushy Run Research Center. 1989b. Developmental toxicity study of inhaled cumene vapor in
CD (Sprague-Dawley) rats. Final project report 52-621. TSCATS/0522881; EPA/OTS Doc.
No. 40-8992172.
Bushy Run Research Center. 1989c. Developmental toxicity study of inhaled cumene vapor in
New Zealand White rabbits. Final project report 52-622. TSCATS/0522881; EPA/OTS Doc.
No. 40-8992172.
Chemical Manufacturers Association. 1989. A two-week pilot inhalation toxicity study of
cumene vapors in rats with attachments and cover letter dated September 7, 1989.
TSCATS/0522867; EPA/OTS, Doc. No. 40-8992168.
20
-------
Curren, R.D. 1992. Unscheduled DNA synthesis in rat primary hepatocytes - test article:
Cumene. Microbiological Associates, Inc., Study No. T4786.380005, May 28, 1987.
Cushman, J.R., J.C. Norris, D.E. Dodd, K.I. Darmer, and C.R. Morris. 1995. Subchronic
inhalation toxicity assessment of cumene in Fischer 344 rats. J. Am. Coll. Toxicol.
14(2): 129-147.
Dourson, M.L. and J.F. Stara. 1983. Regulatory history and experimental support of uncertainty
(safety) factors. Reg. Toxicol. Pharmacol. 3: 224-238.
Fabre, R., R. Truhaut, J. Bernuchon, and F. Loisillier. 1955. Toxicologic studies of solvents to
replace benzene. III. Study of isopropyl benzene or cumene. Arch. Mai. Prof. 16(4): 285-299.
Florin, I, L. Rutberg, M. Curvall, and C.R. Enzell. 1980. Screening of tobacco smoke
constituents for mutagenicity using the Ame's test. Toxicology. 18: 219-232.
Gulf Life Sciences Center. 1985a. CHO/HGPRT test of cumene. Gulf Project No. 84-2128.
Gulf Life Sciences Center. 1985b. Micronucleus test of cumene. Gulf Project No. 84-2129.
EPA/OTS Doc. No. 878216015.
Gulf Oil Corporation. 1984a. TSCA 8(e) submission 8EHQ-11840536 88-8500694. Project
No. 84-2131: Cell transformation test of cumene. Office of Toxic Substances, U.S. EPA,
Washington, DC (also Fiche No. OTS 0509712).
Gulf Oil Corporation. 1984b. TSCA 8(e) submission 8EHQ-11840536 88-8500694. Project
No. 84-2130: Hepatocyte primary culture/DNA repair test of cumene. Office of Toxic
Substances, U.S. EPA, Washington, DC (also Fiche No. OTS 0509712).
Gulf Oil Corporation. 1985. Five-day repeated dose inhalation toxicity study in rats of cumene
with cover letter. TSCATS/0206783; EPA/OTS, Doc. No. 87-8216016.
Hansch, C. and A.J. Leo. 1985. Medchem Project. Issue No. 26. Pomona College, Claremont,
CA.
Hard, G.C., IS. Rodgers, K.P. Baetcke, W.L. Richards, R.E. McGaughy, and L.R. Valcovic.
1993. Hazard evaluation of chemicals that cause accumulation of a2u-globulin, hyaline droplet
nephropathy, and tubular neoplasia in the kidneys of male rats. Environmental Health
Perspectives. 99: 313-349.
ICF Kaiser, Inc. 1990a. THC: A computer program to compute a reference dose from
continuous animal toxicity data using the benchmark dose method. K.S. Crump Division,
Ruston, LA.
21
-------
ICF Kaiser, Inc. 1990b. THWC: A computer program to compute a reference dose from
continuous animal toxicity data using the benchmark dose method. K.S. Crump Division,
Ruston, LA.
Jenkins, L.J., Jr., R.A. Jones, and J. Siegel. 1970. Long-term inhalation screening studies of
benzene, toluene, o-xylene, and cumene on experimental animals. Toxicol. Appl. Pharmacol.
16(3): 818-823.
Kristiansen, U., L. Hansen, G.D. Nielsen, and E. Hoist. 1986. Sensory irritation and pulmonary
irritation of cumene and n-propanol: Mechanisms of receptor activation and desensitization.
Acta Pharmacol. Toxicol. 59: 60-72.
Lawlor, T.E. and Wagner, V.O. 1987. Salmonella/Mammalian-microsome preincubation in
mutagenicity assay (Ames test); test article: Cumene. Microbiological Associates, Inc., Study
No. T4786.502009, March 23, 1987.
Mackay, D. and W.Y. Shui. 1981. A critical review of Henry's Law constants for chemicals of
environmental interest. J. Phys. Chem. Ref Data. 19: 1175-1199.
McKee, R.H., Z.A. Wong, S. Schmitt, P. Beatty, M. Swanson, C.A. Schreiner, and J.L.
Schardein. 1990. The reproductive and developmental toxicity of high flash naphtha. Toxicol.
Ind. Hlth. 6: 441-460.
Mitruka, B.M. and H.M. Rawnsley. 1981. Clinical Biochemical and Hematological Reference
Values in Normal Experimental Animals and Normal Humans, 2nd ed. Masson Publishing,
New York.
Monsanto Company. 1986. One-month study of cumene vapor administered to male and female
Sprague-Dawley rats by inhalation. U.S. EPA/OTS Public Files, 8D submission. Microfiche
No. OTS0513229.
Montgomery, C.A., Jr. and J.C. Seely. 1990. Chapter 10, Kidney, in Pathology of the Fischer
Rat, Reference and Atlas (G.A. Boorman et al, eds.), p. 127-153, Academic Press.
NRC (National Research Council). 1983. Risk Assessment in the Federal Government:
Managing the Process. National Academy Press.
NTP (National Toxicology Program). 1990. Toxicology and carcinogenesis studies of toluene in
F344/N rats and B6C3F1 mice. (Available from National Toxicology Program, NIEHS,
Research Triangle Park, NC.)
NTP (National Toxicology Program). 1996. In-vivo cytogenetics testing results for cumene,
micronucleus induction results. Available from National Toxicology Program, NIEHS, Research
Triangle Park, NC 27709.
22
-------
Putnam, D.L. 1987. Chromosome aberrations in Chinese hamster ovary (CHO) cells - test
article: Cumene. Microbiological Associates, Inc. Study No. T4786.337012, May 12, 1987.
Research Triangle Institute. 1989. Metabolism, disposition and pharmacokinetics of cumene in
F-344 rats following oral, IV administration or nose-only inhalation exposure. Report
No. RTI/4353-01F. CM A Reference No. CU-5.0-PK-RTI.
Sato, A. and T. Nakajima. 1987. Scand. J. Work Environ. Health. 13: 81-93.
Seticzuk, W. and B. Litewka. 1976. Absorption of cumene through the respiratory tract and
excretion of dimethylphenylcarbinol in urine. Br. J. Ind. Med. 33: 100-105.
Sipes, I.G. and AJ. Gandolfi. 1991. Biotransformation of toxicants, in Casarett and Doull's
Toxicology, 4th ed. (C.D Klassen, M.O. Amdur, and J. Doull, eds.), p. 88-126. McGraw-Hill.
Tegeris, J.S. and R.L. Balster. 1994. A comparison of the acute behavioral effects of
alkylbenzenes using a functional observational battery in mice. Fund. Appl. Toxicol.
22: 240-250.
U.S. EPA. 1987a. Risk Assessment Guidelines of 1986 (EPA/600/8-87/045, dated
August 1987).
U.S. EPA. 1987b. Health and Environmental Effects Document on Cumene. Prepared by the
Office of Health and Environmental Assessment, Environmental Criteria and Assessment Office,
Cincinnati, OH, for the Office of Solid Waste and Emergency Response, Washington, DC, dated
August 1987.
U.S. EPA. 1988. Recommendations for and Documentation of Biological Values for Use in
Risk Assessment. EPA 600/6-87/008, NTIS PB88-179874/AS, February 1988.
U.S. EPA. 199la. a2u-globulin: Association with Chemically Induced Renal Toxicity and
Neoplasia in the Rat. EPA/625/3-91/019F, September 1991.
U.S. EPA. 1991b. Memorandum dated November 23, 1991, from Jennifer Seed, Health and
Environmental Review Division, to Gary E. Timm, Chemical Testing Branch, Existing Chemical
Assessment Division, on increased incidence of ecchymosis in a developmental toxicity study of
inhaled cumene vapor in New Zealand White rabbits (TSCATS/0522881; EPA/OTS Doc.
No. 40-8992172, see Bushy Run Research Center, 1989c, this report).
U.S. EPA. 1991c. Guidelines for Developmental Toxicity Risk Assessment, dated December 5,
1991. Fed. Reg. 56, No. 234: 63798-63826.
23
-------
U.S. EPA. 1994a. Peer review and peer involvement at the U.S. Environmental Protection
Agency, signed by U.S. EPA Administrator, Carol M. Browner, dated June 7, 1994.
U.S. EPA. 1994b. Interim Policy for Particle Size and Limit Concentration Issues in Inhalation
Toxicity: Notice of Availability, dated October 26, 1994. Fed. Reg. 59, No. 206: 53799.
U.S. EPA. 1994c. Methods for Derivation of Inhalation Reference Concentrations and
Application of Inhalation Dosimetry, EPA/600/8-90/066F, dated October 1994.
U.S. EPA. 1995a. Guidance on Risk Characterization, memorandum of the Administrator,
Carol Browner, dated March 21, 1995.
U.S. EPA. 1995b. (proposed) Guidelines for Neurotoxicity Risk Assessment, dated October 4,
1995. Fed. Reg. 60(192): 52032-52056.
U.S. EPA. 1995c. Use of the Benchmark Dose Approach in Health Risk Assessment,
EPA/630/R-94/007, dated February 1995.
U.S. EPA. 1996a. (new proposed) Guidelines for Carcinogen Risk Assessment, 1996.
(Currently, these guidelines are available only as a draft.)
U.S. EPA. 1996b. Guidelines for Reproductive Toxicity Risk Assessment dated October 31,
1996. Fed. Reg. 61(212): 56274-56322.
U.S. EPA. 1997. Integrated Risk Information System (IRIS) Online. NCEA, Cincinnati, OH.
Wolf, M.A., V.K. Rowe, D.D. McCollister, R.L. Hollingsworth, and F. Oyen. 1956.
Toxicological studies of certain alkylated benzenes and benzene. Arch. Ind. Health.
14: 387-398.
Yang, L.L. 1987. CHO/HGPRT mutation assay; test article: Cumene. Microbiological
Associates, Inc., Study No. T4786.332010, June 1, 1987.
24
-------
8.0 APPENDIXES
Appendix A: Benchmark Concentration Analyses of Data from Cushman et al. (1995)
(1) Computational Models — Continuous Data
The polynomial mean response regression model (THC, ICF Kaiser, 1990a) and the Weibull
power mean response regression model (THWC; ICF Kaiser, 1990b) were used.
THC F(d) = q0 + SIGN x [qi(d - d0) + ... + qk(d - d0)k]
THWC F(d) = q0 + SIGN x qi(d - d0)q2
where
d = dose,
F(d) = average response at dose d,
q0, ql3 q2, k = estimated parameters (used to determine degrees of freedom), and
SIGN = input indicating an increasing or decreasing dose-response function.
For THC data inputs, the degree of the polynomial was set to the number of dose groups
minus one, the corrected sum of squares for each group = (n - 1) x (standard deviation)2, the
response type was relative [F(d) - F(0)] / F(0), and no threshold was estimated. For THWC, data
inputs were the same, except that the lower limit of q2 was set at 1. Although lower values of
q2 may produce a better fit to the data (i.e., lower SSf), the shapes of dose-response curves
generated from the lower values often lack a reasonable biological motivation.
(2) Data
Group mean absolute organ weights for female kidneys and female and male adrenals listed
in the principal study of Cushman et al. (1995) were modeled.
(3) Model Fit
Model fit was judged by comparison of a test statistic (F') with F distribution at specified
degrees of freedom [dff (F table numerator); dfe (F table denominator)]. When F' equals or
exceeds the appropriate value in the F distribution tables at 0.01, it is concluded that the model
did not fit the data.
F' = (Ssf / dff) / MSe)
where
SSf = sum of squares lack of fit (generated by THC or THWC),
Mse = pooled mean square pure error (generated by THC or THWC),
dff = dose groups minus number of parameters [see (1) above] = 5-3=2 (numerator
25
-------
in F table), and
dfe = degrees of freedom generated by THC or THWC (denominator in F table).
(4) Results
The critical effect most correlated with the cumene database, female kidney weight gain,
was not modeled successfully (F > 0.01, Table 1). Inspection of the modeling results (Table 2)
showed that the model did not predict the dose-response discontinuity in weight observed at the
low dose in the female kidney. A similar discontinuity in dose-response occurred with female
adrenal weights, such that the model did not fit the data at 500 ppm, where there was a decrease,
rather than increase, in weight gain relative to the lower dose.
Organ Weight
Data Modeled
Male adrenal
Female kidney
Female adrenal
THC BMC10
(MLE), ppm
484 (656)
1,067 (1,229)
906 (1,067)
THC Fit (F', F)
0.01 (0.9, 4.8)
>0.01 (5.6, 4.8)
>0.01 (7.1,4.8)
THWC BMC10
(MLE), ppm
484 (656)
1,072 (1,239)
924(1,168)
THWC Fit (F', F)
0.01 (0.9, 4.8)
>0.01 (5.6, 4.8)
>0.01 (6.5, 4.8)
Table 1. BMC10 values and statistical analysis of model fits to weight gain data from
Cushman et al. (1995), where BMC10 is the lower 95% confidence bound on the
concentration of the maximum likelihood estimate (MLE) of a 10% relative
weight change. Fits for both THC and THWC were based on calculations from
ICF Kaiser, Inc. (1990a,b)
(5) Discussion
Rather than rely on results from unsuccessful modeling (F'/F > 0.01) or on results from a
possibly inappropriate endpoint, the NOAEL of 496 ppm is used for further quantitative analysis.
This NOAEL is nearly the same as the BMC 10 of 484 ppm for the only data set that was
modeled successfully, male adrenal weight gain. The critical effect that was most correlated with
the cumene database, female kidney weight gain, was not modeled successfully.
Appendix B: Summary of and Response to External Peer Review Comments
The Toxicological Review for Cumene (except for Sections 4.7 and 6.0, which were
rewritten subsequent to external peer review) and all individual cumene assessments have
undergone both internal peer review performed by scientists within EPA or other Federal
agencies and a more formal external peer review performed by scientists chosen by EPA in
accordance with U.S. EPA (1994a). Comments made by the internal reviewers were addressed
prior to submitting the documents for external peer review and are not part of this
26
-------
Organ Data Modeled
Male adrenal
Female kidney
Female adrenal
Dose
(ppm)
0
50
100
496
1,202
0
50
100
496
1,202
0
50
100
496
1,202
Observed Mean Weight
(g)
0.039
0.041
0.040
0.044
0.047
1.40
1.49
1.41
1.44
1.56
0.047
0.049
0.048
0.043
0.056
Predicted Mean Weight
THC and THWC (g)
0.040
0.040
0.041
0.043
0.047
1.43
1.43
1.43
1.44
1.56
0.048
0.048
0.048
0.048
0.056
Table 2. Benchmark dose modeling of organ weight data from the study of Cushman et
al. (1995). The actual data from the study (Oberved Mean Wt.) is compared
against the results obtained from applying both the THC and THWC models
(Predicted Mean Wt.). Bolded text highlights differences between predicted
and observed values.
appendix. Public comments also were read and carefully considered. The external peer
reviewers were tasked with providing written answers to general questions on the overall
assessment and on chemical- specific questions in areas of scientific controversy or uncertainty.
A summary of comments made by the external reviewers and EPA's response to these comments
follows. All three external peer reviewers (see Contributors and Reviewers) recommended that
this document and the accompanying assessments were acceptable with minor revision.
(1) General Comments
The three external reviewers offered editorial comments and many minor, but valuable
suggestions, all of which have been incorporated into the text to the extent feasible. Substantive
scientific comments are addressed below.
27
-------
A. Comment: The appropriateness of Wolf et al. (1956) as principal study for deriving the oral
RfD
One reviewer states that EPA is forced to rely on the marginal Wolf et al. (1956) study
because of the paucity of other studies that use the oral route, but the use of uncertainty factors
may in part compensate for the deficiencies of this study. Another reviewer states that EPA
appropriately recognizes the severe limitations of the study. This same reviewer suggests the use
of the Cushman et al. (1995) inhalation study as an alternative method for deriving an RfD.
Response to Comment: The proposal that the inhalation study of Cushman et al. study be
used for derivation of the oral RfD has merit. The reviewer notes correctly that cumene is both
readily absorbed and has a similar disposition by both oral and inhalation routes. The EPA,
however, feels that this option is outweighed by the short term (90 days) of the Cushman et al.
inhalation study (the Wolf et al. study lasted 7 mo and is therefore more in concordance with the
intention of the RfD). Therefore, the Wolf et al. (1956) study is retained as the principal study
for the RfD assessment.
B. Comment: The potential for hematotoxicity
One reviewer cautioned that myelotoxicity from several compounds, including benzene,
has been difficult to reproduce in rodents, and the absence of distinct blood effects in the rat
studies does not completely exonerate cumene as a potential myelotoxic agent in humans.
Response to Comment: Minor blood effects noted in the principal, 90-day inhalation study
of Cushman are described in this IRIS file. A comparison of these blood effects to those
observed with benzene is somewhat problematic. The metabolism of benzene is exceedingly
complex. The hematotoxic effects of benzene are thought to be mediated through secondary
metabolites, such as catechol and hydroquinone, that can be involved in peroxidative processes
(Irons, 1991). On the other hand, the metabolism of cumene is simple, the principal metabolite
being a secondary alcohol that has little propensity to be involved in peroxidative processes.
It also should be noted that benzene was tested in the same oral study with cumene (Wolf et al.,
1956), and that blood effects (leucopenia and erythrocytopenia) were reported in rats exposed to
benzene but not in rats exposed to cumene. Moreover, the magnitude of blood effects observed
in the Cushman cumene study are within normal limits for rodents. No changes are proposed to
the IRIS file.
C. Comment: The absence of liver effects in the oral study of Wolf et al. (1956)
One reviewer expressed concern that liver weight changes (along with possible
hepatocellular hypertrophy) were present but not recognized in this investigation.
Response to Comment: The study of Wolf et al. (1956) tested and reported on several
benzenoid compounds, in addition to cumene (isopropylbenzene), and the experimental
procedures state that livers were weighed and examined. Results of the study report changes in
liver weight and pathology and in kidney weights and pathology for ethylbenzene, liver and
kidney weight changes for styrene, but only kidney weight changes for cumene. In light of the
28
-------
experimental description and observed results, it is unlikely that liver alterations by cumene
would have been missed. No change is proposed to the IRIS file to accommodate this comment.
(2) Comments on Chemical-Specific Questions
Question 1. Based on the information noted in the currently designated principal study
(Cushman et al, 1995), is the discounting of the renal effects in males justified?
Comments: One reviewer stated that it was not clear that all kidney changes in the males
could be attributed solely to a2u-globulin, but that EPA had presented a scientific basis for
discounting the male effects. Another reviewer indicated that attributing male renal effects to
a2u-globulin without identification of the specific protein was problematic, although the rationale
presented by EPA for the use of female renal effects was adequate.
The response of the third reviewer was that the organ weight changes in both kidney and
liver observed in both sexes were caused primarily by microsomal enzyme induction, with an
a2u-globulin-like response in males being merely ancillary to the weight increases. This reviewer
notes that a2u-globulin responses do not occur with the structurally related compounds toluene
and benzene.
Response to Comments: In response to the third reviewer, liver and kidney weights are
increased in female and male rats exposed to toluene, and kidneys from male rats exposed to
toluene do not show characteristics of a2u-globulin nephropathy in 14- to 15-week exposures
(NTP, 1990). Cumene, in comparison, shows increased liver and kidney weights (and adrenals)
in female and male rats, and kidneys from male rats exposed to cumene show some
characteristics of a2u-globulin nephropathy (perfusion of uncharacterized hyaline droplets)
in 13-week exposures. Thus, responses of rats to cumene exposure show characteristics both of
toluene exposure and of agents causing male-specific a2u-globulin nephropathy. Because the
relevancy to humans is unclear, EPA policy indicates that male-specific a2u-globulin is not an
appropriate toxicological endpoint for use in dose-response assessments. Due to this policy and
to the inconclusiveness of the information on the identity of cumene as an a2u-globulin agent, the
effects in the male kidney are considered to be confounded and are discounted. This logic is
currently presented in the IRIS file, and no change is proposed.
Question 2. Is sufficient rationale given to let stand the organ weight changes in female rats as a
critical effect?
Comments: One reviewer considers the rationale adequate. Another reviewer approves of
the rationale, while stating that associated renal pathology (none was described in the study)
would be more compelling. The third reviewer states that organ weight changes in both kidney
and liver, observed in both sexes, should be considered as critical effects. The third reviewer
also considered the weight increases observed in kidney and liver of both sexes adverse and
caused by microsomal enzyme induction (as apparently is the case for effects from toluene).
Response to Comments: Liver weight increases were carefully considered as a co-critical
effect. As a matter of policy, liver weight increases without accompanying pathology may be
29
-------
indicative of increased liver metabolic capacity and usually are considered by the EPA to be
adaptive, not adverse, in nature. As pointed out by the reviewer, both liver and kidney weights
are increased in female and male rats exposed to toluene by air or by gavage (NTP, 1990) and
liver weight increases are the basis for the current toluene RfD (U.S. EPA, 1997, IRIS Online).
In the case of toluene, the liver weight changes were considered more adverse in character
because liver damage is a documented sequela of toluene exposure in humans. Nevertheless,
increases in hepatic weight are not considered as an adverse, co-critical effect in the case of
cumene toxicity because no parallel evidence exists for human hepatic damage from cumene
exposure, and because liver weight increases do not appear to be a consistent response in animal
studies. Liver weight increases were not observed in the oral study of cumene by Wolf et al.
(1956). Future evidence in the area of cumene liver toxicity may be sufficient to justify inclusion
of liver weight changes as a critical effect. No changes are proposed to the current IRIS file as a
consequence of this comment.
Question 3. Is the information in the toxicological review sufficient to consider that cumene has
low potential for causing reproductive effects?
Comments: All reviewers considered cumene as an unlikely reproductive toxicant. One
reviewer did not consider cumene as a likely reproductive or developmental toxicant, based on
the toxicological evidence (including analysis of the available studies), the rapid elimination
from the body, and results of studies with similar but unspecified compounds. This same
reviewer recommends that the IRIS file should reflect that not only are multigeneration
reproductive studies lacking, but also there are no data regarding cumene exposure prior to
mating, from conception through implantation, or during late gestation, parturition, or lactation.
Response to Comments: The above statement on specific absence of data is incorporated
into the IRIS file at several locations.
REFERENCES
Irons, R.D. 1991. Blood and bone marrow, in Handbook of Toxicologic Pathology (W.M.
Haschek and C.G. Rousseaux, eds.) pp. 389-420.
NTP. 1990. National Toxicology Program Technical Report Number 371. Toxicology and
carcinogenesis studies of toluene. NIH Publication Number 90-2826. (NCEA CRIB No. 65618).
U.S. EPA. 1997. Integrated Risk Information System (IRIS) Online. NCEA, Cincinnati, OH.
30
------- |