EPA635/R-03/001
&EPA
TOXICOLOGICAL REVIEW
OF
XYLENES
(CAS No. 1330-20-7)
In Support of Summary Information on the Integrated Risk
Information System (IRIS)
January 2003
U.S. Environmental Protection Agency
Washington, D.C.
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use. Note: This document may undergo
revisions in the future. The most up-to-date version will be made available electronically via the
IRIS Home Page at http://www.epa.gov/iris.
11
-------
CONTENTS — TOXICOLOGICAL REVIEW for XYLENES (CAS No. 1330-20-7)
DISCLAIMER ii
FOREWORD v
AUTHORS, CONTRIBUTORS, AND REVIEWERS vi
1. INTRODUCTION 1
2. CHEMICAL AND PHYSICAL INFORMATION
RELEVANT TO ASSESSMENTS 2
3. TOXICOKINETICS RELEVANT TO ASSESSMENTS 4
3.1. ABSORPTION 4
3.2. DISTRIBUTION 7
3.3. METABOLISM 9
3.4. EXCRETION 14
3.5. PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS 16
4. HAZARD IDENTIFICATION 17
4.1. STUDIES IN HUMANS—EPIDEMIOLOGY, CASE REPORTS 17
4.1.1. Cancer Studies 17
4.1.2. Noncancer Studies 19
4.1.2.1. Cohort and Case-Control Studies 19
4.1.2.2. Case Reports 21
4.2. PRECHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS 23
4.2.1. Prechronic 23
4.2.1.1. Prechronic Oral Studies 23
4.2.1.2. Prechronic Inhalation Studies 25
4.2.2. Chronic Studies and Cancer Assays 28
4.2.2.1. Oral Studies 28
4.2.2.2. Inhalation Studies 30
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES 30
4.3.1. Reproductive Studies 30
4.3.1.1. Oral Reproductive Studies 30
4.3.1.2. Inhalation Reproductive Studies 30
4.3.2. Developmental Studies 32
4.3.2.1. Oral Developmental Studies 32
4.3.2.2. Inhalation Developmental Studies 33
4.4. OTHER STUDIES 42
4.4.1. Neurotoxicity Studies 42
4.4.1.1. Prechronic Oral Studies 42
4.4.1.2. Prechronic Inhalation Studies 43
4.4.2. Genotoxicity 47
iii
-------
4.4.3. Comparison of the Toxicity of Individual Xylene Isomers 48
4.5. SYNTHESIS AND EVALUATION OF MAJORNONCANCER EFFECTS AND
MODE OF ACTION—ORAL AND INHALATION 50
4.5.1. Oral Exposure 50
4.5.2. Inhalation Exposure 51
4.6. WEIGHT-OF-EVIDENCE EVALUATION AND CANCER
CHARACTERIZATION 59
4.7. SUSCEPTIBLE POPULATIONS 60
5. DOSE-RESPONSE ASSESSMENTS 61
5.1. ORAL REFERENCE DOSE (RfD) 61
5.1.1. Choice of Principal Study and Critical Effect 61
5.1.2. Methods of Analysis 62
5.1.3. Oral Reference Dose Derivation 63
5.2. INHALATION REFERENCE CONCENTRATION (RfC) 64
5.2.1. Choice of Principal Study and Critical Effect 64
5.2.2. Methods of Analysis 66
5.2.3. Inhalation Reference Concentration Derivation 66
5.2.3.1. Principal RfC Derivation 66
5.2.3.2. PBPK Model Applications 68
5.3 CANCER ASSESSMENT 69
5.3.1. Oral Exposure 70
5.3.2. Inhalation Exposure 70
6. MAJOR CONCLUSIONS IN CHARACTERIZATION OF
HAZARD AND DOSE-RESPONSE 70
6.1. HUMAN HAZARD POTENTIAL 70
6.2. DOSE-RESPONSE 71
6.2.1. Noncancer/Oral 71
6.2.2. Noncancer/Inhalation 72
6.2.3. Cancer/Oral and Inhalation 73
7. REFERENCES 74
APPENDIX A. EXTERNAL PEER REVIEW—SUMMARY OF
COMMENTS AND DISPOSITION 85
APPENDIX B. PBPK MODELS FOR m-XYLENE 95
APPENDIX C. BENCHMARK DOSE ANALYSIS OF WOLFE ET AL. (1988A) 103
IV
-------
FOREWORD
The purpose of this Toxicological Review is to provide scientific support and rationale
for the hazard and dose-response assessment in IRIS pertaining to chronic exposure to xylenes.
It is not intended to be a comprehensive treatise on the chemical or toxicological nature of
xylenes.
In Section 6, EPA has characterized its overall confidence in the quantitative and
qualitative aspects of hazard and dose response. 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 assessment 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 IRIS,
the reader is referred to the IRIS Hotline at 202-566-1676.
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
Chemical Managers/Authors
Chemical Managers
Michael W. Broder, Ph.D.
Office of Research and Development
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Lynn Flowers, Ph.D., DABT
Office of Research and Development
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Authors
Lynn Flowers, Ph.D, DABT
Office of Research and Development
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Michael W. Broder, Ph.D.
Office of Research and Development
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Peter McClure, Ph.D., DABT
Syracuse Research Corporation
North Syracuse, NY
Mark Osier, Ph.D.
Syracuse Research Corporation
North Syracuse, NY
VI
-------
Claudia Troxel, Ph.D.
Oak Ridge National Laboratory
Oak Ridge, TN
Reviewers
This document and summary information on IRIS have received peer review both by
EPA scientists and by independent scientists external to EPA. Subsequent to external review
and incorporation of comments, this assessment has undergone an Agency-wide review process
whereby the IRIS Program Director has achieved a consensus approval among the Office of
Research and Development; Office of Air and Radiation; Office of Prevention, Pesticides, and
Toxic Substances; Office of Solid Waste and Emergency Response; Office of Water; Office of
Policy, Economics, and Innovation; Office of Children's Health Protection; Office of
Environmental Information; and the Regional Offices.
Internal EPA Reviewers
Elaina Kenyon, Ph.D.
National Health and Environmental Effects Research Laboratory
Bruce Rodan, MD
National Center for Environmental Assessment
Kevin Crofton, Ph.D.
National Health and Environmental Effects Research Laboratory
External Peer Reviewers
Robert A. Howd, Ph.D.
California Environmental Protection Agency
Sacramento, CA
Patricia McGinnis, Ph.D.
Syracuse Research Corporation
Ft. Washington, PA
Bonnie R. Stern, Ph.D., M.P.H.
BRC Associates
Washington, DC
Summaries of the external peer reviewers' comments and the disposition of their
recommendations are in Appendix A.
vn
-------
1. INTRODUCTION
This document presents background and justification for the hazard and dose-response
assessment summaries in EPA's Integrated Risk Information System (IRIS). IRIS summaries
may include an oral reference dose (RfD), inhalation reference concentration (RfC), and a
carcinogenicity assessment.
The RfD and RfC provide quantitative information for noncancer dose-response
assessments. The RfD is based on the assumption that thresholds generally exist for noncancer
effects. It is expressed in units of mg/kg-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 noncancer effects during a lifetime. The inhalation RfC is analogous to the oral RfD,
but provides a continuous inhalation exposure estimate. The inhalation RfC considers toxic
effects for both the respiratory system (portal-of-entry) and for effects peripheral to the
respiratory system (extrarespiratory or system effects). It is generally expressed in units of
mg/m3.
The carcinogenicity assessment provides information on the carcinogenic hazard
potential of the substance in question and quantitative estimates of risk from oral exposure and
inhalation exposure. The information includes 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 \igfL drinking
water or risk per |_ig/m3 air breathed. Another 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 identification and dose-response assessments for xylenes
has followed the general guidelines for risk assessment as set forth by the National Research
Council (NRC, 1983). EPA guidelines that were used in the development of this assessment
may include the following: Guidelines for the Health Risk Assessment of Chemical Mixtures
(U.S. EPA, 1986a), Guidelines for Mutagenicity Risk Assessment (U.S. EPA, 1986b), Guidelines
for Developmental Toxicity Risk Assessment (U.S. EPA, 1991), Guidelines for Reproductive
Toxicity Risk Assessment (U.S. EPA, 1996), Guidelines for Neurotoxicity Risk Assessment (U.S.
EPA, 1998a), Draft Revised Guidelines for Carcinogen Risk Assessment ((3.$. EPA, 1999),
Recommendations for and Documentation of Biological Values for Use in Risk Assessment (U.S.
EPA, 1988), (proposed) Interim Policy for Particle Size and Limit Concentration Issues in
Inhalation Toxicity (U.S. EPA, 1994a), Methods for Derivation of Inhalation Reference
Concentrations and Application of Inhalation Dosimetry (U.S. EPA, 1994b), Use of the
Benchmark Dose Approach in Health Risk Assessment (U.S. EPA, 1995), Science Policy Council
Handbook: Peer Review (U.S. EPA, 1998b, 2000a), Science Policy Council Handbook: Risk
Characterization (U.S. EPA, 2000b), Benchmark Dose Technical Guidance Document (U.S.
1
-------
EPA, 2000c) and Supplementary Guidance for Conducting Health Risk Assessment of Chemical
Mixtures (U.S. EPA 2000d).
The initial literature search strategy employed for this compound was based on the
CASRN and the common name for individual isomers as well as the mixture. The large number
of citations for the CASRN and common name necessitated a refinement of the search strategy
that involved identifying older research from reviews and chemical assessments combined with a
thorough review of the recent publications. The following data bases were searched: TOXLINE
(all subfiles), MEDLINE, CANCERLIT, TOXNET [HSDB, IRIS, CCRIS, EMIC (1991-2002),
and GENE-TOX], and RTECS, in conjunction with a comprehensive DIALOG search. Any
pertinent scientific information submitted by the public to the IRIS Submission Desk was also
considered in the development of this document.
2. CHEMICAL AND PHYSICAL INFORMATION
RELEVANT TO ASSESSMENTS
Commercial or mixed xylenes are composed of three isomers: meto-xylene (m-xylene),
ortho-xylene (o-xylene), and/>ara-xylene (p-xylene), of which the m-isomer usually
predominates (44-70% of the mixture) (Fishbein, 1988; ATSDR, 1995). The exact composition
of the isomers commonly depends on the source. Ethylbenzene is commonly present in mixed
xylenes; in fact, the technical product contains approximately 40% m-xylene and approximately
20% each of o-xylene, p-xylene, and ethylbenzene (Fishbein, 1988). Thus, most of the
environmental and occupational exposures and toxicological studies are conducted on this
mixture of xylenes containing ethylbenzene. Other minor contaminants of xylenes include
toluene and C9 aromatic fractions. Some physicochemical data for xylenes are shown in Table 1.
Mixed xylenes are used in the production of the individual isomers or ethylbenzene, as a
solvent, in paints and coatings, or as a blend in gasoline (Fishbein, 1988; ATSDR, 1995). The
annual U.S. manufacturers' production capacity of mixed xylenes has been estimated to be 13.1
billion pounds, based on maximum plant production volumes (ATSDR, 1995). The annual U.S.
production of xylenes for 1990-1991 has been estimated at about 6-12 billion pounds for mixed
xylenes, 900 million pounds for o-xylene, 5-8 billion pounds for p-xylene, and 169 million
pounds for m-xylene (ATSDR, 1995). The nonconfidential U.S. aggregate production volumes
for 1998, based on industry submissions to U.S. EPA, are: mixed xylenes >1 billion pounds, m-
xylene >100-500 million pounds, o-xylene >1 billion pounds, and p-xylene >1 billion pounds
(U.S. EPA, 1998c).
-------
Table 1. Physicochemical data for xylenes
Parameter
Synonyms
CAS registry no.
Chemical formula
Molecular weight
Physical state
Vapor pressure at 20°C
Density
Melting point
Boiling point
Solubility in water
LogKow
Conversion factors in air
Odor threshold in air
(absolute)
Value
dimethylbenzene (1,2-; 1,3-; or 1,4-); xylol (mixture),
m-, o-, or p-xylene (isomers); methyltoluene
1330-20-7 mixture
108-38-3 m-isomer
95-47-6 o-isomer
106-42-3 p-isomer
C8H10
106.17
liquid
6-16 mmHg mixture
5-6.5 mmHg individual isomers
0.864 g/cm3 mixture
0.8642 g/cm3 m-isomer
0.8801 g/cm3 o-isomer
0.8611 g/cm3 p-isomer
No data for mixture
-47.4°C m-isomer
-25°C o-isomer
13-14°C p-isomer
137-140°C mixture
139°C m-isomer
144.4°C o-isomer
138. 37°C p-isomer
130 mg/L mixture
134-146 mg/L m-isomer
178-213 mg/L o-isomer
185-198 mg/L p-isomer
3. 12-3. 20 mixture
3.20 m-isomer
3.12, 2. 77 o-isomer
3.15 p-isomer
1 ppm = 4.34 mg/m3
1 mg/m3 = 0.23 ppm
1.0 ppm mixture
3. 7 ppm m-isomer
0.08-0.17 ppm o-isomer
0.47 ppm p-isomer
Reference
Budavarietal. (1996);
ACGIH (1991)
Budavarietal. (1996)
Budavarietal. (1996)
Budavarietal. (1996)
ATSDR (1995)
ATSDR (1995)
Budavarietal. (1996)
Budavarietal. (1996);
ATSDR (1995)
ATSDR (1995)
ATSDR (1995)
NRC (1984)
ATSDR (1995)
-------
3. TOXICOKINETICS RELEVANT TO ASSESSMENTS
The available studies indicate that xylenes are rapidly absorbed following both inhalation
and oral exposure. Following absorption, considerable metabolism occurs, with the liver being
the primary site of metabolism. Xylenes are distributed throughout the body, but show the
greatest affinity for lipid-rich tissues such as adipose tissue or the brain. Elimination is rapid and
occurs primarily in the urine, with the predominant form being the glycine conjugate of
methylbenzoic acid (methylhippuric acid). In humans exposed by inhalation, the loss of xylene
from the blood has been shown to follow biphasic, first-order kinetics with half-lives of about
0.5-1 hour and 20-30 hours.
3.1. ABSORPTION
The rapid absorption of xylenes into the body is demonstrated by numerous experimental
observations. Sato and Nakajima (1979) conducted studies to determine the partition
coefficients of each of the three isomers of xylene. The blood/air and oil/blood partition
coefficients are useful as surrogates for assessing the relative solubility of the respective
chemical for movement into the blood from inhaled air and movement from the blood into tissue.
The study authors employed olive oil and blood (source not provided). The blood/air partition
coefficient for the three isomers ranged from 26.4 to 37.6, and the oil/blood partition coefficient
ranged from 98 to 146. These data indicate that xylene entering the body would be readily
absorbed into the blood and would be expected to move from the blood into tissues in which
neutral lipids predominate, such as adipose tissue. The low water solubility of xylenes suggests
that xylene would move into portions of tissues that have a high lipid fraction. Accordingly, low
water solubility may also serve as a boundary by impeding movement of xylenes from the
gaseous phase into tissues that contain a liquid coating, most notably pulmonary tissue.
Several studies have been conducted on the uptake of xylenes by inhalation. Ogata et al.
(1970) exposed human volunteers to either m- or p-xylene in an exposure chamber for either 3
hours or 7 hours with a 1-hour break at midday. The authors provided no information on the
level of activity of the subjects (i.e., whether the subjects were sedentary or physically active).
During the last 2 hours of exposure the researchers determined that the retention of xylene in
lungs was 87%. This value is considerably higher than values found in other studies and may
reflect the quality of instrumentation used in this study. The authors noted that their methods
were not state of the art, but they were used to maintain consistency with earlier studies.
Sedivec and Flek (1976a) made direct measurements of the level of absorption of xylenes
by subjecting volunteers to 200 or 400 mg/m3 (46 or 92 ppm) of either individual isomers or
mixtures of the three isomers of xylenes vapor for 8 hours without interruption and measuring
the difference in the concentration of xylenes in the inspired air relative to the amount expired.
The amount of the individual isomers absorbed over time was consistent for all three isomers and
ranged from 62.4 to 64.2% of the inhaled volume, reflecting a high solubility of xylenes in
blood.
-------
Riihimaki and Savolainen (1980) conducted studies on human subjects both at rest and
during exercise to measure the kinetics resulting from exposure to mixed xylenes. Healthy male
subjects were exposed to xylene for 5 days, 6 hours per day with a 1-hour break at midday and
then for an additional 1 to 3 days after a 2-day weekend break. The exposure scenarios included
either constant exposure to 100 or 200 ppm or fluctuating exposure with peaks of 200 or 400
ppm that lasted for 10 minutes. The subjects were either sedentary or exercising on a stationary
bike for short periods of time.
Regardless of the exposure scenario (constant or fluctuating or with different xylene
concentrations), retention consistently remained around 60% (i.e., 60% of the inhaled xylene was
retained in the blood and 40% was expired). The results indicate that partitioning of xylene
between the tissues and the air occurs, but it is limited by the solubility in the tissue lipids and
the rate of passive diffusion through the matrix. Overall, the lowest uptake rate was noted with
100 ppm exposure during sedentary conditions (22 |amol/min) and the highest uptake was seen
with the fluctuating concentrations in which the peaks reached 400 ppm during exercise (266
|j,mol/min). Given the constant retention values, the two factors that appeared to control the total
uptake of xylenes were the ambient concentration of xylene and ventilation rates of the subjects.
Astrand et al. (1978) subjected volunteers to vapors of mixed xylenes for four periods of
30 minutes each. Volunteers in the first group were exposed to 870 mg/m3 (200 ppm) vapors for
30 minutes at rest and 90 minutes during light exercise that required 30% of the subjects'
maximal oxygen uptake. The second group was exposed to 435 mg/m3 (100 ppm) with no
activity for 30 minutes followed by three successive 30-minute step intervals of increasingly
demanding workload that required up to 50% of the subjects' maximal oxygen uptake. The
authors monitored the amount of xylene in the inspired and expired air and in the arterial and
venous blood to measure the uptake of xylenes into the blood.
The amount of xylenes taken up in the group with continuous light exercise was constant
over the 2-hour period, with about 65% of the inspired xylenes absorbed by the body. With
regard to increasing workload, retention started at about 65% but dropped to 50% at the higher
workloads and the corresponding increase with ventilation rate. Over the 2-hour period, the
volunteers subjected to 870 mg/m3 and the light workload absorbed 1.4 g of the mixed xylenes
and the group exposed to 435 mg/m3 and the increasingly demanding workload absorbed about
1.0 g. The rate of absorption of xylenes in the first group remained constant over the final 90
minutes, indicating that for a 2-hour exposure, equilibrium between the blood and air had not
been reached. The fact that the constant absorption rate was 64-65% for the first group and the
rate never dropped below 50% for the second group indicates both the high affinity of xylene for
blood and the rapid metabolism of xylene in the body. The lower retention observed in the
second group reflects the ventilation rate. However, the authors estimated that the two groups
inhaled a total of 2.2 and 1.7 g xylene and retained 1.4 and 1.0 g, respectively.
Senczuk and Orlowski (1978) conducted three measurements each on 10 healthy
volunteers (5 men and 5 women) between the ages of 17 and 33 years (for a total of 30
experiments). The individuals were exposed to m-xylene vapor in an inhalation chamber at three
concentrations (100, 300, and 600 mg/m3) for 8 hours with two half-hour breaks. The
-------
investigators monitored the concentration of m-xylene in the vapor and expired air. Xylene
retention and urinary levels of m-methylhippuric acid were also measured. The investigators
found that the retention of m-xylene in the lung varied with the concentration of m-xylene and
duration of exposure. At 300 mg/m3, retention decreased from 83% at the start of the study to
67% at the end of the exposure period (mean, 75%). At 600 mg/m3, retention decreased from
78% at the start of the study to 65% (mean, 71%) at the end of the exposure period. At
100 mg/m3 there was relatively little change in retention rate: 87% at the start and 84% at the end
of exposure. The time periods at which the retention measurements occurred were not specified.
The total amount of m-xylene absorbed was 272, 724, and 1359 mg for women and 342, 909,
and 1712 mg for men at exposures of 100, 300, and 600 mg/m3, respectively.
David et al. (1979) conducted comparative studies on the uptake and metabolism of
m-xylene by inhalation in humans and rats. The objective of the study was to evaluate the
effects of induction of metabolizing enzymes on the ability of the body to clear m-xylenes at
different concentrations. The human component involved five healthy volunteers between the
ages of 46 and 55 years who, during one stage of the study, were exposed to m-xylene without
pretreatment with phenobarbital and during another stage with the equivalent of 2 mg/kg-day
phenobarbital for 11 days prior to treatment with m-xylene. The subjects were exposed to 400
mg/m3 of m-xylene for 8 hours in a chamber (no description of the level of activity or ventilation
rates of the subjects was provided). The retention rates for the controls and phenobarbital-
treated subjects averaged 58% and 59%, respectively, over the course of exposure with no
difference between the morning and afternoon exposure periods, indicating that equilibrium is
not reached at these exposure concentrations and times.
Animal studies (Turkall et al., 1992; Kaneko et al., 1995) have demonstrated rapid and
extensive uptake via the oral route through the detection of parent material in blood or
metabolites in the urine.
Following oral administration of 0.081 or 0.81 mmol/kg (8.6 or 86.4 mg/kg) m-xylene in
corn oil to male Wistar rats, Kaneko et al. (1995) found that blood concentrations of m-xylene
rapidly increased within 4-6 hours and declined thereafter, indicating rapid absorption by the
gastrointestinal tract. Peak concentrations were about 2.5 |j,M for the low dose and about 55 |j,M
for the high dose. Cumulative amounts of the metabolite m-methylhippuric acid in urine
increased through about 12 hours after dose administration and reached an apparent plateau
thereafter. Final cumulative amounts of m-methylhippuric acid in the urine (through 48 hours)
were about 22 |_imoles for the low dose and 160 |_imoles for the high dose. The authors noted
that because m-xylene is subject to metabolism in the liver and has a high blood/air partition
coefficient, more than 90% of the dose would be expected to be excreted as metabolites in the
urine and less than 10% excreted unchanged in expired air.
To examine the influence of soil adsorption on oral absorption, distribution, and
elimination of m-xylene, groups of six male or six female Sprague-Dawley rats were orally
administered doses of 14C-ring-labeled m-xylene in gum acacia alone or with samples of a sandy
soil or a clay soil (Turkall et al., 1992). Each dose contained 150 |^L m-xylene (about 130 mg,
-------
assuming a density of 0.864 g/mL). Radioactivity was measured in samples of blood (plasma),
feces, urine, and expired air collected at several intervals up to 48 hours after dosing. Five rats
of each sex were sacrificed at 24 hours after dosing, and levels of radioactivity were determined
in fat, stomach, pancreas, and skin.
Absorption from the gastrointestinal tract was rapid for both sexes in all treatment
groups; maximum peak plasma concentrations of radioactivity occurred within 20 minutes after
dosing. From the time course of plasma radioactivity, absorption and elimination half-times
were calculated. In the xylene-alone treatment groups, the mean absorption half-time in females
(0.31 hours) was statistically significantly shorter than that in male rats (0.64 hours), whereas the
elimination half-time was longer in females (11.42 hours) than in males (6.77 hours). These
results suggest that absorption may be faster and elimination slower in female rats than in male
rats. Both soil treatments resulted in statistically significantly increased absorption half-times
over xylene alone in female rats, but no effect of soil treatment on this variable was apparent in
male rats. Female rats treated with the sandy soil showed a significantly increased area under
the curve (AUC) for plasma-concentration-time as compared with AUCs for other female rat
groups, indicating that this soil may have increased bioavailability of xylene in females.
No statistically significant differences were observed between males and females in the
amount of radioactivity in the urine collected over 48 hours. Mean levels of radioactivity
expressed as a percentage of the administered dose were 73.7 ± 4.9, 73.2 ± 16.3, and 78.2 ± 0.6
for the xylene alone, xylene and sandy soil, and xylene and clay soil female groups, respectively,
and 96.2 ± 8.8, 83.3 ± 2.9, and 79.4 ±3.3 for the male groups. These results suggest extensive
absorption by the gastrointestinal tract in both sexes ranging from about 74 to 96% of the
administered dose of m-xylene. Radioactivity as parent compound in expired air was a
secondary route of elimination, representing about 14-22% of the administered dose in females
and 9-20% in males. Tissue concentrations of radioactivity at 24 hours were highest in the fat
for both sexes in all treatments.
3.2. DISTRIBUTION
The Kow of xylene indicates that xylene is expected to partition primarily into tissues
containing a higher proportion of neutral lipids, such as adipose, liver, and brain tissue.
Kumarathasan et al. (1998) conducted studies with organs and blood from Sprague-Dawley rats
to determine tissue/blood partition coefficients (Kd) for brain, muscle, kidney, liver, and fat. The
rats were sacrificed and the organs of interest were removed. The tissues were trimmed to
remove extraneous tissue and to achieve uniform size, after which they were spiked with varying
concentrations of a mixture of ethylbenzene and o-, m-, and p-xylenes. Following treatment, the
tissues were placed into vials that were sealed and allowed to equilibrate for either 1 or 2 days.
The concentration of the ethylbenzene and individual isomers of xylene in the head space was
determined by gas chromatography. The Kd for xylene in brain, muscle, and kidney were
comparable and ranged from 1.5 to 3.7. The range of Kd includes ranges in the administered
tissue dose for the three isomers. The Kd for liver was slightly higher, ranging from 3.2 to 5.7.
The Kd for fat was the highest in all tissues tested, with values ranging from 37 to 67; one value
7
-------
of 26 appeared to be an outlier. There were no differences across administered tissue doses or
between isomers.
In their study on the uptake and distribution of ethylbenzene and xylenes, Riihimaki and
Savolainen (1980) found that 10-20% of the xylene was distributed to the adipose tissue.
Adipose has the highest concentration of neutral fat and the highest affinity for xylene of all
tissues. Therefore, once sequestered in adipose tissue, xylene is expected to have the lowest rate
of metabolism, the slowest movement to blood, and the longest persistence in the body. The
concentration of xylene in gluteal subcutaneous fat was about 10-fold higher than in venous
blood following the last day of exposure (5 days exposure + weekend without exposure + 1 day
of exposure).
Astrand et al. (1978), in their study discussed above, found that following rapid uptake of
xylene vapors, concentrations increased in the arterial and venous blood. An initially rapid but
persistent loss of xylenes from the blood followed cessation of exposure. Despite the rapid
absorption of xylenes, the amount found in the blood generally constituted 2-3% of the total
xylenes absorbed. The authors postulate that these factors reflect the high lipid solubility of
xylenes, resulting in the distribution and storage of xylenes, in various tissues. As xylenes are
lost from the blood, residual xylene stored in tissue is eluted into the blood. The higher affinity
of xylene for lipid indicates that the loss of xylene from the tissue is a slow process.
In a study similar to that of Astrand et al. (1978), Engstrom and Bjurstrom (1978)
exposed volunteers to either 870 mg/m3 of xylene vapors during a 30-minute resting period
followed by light exercise or 435 mg/m3 of xylene vapors during a 30-minute resting period
followed by 90 minutes of increasingly strenuous exercise. Adipose tissue was sampled for
xylenes from volunteers at 0.5, 2, 4, and 20-24 hours following conclusion of exposure. The
amount of solvent stored in the body was highly correlated with the amount of body fat. A direct
correlation was found between the amount of xylene taken up and body fat when the two
exposure groups were analyzed together. The mean of the high-exposure group was higher than
that seen with the low-exposure group during the first 4 hours of the study, despite the higher
rate of ventilation in the lower-exposure group. However, at the 20-24 hour sampling, the
amount of xylene in the adipose tissue of the low-exposure group was slightly, but not
significantly (p>0.10), higher than the high-exposure group. The concentration of xylenes in the
adipose tissue was comparable or higher at the 20-24 hour sampling than at the 4-hour sampling.
These data reflect a high absorption of xylenes from the blood into the tissue extending beyond
the exposure period.
Additional information on the distribution of xylenes in the body is available from rodent
studies. Kumarathasan et al. (1997) exposed Sprague-Dawley rats to 1100 ppm m-xylene vapor
in an inhalation chamber for 2 hours. After exposure, the rats were removed from the inhalation
chamber, treated with anaesthetic, and returned to the chamber to avoid loss of xylene. After the
anaesthetic had taken effect, a blood sample was taken from the aorta, the animals were
sacrificed, and the kidney, liver, brain, and fat were harvested. Head-space samples were taken
and analyzed by gas chromatography for m-xylene after 1 day for fat and 2 days for the other
-------
tissues. On a per-gram-of-tissue basis, brain and kidney had the lowest level of m-xylene,
followed by liver; fat contained a considerably higher level than the other tissues.
Carlsson (1981) determined tissue concentrations of m-xylene in male rats following
inhalation exposure to 48 ppm radiolabeled p-xylene for 1, 2, 4, or 8 hours. The greatest
concentration of xylene equivalents (combined concentration of xylene and its metabolites) was
found in the kidneys immediately following the 4-hour exposure (1080 ± 366 nmol/g tissue),
with the next highest concentration found in the subcutaneous fat. The concentration in the
subcutaneous fat continued to increase, reaching peak concentration following the 8-hour
exposure (270 ± 7 nmol/g tissue). For the remaining tissues, the relative m-xylene
concentrations were ischiadic nerve > blood = liver > lungs > cerebrum = cerebellum = muscles
= spleen. Concentrations of xylene-equivalents in the cerebellum, cerebrum, muscles, spleen,
and lungs paralleled the concentrations of xylenes in the arterial blood throughout the entire
exposure period. The distribution of xylenes in tissues parallels that seen in Kumarathasan et al.
(1998).
Bergman (1983) investigated the distribution of radiolabeled m-xylene in mice using
low-temperature, whole-body autoradiography and found high radioactivity levels in the body
fat, bone marrow, white matter of the brain, spinal cord, spinal nerves, liver, and kidney
immediately following inhalation. High levels of metabolites were present in the blood, liver,
lung, kidney, and adrenal medulla; only the parent compound was found in the body fat, bone
marrow, and white matter of the brain. High levels of metabolites were observed in the kidneys
up to 4 hours, in the liver up to 2 hours, in the bile from 2 to 8 hours, in the nasal mucosa and
bronchi from 2 to 24 hours, and in the adrenal medulla immediately after exposure (with no
detectable levels by 30 minutes). No radioactivity was detected in the body by 48 hours after
exposure. Additionally, the author reported that no metabolites of m-xylene were firmly bound
in the tissues.
3.3. METABOLISM
Proposed metabolic pathways for o-xylene are shown in Figure 1 as a model for all
xylene isomers. The principal metabolic fate involves oxidation of one of the methyl groups to a
methylbenzoic acid derivative via methylbenzyl alcohol and methylbenzaldehyde intermediates
(only methylbenzyl alcohol is shown in Figure 1). The methylbenzoic acid derivative is mostly
conjugated to glycine, producing methylhippuric acid derivatives that can be excreted in urine.
Conjugation to glucuronic acid is a minor pathway. Oxidation of the benzene ring to produce
xylenols (i.e., dimethylphenols) is expected to be a negligible metabolic pathway, based on
-------
-o
CH2OH
COOH
o-methylbenzyl
alcohol
glucuronidation
(trace amounts)*
O
o-methyl
benzole
acid
Oxidation
acid)
conjugation
(95-97%)
glucuronide
conjugate of
o-methyl
benzole acid
CONHCH2COOH
o-xylene
expelled unchanged
o-methyl hippuric acid
(o-toluric acid)
exhalation (3-5%)
urine (trace amounts)
o-xylene
o-xylene used as a model for all isomers of xylene
* significant production of glucuronic derivative under conditions of high levels of administration
Figure 1. Metabolic pathways for xylenes.
Source: Adapted from Ogata et al., 1970; Riihimaki and Savolainen, 1980;
Riihimaki, 1979; Bray et al., 1949; Sedivec and Flek, 1976a,b; Ogata et al.,
1980; Carlsson, 1981; Senszuk and Orlowski, 1978; David et al., 1979.
analysis of urinary metabolites, and is not included in Figure 1. The liver is expected to be the
principal site of metabolism for xylenes.
10
-------
In the study described above, Ogata et al. (1970) exposed human volunteers to differing
concentrations of either m- or p-xylene for varying periods of time. The data demonstrated a
linear relationship between exposure to xylenes expressed either in concentration or amount of
methylhippuric acid excreted in the urine over an 18-hour period. These data also demonstrated
that the rates of excretion for the two isomers were similar. The authors used these data to
demonstrate that methylhippuric acid could serve as a marker of exposure to xylene.
Riihimaki and Savolainen (1980), in their study of inhalation exposure to xylene by
human subjects under sedentary and physically active conditions, found that 95% of the
eliminated xylene was in the form of methylhippuric acid, with the remainder lost as
unmetabolized xylene in expired air. No deposition sites, such as lipid-rich tissues, were studied.
Riihimaki (1979) evaluated the metabolism and excretion of xylene and toluene
derivatives in humans. A volunteer was administered a single dose of 7.4 mmole m-
methylbenzoic acid or 7.8 mmole m-methylhippuric acid. Urine was analyzed for 30 hours
following administration for the presence of metabolites. All of the administered xylene
derivatives appeared in urine as methylhippuric acid, indicating that, under the conditions of this
study, once xylene has been oxidized to methylbenzoic acid, the only route of metabolism was as
the glycine conjugate. The relevance of metabolism to the rate of excretion is discussed below.
Bray et al. (1949) intubated and fed rabbits either o-, m-, or p-xylene or the
corresponding toluamide or toluic acid (methylbenzoic acid) and analyzed for metabolites in the
urine. The authors relied on rudimentary colorimetric methods. Higher levels of the glucuronide
metabolite were found in the urine of rabbits that were force-fed the methylbenzoic acid
metabolite and relatively small amounts of the glucuronide following force-feeding with xylene.
From these data, the authors concluded that xylenes are metabolized first to methylbenzoic acid,
which is subsequently conjugated with glycine; the rate-limiting step of this process is the
conversion of xylene to methylbenzoic acid. In the absence of sufficient amounts of glycine, as
with the bolus administration of methylbenzoic acid, the acid reacts with other possible
reactants. The authors noted that there was a difference in the metabolism of o-xylene compared
with the other two isomers that may have been related to unexplained differences in the
excretion of the chemical. Nonetheless, with bolus administration of the xylenes, most of the
chemical was excreted as either the acid or glycine conjugate (i.e., methylhippuric acid).
Sedivec and Fleck (1976a) measured the production of conjugates of methylbenzoic acid
in the urine of subjects exposed to 200 and 400 mg/m3 of individual isomers of xylene vapor. In
contrast with the findings of Bray et al. (1949), all of the methylbenzoic acid derivatives in the
urine were in the form of glycine conjugates (hippuric acid and methylhippuric acid), with no
evidence of glucuronic conjugates. The lack of glucuronide conjugation was attributed to the
method of dosing: Bray et al. used an oral bolus administration of 0.6 g/kg as compared with
their estimated administration by Sedivec and Fleck of 0.019 g/kg by the inhalation route. The
metabolic differences may be species-specific (rabbits and humans), with dosing as the decisive
factor. In addition, xylenols were formed at considerably lower concentrations than were the
methylhippuric acids.
11
-------
A similar finding was reported by Ogata et al. (1980). The researchers conducted studies
under three different scenarios: intraperitoneal (i.p.) injection of xylenes to rats (11.3 mmol/kg),
oral administration in human volunteers (0.368 and 0.736 mmol/kg), and inhalation exposure by
human volunteers (138 ppm for 3 hours) followed by analysis of urine over time. The results
demonstrated that at the highest doses (e.g., rats at 11.3 mmol/kg i.p. xylene), the relative
amount of methylbenzoic acid glucuronide was the highest and the glycine conjugate,
methylhippuric acid, was the lowest. The inhalation exposure demonstrated the highest relative
amounts of methylhippuric acid and almost no detectable methylbenzoic acid glucuronide. One
explanation for these results may be differences in the toxicokinetics between species, or they
could reflect differences arising from routes of administration, as the rats received i.p. injections,
which may have led to saturation of enzyme systems or depletion of glycine stores. The lowest
dose was through inhalation, which occurred over 3 hours, as compared with bolus
administration by i.p. or oral ingestion. The administration by i.p. injection of o-xylene in rats
reflected the highest amount and proportion of unmetabolized xylene in the urine of the three
scenarios. Although these data may reflect species specificity, the differences in the dosing
levels and routes of administration appear to place different burdens on the respective enzyme
systems.
In a study noted above (Senczuk and Orlowski, 1978), five male and five female
volunteers were exposed to three concentrations of m-xylene vapor. The study was designed to
develop a method for quantifying the exposure of individuals to m-xylene using methylhippuric
acid as an indicator; therefore, no measurement was made of other metabolites of m-xylene nor
was there a measurement of the amount of xylene expired following cessation of exposure. The
study found that about 90% of the xylene absorbed was converted to methylhippuric acid.
Astrand et al. (1978) subjected volunteers to concentrations of mixed xylene vapor and
measured the persistence of xylene in the body. One procedure involved exposure to xylenes at
870 mg/m3 while resting for 30 minutes followed by 90 minutes of light exercise and the other
exposure to 435 mg/m3 while resting followed by 90 minutes of increasingly demanding
exercise. The authors tracked the amount of xylenes in both venous and arterial blood and the
amount expelled through the lungs. They found that following cessation of exposure the amount
of xylenes in the arterial and venous blood decreased rapidly. A total of 5% and 4% of the
xylenes absorbed during exposure while resting followed by light exercise and resting with
variable exercise exposures, respectively, were exhaled as unmetabolized xylenes. The
remainder was assumed to be excreted in the urine.
The authors attributed the rapid removal of xylene from the blood to metabolism of the
xylenes (although they did not measure the production of methylhippuric acid or the amount of
xylene or metabolites in the urine). Despite the rapid initial removal of xylenes, the authors were
able to detect the presence of xylenes 4 to 5 days following exposure. The presence of xylenes
in
12
-------
the blood most likely reflects their high solubility for the lipid component in tissues. As the
xylenes in the blood decrease over time, the equilibrium shifts from movement from blood to
tissues to movement out of the tissues and into the blood, albeit at a slower rate.
In the study mentioned above, Carlsson (1981) measured the distribution of xylene and
its metabolites. During exposure, the highest proportions of the dose present as a metabolite
were found in liver and blood; each contained 60-70% of the total xylene equivalent as
metabolites (mostly methylhippuric acid). Subcutaneous fat had the lowest proportion (20%) of
metabolized xylene. Three hours following cessation of exposure, the concentration of xylene
equivalents in the form of metabolites increased from 50-67% to about 95% in the liver.
However, in the muscle, the proportion of xylene equivalents as metabolites ranged from 61 to
71% during exposure, then subsequently decreased to 40% at 3 hours following exposure. In
subcutaneous fat, the relative concentration of xylene equivalents in the metabolite form
remained constant at about 20%.
Results from rat studies demonstrate that orally administered xylenes are subject to a
first-pass metabolic effect that limits the amount of absorbed parent material reaching the
general circulation (Kaneko et al., 1995). The results, however, do not identify oral dose levels
that saturate hepatic metabolism and overwhelm the first-pass effect. Oral doses of 0.081 or 0.81
mmol/kg (8.6 and 86.4 mg/kg) m-xylene in corn oil were given to groups of five male Wistar
rats that were pretreated for 3 days with 80 mg/kg/day phenobarbital in saline or with saline
alone. Other groups of five rats with or without phenobarbital pretreatment were exposed by
inhalation to 40 or 400 ppm m-xylene for 6 hours. At preselected intervals up to 6 hours after
inhalation exposure and up to 12 hours after oral dose administration, blood samples were
collected from the tail. Blood concentrations of m-xylene were measured by a syringe
equilibration method. Phenobarbital pretreatment was administered to examine the effect of
induction of hepatic CYP2B1 on the amounts of parent material in the blood. The AUCs of the
plot of m-xylene blood concentrations versus time after exposure were taken as an index of the
amount of parent material in the blood.
Pretreatment with phenobarbital did not influence the AUCs for rats exposed to the low
air concentration, but it statistically significantly decreased the AUCs for rats exposed to the
high concentration. At 40 ppm, AUCs (in units of |J,M x hour) were 15.3 ±1.5 and 15.6 ±1.2 for
nonpretreated and pretreated rats; at 400 ppm, respective AUCs were 361 ± 58 and 208 ± 25.
Pretreatment with phenobarbital statistically significantly decreased AUCs for rats exposed to
either oral dose level when compared with nonpretreated rats. At 0.081 mmol/kg, AUCs were
14.3 ±1.2 and 4.81 ± 0.23 for nonpretreated and pretreated rats; at 0.81 mg/kg, respective AUCs
were 326 ± 29 and 28.6 ± 1.9. These results demonstrate that first-pass metabolism under an
"induced" state can greatly influence the amount of orally administered m-xylene that reaches
the blood, but they do not identify dose levels at which the first-pass effect is exceeded. For
inhalation exposure, the results show that hepatic enzyme induction influences the amount of
parent material in the blood only at a high air concentration (400 ppm) and not at a low
concentration (40 ppm).
13
-------
3.4. EXCRETION
Results from the study by Riihimaki (1979) (described above) indicate that the rate-
limiting step for the excretion of xylene is the conjugation of methylbenzoic acid with glycine.
A single volunteer was, at different times, administered methylbenzoic acid or its glycine
conjugate, methylhippuric acid. Urine was collected over 30-hour periods, and the identity of
urinary metabolites and the rate of loss (i.e, excretion rate) were determined. Following
methylbenzoic acid administration, only methylhippuric acid was detected in urine. The rate of
loss was greater with the methylhippuric acid treatment than with the methylbenzoic acid
treatment. This study is limited by the fact that it was conducted on a single individual. The
determination that loss of xylene is limited by the availability of glycine suggests that the rate of
utilization of glycine may vary with such factors as age and nutritional status of the individual.
Riihimaki and Savolainen (1980) exposed volunteers to constant levels of 100 ppm or
200 ppm or to fluctuating concentrations with peak concentrations of 200 ppm or 400 ppm
xylene, either while at rest or with intermittent periods of exercise on an ergonomic bicycle. The
uptake of xylene varied with ventilation and exercise. The loss of xylene from the blood
followed biphasic, first-order kinetics, with the initial loss of xylene having a half-life in the
venous blood of 0.5-1 hour, followed by a second phase with a half-life of 20-30 hours. The
authors proposed that the two phases representing the rapid loss of xylene from the blood, mostly
through conversion to methylhippuric acid followed by excretion, indicate that well-perfused
organs reach equilibrium within minutes and muscles reach equilibrium within a few hours,
whereas adipose tissues may require several days of continuous exposure to reach equilibrium.
In a study by Sedivec and Flek (1976a), workers who had been exposed to 200 mg/m3
and 400 mg/m3 xylene vapor for 8 hours were tracked at 10-minute and then 30-minute intervals
for loss of xylenes in expired air and metabolites in urine and, on the second day, at 4- and 8-
hour intervals for loss in expired air. The loss of xylene through the lungs gave a standard
desaturation curve reflecting first-order kinetics, where the amounts of xylene recovered from
the lungs decreased continuously for 24 hours following termination of exposure. Detectable
amounts of unmetabolized xylene were present in expired air on the second day postexposure.
The values given for the amounts of absorbed xylene lost through the lungs were 5.3, 5.8, and
3.5% for the o-, m-, and p- isomers, respectively. The authors considered the last value to reflect
a higher rate of metabolism for the p-isomer relative to the other isomers. When administered as
a mixture of isomers, 4.8% of the absorbed p-xylene was lost through the lungs. The loss of
unmetabolized xylene in urine appeared 2 hours following the beginning of exposure but
remained low throughout the study.
In the aforementioned study by Carlsson (1981), male rats were exposed to 208 mg/m3
14C- p-xylene for 1, 2, 4, and 8 hours and the amount of xylene equivalents (xylene and its
metabolites) were determined. Following rapid absorption of xylene and its broad distribution
throughout the body, there was a rapid loss from the various tissues. The rate of loss of the
parent material varied with concentration and period of exposure. Tissue concentrations
determined 1 to 6 hours after the end of exposure were consistently lower than those recorded in
14
-------
the same tissues immediately after cessation of exposure. The half-life of xylene equivalents in
the subcutaneous fat was estimated to be 2.2 hours following 1-hour exposure and 6.9 hours
following 8-hour exposure. The high concentration of xylene in the fat and the longer
persistence of xylene during the post-exposure period most likely represents its high lipophilicity
and hence its high Kow.
The elimination of xylenes in urine was measured as unmetabolized xylene, toluic acid
(methylbenzoic acid), and toluric acid (methylhippuric acid) (Sedivec and Flek, 1976a).
Analysis of fresh urine did not indicate any toluic acid. However, following storage at ambient
temperatures for "several days," toluic acid appeared in urine from the same sampling. The
authors attributed this to the enzymatic hydrolysis of methylhippuric acid to the acid and the
amine, mediated by microbial contamination. In other studies (Bray et al., 1949; Ogata et al.,
1980), glucuronic acid conjugates have been detected in urine. The authors stated that with their
analytical capabilities they were unable to directly measure the amount of glucuronic acid
conjugates in the urine, but they noted that since the amount of toluic acid in the aged sample
was the same as that of methylhippuric acid in the fresh urine, they presumed that no glucuronic
acid or other conjugates were produced.
Although the differences between this study and others demonstrating the presence of
glucuronic acid conjugates in rabbits following xylene exposure may be due to species-specific
differences in metabolism, the authors stated that the dosing regimen is the decisive factor. In
Bray et al. (1949), the administered dose was 0.6 g/kg; in Sedivec and Flek (1976a) the animals
absorbed 0.019 g/kg xylene, or doses 30-fold lower. Overall, loss of unmetabolized xylene in
urine accounted for less than 0.005% of the amount of xylene retained in the system.
In addition to monitoring the loss of unmetabolized xylene in expired air and urine and
the corresponding acid or glucuronic acid, Sedivec and Flek (1976a) tracked the course of
excretion of methylhippuric acid following inhalation exposure. They detected methylhippuric
acid in urine following 2 hours of exposure; concentrations increased during exposure and
peaked 2 hours following termination of exposure. Following the 2-hour post-exposure
sampling, the amount of methylhippuric acid in urine decreased rapidly, but it was still
detectable in the urine 4-5 days following exposure. When administered individually, there
were no significant differences between the three isomers in the amount of methylhippuric acid
excreted; however, when administered in a 1:1:1 ratio, the authors found that p-toluic acid
derivatives constituted a larger proportion (41.5%) than m-toluic acid (34.7%) or o-toluic acid
(23.8%) in the first 2-hour sampling, which may reflect a preferential oxidation of the p-isomer.
Finally, in addition to the methylhippuric acid derivatives, the authors also recovered a small
amount of xylenols, which appeared in the initial 2-hour sampling but did not increase over time.
Overall, methylhippuric acids accounted for 97.1, 99.2, and 95.1% of the absorbed o-, m-, and p-
isomers, respectively, and xylenols for 0.86, 1.98, and 0.05%, respectively.
Recent studies have used the level of methylhippuric acids in the urine as a quantitative
indicator of xylene. Sedivec and Flek (1976b) measured the amount of toluic acids
(methylhippuric acids) in urine as an indicator of exposure. Human volunteers were exposed to
15
-------
200 or 400 mg/m3 for 4 or 8 hours with a 2-hour break. The amount of excreted metabolites
increased exponentially, reaching a maximum at the end of the exposure and decreasing
exponentially thereafter. The authors noted that the most reliable correlation of exposure with
methylhippuric acid was found by relating the concentration of the metabolite to ventilation rate,
a factor that most likely has a direct relationship with the amount of xylene taken into the body.
David et al. (1979) tracked the loss of m-xylene through urine in the form of
methylhippuric acid in humans exposed to 400 mg/m3 m-xylene. The methylhippuric acid
metabolite first appeared in urine within 2 hours of the start of exposure and reached a peak
between 7 to 8 hours into the study (towards the end of the exposure period). Immediately
following cessation of exposure the concentration of methylhippuric acid decreased
significantly. Levels of methylhippuric acid were low but still detectable for 20 hours following
the start of the study. The authors conducted comparable studies on Wi star-derived male rats.
The treated rats received the equivalent of 50 mg/kg-day for 3 days prior to the start of the study.
The rats were exposed to m-xylene concentrations ranging from 400 to 2000 mg/m3. At
exposure concentrations above 800 mg/m3, rats pretreated with phenobarbital demonstrated an
increased ability to metabolize xylene.
3.5. PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS
Physiologically based pharmacokinetic (PBPK) models for m-xylene inhalation have
been developed for both rats (Tardif et al., 1991; 1992; 1993a; Kaneko et al., 2000) and humans
(Tardif et al., 1993b, 1995; Haddad et al., 1999) (Appendix B). Conceptually, the models consist
of five dynamic tissue compartments representing the lung, adipose tissue, slowly perfused
tissues, richly perfused tissues, and the liver. Inhalation of m-xylene is represented by addition
of m-xylene to the system via the lung component. It should be noted that the models lack an
oral input component; thus, their use in extrapolating toxicity data across these two routes is
precluded.
Concentration in arterial blood is predicted on the basis of the existing venous blood
concentration, the rate of m-xylene exhalation, the inhaled m-xylene concentration, and the
blood/gas partition coefficient. The concentration in each tissue compartment is predicted on the
basis of the existing concentration and the arterial concentration, using appropriate tissue-blood
coefficients. Metabolism is assumed, for the purposes of the models, to occur only in the liver
compartment and is described by a series of equations. The venous concentration is calculated
as a mean concentration, based on the blood flow rates from each compartment and the
concentration of blood leaving each compartment.
Validation of the models following inhalation exposure in both rats (Tardif et al., 1993a,
1997) and humans (Tardif et al., 1995, 1997) has been reported. These models have also been
applied to mixtures containing xylenes and other aromatic solvents (Haddad et al., 1999; Tardif
etal., 1991, 1992, 1993a,b, 1995, 1997).
16
-------
4. HAZARD IDENTIFICATION
4.1. STUDIES IN HUMANS—EPIDEMIOLOGY, CASE REPORTS
On December 14, 2001, after internal peer review of this document, the Agency
articulated its interim policy on the use of third-party studies submitted by regulated entities
(U.S. EPA, 2001). For these purposes, EPA is considering "third party studies" as studies that
have not been conducted or funded by a federal agency pursuant to regulations that protect
human subjects. Under the interim policy, the Agency will not consider or rely on any such
human studies (third-party studies involving deliberate exposure of human subjects when used to
identify or quantify toxic endpoints such as those submitted to establish a NOAEL or NOEL for
systemic toxicity of pesticides) in its regulatory decision making, whether previously or newly
submitted. Some of the supporting studies discussed in this Toxicological Review are third-party
studies; however, the scientific and technical strengths and weaknesses of these studies were
described before this Agency policy was articulated. In addition, the studies cited provide data
which suggests and inform a public health concern for xylenes, but were not designed or used as
principal studies in the derivation of any quantitative value for xylenes based on NOAELs or
LOAELs. The Agency is requesting that the National Academy of Sciences conduct an
expeditious review of the complex scientific and ethical issues posed by EPA's possible use of
third-party studies that intentionally dose human subjects with toxicants to identify or quantify
their effects.
4.1.1. Cancer Studies
Arp et al. (1983) conducted a reanalysis of a cohort of rubber industry workers to study
the relationship between occupational exposure to benzene and other solvents and lymphocytic
leukemia. Worker exposure was reconstructed using company records. The cohort from which
the cases and controls were identified was defined as all active or retired hourly workers 40 to 84
years of age who were alive as of January 1, 1964, with mortality followup through December
31, 1973. Solvent exposures were inferred from groupings of occupational titles and their
associated activities, which were used to estimate the potential for exposures. The authors used
titles and longevity in the position to estimate periods of exposure to specific solvents, but they
authors provide no indication of exposure levels. Solvent composition was inferred from records
of formulations of raw materials. The time periods for which the cohorts were exposed are
important because of the change from coal-based to petroleum-based sources of commercial
aromatic solvents that occurred during the 1940s. The authors defined exposure as cumulative
periods greater than 12 months.
The reporting of the data does not specifically address exposure to petroleum-based
xylenes but does consider exposure to benzene and to secondary solvents, including xylenes.
17
-------
The odds ratios (OD) for lymphocytic leukemia from exposure to solvents other than
benzene, including xylenes, was 4.50 (p=0.08). However, when this value was broken down,
coal-based xylenes produced an OD of 5.50 (p=0.02, 6 cases), compared with all petroleum-
based solvents (principally xylenes) (OD = 1.50, p=OAl, 11 cases). No data were provided for
direct exposure to xylene.
The most pronounced effect noted was the difference between exposure to coal-based
solvents (OD = 6.67,/>=0.01, 8 cases) and exposure to petroleum-based solvents (OD = 1.50,
p=OAl, 11 cases). The source of the solvents (which, as noted above, changed during the 1940s)
indicates the composition of contaminants in the mixture. Petroleum-based solvents typically
have straight-chain aliphatic alkanes, whereas coal-based solvents are expected to have
poly cyclic aromatics as contaminants, several of which (dimethylbenzanthracene,
benzo(a)pyrene and methylcholanthrene) have been shown to cause cancer in laboratory
animals.
Several shortcomings of the study are noted. The small sample size (15 cases and 30
controls) limited the reliability of the observed associations, and chance could not be ruled out as
a possible explanation. Some exposure misclassification was likely because exposures were
inferred from solvent use histories that were of limited precision. Finally, it was noted that there
may not have been a sufficient latent period for the development of tumors in workers exposed
to petroleum-based solvents.
Wilcosky et al. (1984) conducted a reanalysis using the same cohort as Arp et al. (1983).
This study also employed workers of the same age group of 40 to 79 for the same 10-year period,
and exposures were estimated on the basis of occupational titles. Individuals were considered
exposed when they had held a job title that involved direct contact with a chemical for 12
months. However, in the Wilcosky et al. study exposures were broken down into individual
chemicals. The authors found a statistically significantly increased OD of 3.7 (p<0.05, 4 cases)
for lymphosarcoma and a nonstatistically significant increase (OD = 3.3, 4 cases) for lymphatic
leukemia resulting from exposure to xylenes. Despite the findings in Arp et al., no breakout for
exposure to coal-based solvents or petroleum-based solvents was performed. Confounding
issues associated with this study include the finding of a negative correlation between solvent
exposure and lung cancer and the limited number of cases for each.
In brief, although the entire cohort comprised 6678 workers, the number of workers who
were exposed to individual solvents was considerably less (Arp et al., 1983). The authors found
an increased relative-risk estimate (5.5;/><0.02; 6 cases) for lymphocytic leukemia in workers
exposed to coal-based xylenes. The reanalysis by Wilcosky et al. (1984) found an increased OD
for lymphosarcomas (3.7;/><0.05) and a statistically insignificant increase for lymphatic
leukemia (3.3) in workers exposed to xylenes; only 4 cases of each were detected.
Spirtas et al. (1991) conducted a retrospective cohort mortality study of 14,457 workers
who worked for at least 1 year between January 1, 1952, and December 31, 1956, at an aircraft
maintenance facility. The members of the cohort were followed for mortality determinations
until 1982. The study was designed to evaluate health effects arising from exposure to
18
-------
trichloroethylene, although exposures to all solvents were evaluated by surveying the base,
interviewing long-term employees, and looking at historical files. An increased standardized
mortality ratio (SMR) for cancer of the central nervous system (CNS) was observed in male
workers (SMR =1436; 95% CI = 174-5184). No increases in SMRs for multiple myeloma or
non-Hodgkin's lymphoma were observed in male or female employees exposed to xylene.
This study has several limitations: the number of person-years of the workers exposed to
xylene was small (1837 for males and 444 for females), the composition of xylene was not
specified, concentrations to which the workers were exposed were not determined, and the
confounding effect of concurrent exposure of workers to other solvents was not accounted for.
Gerin et al. (1998) conducted a population-based case-control study in Montreal, Canada.
The authors identified sites of "high" incidence of cancer. The study involved the administration
of questionnaires about hospitalized individuals who were being treated for cancer. The
questions involved information about the lifestyles and work habits of the patients. This
information was used to identify potential exposure to benzene, toluene, xylene, and styrene;
exposure was semi quantitatively categorized into low, medium, or high. The researchers used
randomly selected individuals to serve as controls. Although an increased OD was reported for
exposure to "high" concentrations of xylene and cancer of the colon (SMR = 5.8; 8/429 for
cases, compared with 3/955 for controls) and rectum (SMR = 2.7; 5/213 for cases, compared
with 8/937 for controls), the number of cases was small, exposure concentrations were not
defined, the xylene composition was not characterized, and approximately 88% of those exposed
to xylene were also exposed to toluene and benzene. This last point is significant because the
authors note that statistically significant associations were found between each of the four
compounds and rectal cancer.
4.1.2. Noncancer Studies
4.1.2.1. Cohort and Case-Control Studies
Surveys conducted by Uchida et al. (1993) in factories in China identified workers
exposed to solvents in the production of rubber boots or plastic coated wires or in printing work.
The surveys identified 994 solvent-exposed workers. To identify and quantify solvent
exposures, the workers were equipped with a diffusive air sampler for an entire 8-hour working
shift. A total of 175 xylene-exposed workers (107 men, 68 women) for whom the sum of the
three isomers accounted for 70% or more of the total solvent exposure (on a ppm basis) were
selected for the study. The next day these workers underwent a medical examination that
included subjective symptoms, clinical signs, and quantitative health measurements of
hematology, serum biochemistry, and urinalysis. Controls were 241 nonexposed workers from
the same factories or from factories in the same region. Both groups had worked for an average
of 7 years with no change in the workplace during their working life, were of similar ages, and
had comparable drinking rates and smoking habits. Subjective symptoms were evaluated by
means of a survey inquiring about symptoms experienced during the work shift and another
19
-------
survey of symptoms observed outside of work in the previous 3-month period. The prevalence
of the subjective symptoms was calculated as
# of affirmative answers / (# of people in groups x # of questions) x 100%.
With the measurements of the three isomers combined, workers were exposed to a
maximum concentration of 175 ppm xylenes, with a geometric mean of 14 ppm. m-Xylene was
the most prevalent isomer, accounting for approximately 50% of the xylene exposure, followed
by p-xylene (-30%) and o-xylene (~ 15%). Workers were also exposed to ethylbenzene
(geometric mean of 3.4 ppm) and toluene (geometric mean of 1.2 ppm). n-Hexane was rarely
detected, and benzene was never detected. There was little difference between men and women
in the amount of solvent exposure.
The prevalence of subjective symptoms during the work shift and in the previous 3-
month period was significantly higher (p<0.01) in exposed workers when compared with that of
nonexposed workers for both men and women and both sexes combined. During the work shift,
eye and nasal irritation, sore throat, and a floating sensation were increased among exposed
worker of both sexes, and in the previous 3 months, nausea, nightmares, anxiety, forgetfulness,
inability to concentrate, fainting after suddenly standing up, poor appetite, reduced grasping
power, reduced muscle power in extremities, and rough skin were increased in both sexes.
When the exposed individuals were subdivided according to exposure intensity (1-20 ppm or
>21 ppm xylenes), eye irritation, sore throat, and a floating sensation followed a concentration-
related increase for symptoms reported during the work shift, whereas poor appetite was the only
concentration-dependent symptom reported for the previous 3 months. No significant
differences in measured hematology, clinical biochemistry, or urinalysis parameters were noted
in exposed workers when compared with controls. A no-observed-adverse-effects level
(NOAEL) was not identified.
Taskinen et al. (1986) studied pregnancy outcomes in female workers in eight Finnish
pharmaceutical factories between 1973 and 1980. Among 1795 pregnancies there were 1179
deliveries, 142 spontaneous abortions, and 474 induced abortions. Rates of spontaneous
abortions (100 x [# of spontaneous abortions / (sum of number of spontaneous abortions +
number of births)]) were similar among women employed during the first trimester of their
pregnancy (10.9%) and those workers who were not employed during their first trimester
(10.6%). The corresponding rate of spontaneous abortion for all women in the region was 8.5%
for the study period. A statistical comparison of these rates was not performed.
In a case-control study, each of 44 workers who had had a spontaneous abortion (cases)
was matched by age to 3 workers who had given birth (controls). Information about chemical
exposures during work was collected by questionnaire. In a univariate analysis, OD were not
statistically significantly elevated for exposure to solvents (e.g., benzene, xylene, toluene,
aliphatic hydrocarbons) or other chemicals (e.g., estrogens, "carcinogens"). In this analysis, OD
were calculated for individual solvents. Using a logistic analysis of the collected data (which
included estrogen exposure, solvent exposure by frequency of usage, and heavy lifting as
20
-------
variables), a marginally significant OR was found for exposure to >4 solvents (OR = 3.5, 95%
CI = 1.0-12.4), but the ORs for several other measures of frequency of solvent exposure (e.g., 1-
3 solvents, toluene >once a week) were not statistically significantly elevated.
In a later case-control study (Taskinen et al., 1994), 206 women who had had
spontaneous abortions and 329 age-matched referent women who had had normal births were
identified among Finnish female laboratory workers during the period 1970 to 1986. Exposure
information was collected by questionnaire. In a multivariate analysis that included adjustments
for employment, smoking, alcohol consumption, parity, previous miscarriages, failed birth
control, and febrile disease during pregnancy, statistically significant associations were found for
spontaneous abortions and exposure to 3 out of 20 solvents for which questionnaire information
was collected. The elevated ORs were 3.5 (95% CI = 1.1-11.2) for formalin, 4.7 (95% CI =
1.4-15.9) for toluene, and 3.1 (95% CI = 1.3-7.5) for xylene. Each of these ORs was for an
exposure of 3-5 days/week; ORs were not significantly elevated for 1-2 days of exposure per
week. Most of the women who reported having exposure to formalin or xylene worked in
pathology or histology laboratories.
In a case-control study of 36 women who gave birth to children with congenital
malformations (and 105 referent women), no statistically significant associations were found
with exposure to the solvents for which questionnaire information was collected.
The Taskinen et al. (1986, 1994) studies and several others of similar design reviewed by
IARC (1989) are of limited usefulness in assessing the potential reproductive toxicity of xylenes
because the numbers of cases of spontaneous abortions were small and the women had been
exposed to a number of chemicals.
4.1.2.2. Case Reports
Morley et al. (1970) reported the cases of three workmen exposed to approximately
10,000 ppm xylene for 19 hours. One man was dead upon arrival at the hospital. Autopsy
revealed severe pulmonary congestion with focal alveolar hemorrhage and acute pulmonary
edema, hepatic congestion with swelling and vacuolization of many cells in the centrilobular
areas, and microscopic petechial hemorrhages in both the grey and white matter of the brain. In
addition, evidence of axonal neuronal damage was indicated by swelling and loss of Nissl
substance.
Another man was admitted to the hospital unconscious, exhibiting only a slight response
to painful stimuli. He was also hypothermic, had a flushed face, and had peripheral cyanosis.
Medium-grade moist sounds were present in his lungs, and a chest x-ray revealed patchy diffuse
opacity in both lungs. Five hours following treatment with tracheal aspiration and oxygen, the
patient regained consciousness but was amnesic for 2-3 days. Evidence of renal damage was
indicated by an increase in blood urea of 59 mg/100 mL to 204 mg/100 mL 3 days after
admission. Endogenous creatinine clearance was also reduced at this time. Slight hepatic
21
-------
impairment was indicated by a rise in serum transaminase to 100 i.u. over 48 hours, followed by
a return to normal levels.
The third man recovered consciousness following admission to the hospital. He was
confused and amnesic, had slurred speech, and was ataxic upon walking. Within 24 hours of
admission, he was fully conscious and alert, and the ataxia disappeared over 48 hours. There
was no evidence of renal impairment, and mild hepatic impairment was indicated by a slight rise
in serum transaminase (52 i.u.) over 48 hours, followed by a return to normal levels.
Abu-Al-Ragheb et al. (1986) reported on a 27-year-old man who committed suicide by
ingesting xylene. Histopathologic findings included areas of pulmonary edema and congestion.
The probable cause of death was attributed to respiratory failure and asphyxia, a secondary
response elicited by depression in the respiratory center of the brain. In another case (Recchia et
al., 1985), accidental ingestion of xylene resulted in a deep coma lasting more than 26 hours,
hepatic impairment, hematemesis, acute pulmonary edema, and other pulmonary complications.
Another individual who attempted to commit suicide by the intravenous injection of 8 ml of
xylene developed acute pulmonary failure within 10 minutes of administration (Sevcik et al.,
1992). The individual survived following appropriate treatment in the hospital for the
respiratory effects elicited by the xylene.
Two case reports of seizures following exposure to xylene-based products have been
published. Goldie (1960) reported a case where eight painters were exposed to paint that
contained an 80% xylene/20% methylglycolacetate solvent. The workers complained of
headache, vertigo, gastric discomfort, dryness of the throat, and slight drunkenness after 30
minutes of exposure. After 2 months of exposure, an 18-year-old worker exhibited behavior and
symptoms indicative of a convulsive seizure, including weakness, dizziness, inability to speak,
unconsciousness, eye and head rotation to one side, chewing but no foaming, and kicking
motions. The subject recovered consciousness 20 minutes later. Arthur and Curnock (1982)
reported another case in which that an adolescent worker developed major and minor seizures
following the use of a xylene-based glue used for building model airplanes. Neither case report
provided an exposure concentration, and exposures were not limited to xylene alone.
Klaucke et al. (1982) reported that 15 workers who had been exposed by inhalation to
xylenes were admitted to a small community hospital, each complaining of at least two of the
following symptoms: headache, nausea, vomiting, dizziness or vertigo, eye irritation, and nose or
throat irritation. The frequency of the symptoms were headache: 12/15; nausea: 10/15, eye
irritation: 8/15; nose or throat irritation: 7/15; dizziness or vertigo: 7/15; and vomiting: 6/15.
Fourteen of the 15 affected employees noted an unusual odor 15-30 minutes prior to the onset of
symptoms. It was estimated that the workers were exposed to levels as high as 700 ppm.
Five women occupationally exposed to inhaled xylene from 1.5 to 18 years experienced
symptoms that included chronic headache, chest pain, electrocardiogram abnormalities, dyspnea,
cyanosis of the hands, fever, leukopenia, malaise, impaired lung function, decreased ability to
work, complete disability, and mental confusion (Hipolito, 1980).
22
-------
4.2. PRECHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS
4.2.1. Prechronic Studies
4.2.1.1. Prechronic Oral Studies
Groups of 10 male and 10 female Fischer 344 rats were administered mixed xylenes
(60% m-xylene, 13.6% p-xylene, 9.1% o-xylene, and 17.0% ethylbenzene) in corn oil by gavage
at doses of 0, 62.5, 125, 250, 500, or 1000 mg/kg-day for 5 days per week for 13 weeks (NTP,
1986). At termination of the study, necropsy was performed on all animals and comprehensive
histologic examinations were performed on vehicle and high dose-group animals. High-dose
males and females gained 15% and 8% less body weight, respectively, than did controls, with
final body weights being 89% and 97%, respectively, of those of controls (statistical significance
not reported). No signs of toxicity or treatment-related gross or microscopic pathologic lesions
were observed. The lowest-observed-adverse-effect level (LOAEL) is 1000 mg/kg-day, based
on decreased body weights in male rats, and the NOAEL is 500 mg/kg-day.
In the same study, male and female B6C3FJ mice were treated with mixed xylenes.
Groups of 10 mice of each sex were administered 0, 125, 250, 500, 1000, and 2000 mg/kg-day in
corn oil by gavage 5 days per week for 13 weeks. Two female mice in the high-dose group died
prematurely, although gavage error could not be ruled out as the cause. At 2000 mg/kg-day, the
animals exhibited lethargy, short and shallow breathing, unsteadiness, tremors, and paresis
starting at 5-10 minutes after dosing and lasting for 15-60 minutes. Mean body weight of the
mice in the high-dose group was 7% lower than in the vehicle control for males and 17% lower
for females. Although it was not stated explicitly, the report implies that this was a common
finding among the animals dosed at this level. No treatment-related gross or microscopic
pathologic lesions were seen in this study. The NOAEL is 1000 mg/kg-day and the LOAEL is
2000 mg/kg-day for transient signs of nervous system depression.
In a study by Condie et al. (1988), groups of 10 male and 10 female Sprague-Dawley rats
were administered mixed xylenes (17.6% o-xylene, 62.3% m-xylene and p-xylene (which
coeluted); 20% ethylbenzene) by gavage in corn oil for 90 consecutive days at doses of 0, 150,
750, or 1500 mg/kg-day. Effects of exposure included decreased body weights in high-dose
males (94% of controls'); dose-related increased liver weights and liver-to-body weight ratios in
all exposed groups of males (8, 18, and 29% increase in absolute weight above controls in the
low-, mid-, and high-dose animals, respectively) and in mid- and high-dose females (14 and 30%
increases in absolute weight above controls); and increased kidney weights and kidney-to-body
weight ratios in mid- and high-dose males (16 and 19% increase in absolute weight relative to
controls, respectively) and high-dose females (18% increase in absolute weight relative to
controls). The authors postulated that the modest increases in aspartate aminotransferase (AST)
seen in high-dose females and increases in alanine aminotransferase (ALT) in high-dose males
and in mid- and high-dose females, combined with the lack of significant histopathologic
23
-------
findings in the liver suggest that the enlargement of the liver was an adaptation response to
xylene treatment rather than an adverse toxicological effect.
Hematology analysis revealed a mild polycythemia and leukocytosis in the high-dose
males and females in the absence of any observable changes in the health of the rats.
Microscopic evaluation of the kidneys revealed a dose-related increase in hyaline droplet
formation in male rats (0/9, 3/9, 5/10, 8/10, respectively) and a dose-related increase in the early
appearance of minimal chronic nephropathy only in female rats (1/10, 3/10, 6/10, 7/10,
respectively). Compared with controls, the incidence of minimal nephropathy was statistically
significantly elevated (p<0.05) in the 750 and 1500 mg/kg-day female groups but not in the 150
mg/kg-day group (Fisher Exact test performed by Syracuse Research Corporation). The hyaline
droplet formation in male rats was assumed by the authors to be related to male rat-specific a-
2|a-globulin accumulation and not to be relevant to humans. The LOAEL is 750 mg/kg-day,
based on increased kidney weights and early appearance of nephropathy in female rats, and the
NOAEL is 150 mg/kg-day.
In a study by Bowers et al. (1982), aging (12-19-month-old) Long-Evans hooded, male
rats were administered methylated benzenes, including o-xylene, in feed at a concentration of
200 mg/kg feed (10 mg/kg-day) for 1,2, 3, or 6 months, to access ultrastructural changes in the
liver. No other endpoints were evaluated. Although the liver was grossly normal, electron
microscopic evaluation revealed two types of membrane-bound vacuoles in hepatocytes
appearing 1 month after beginning administration of the feed. The appearance and size of the
vacuoles did not change with continued dietary administration of the compound.
In a study by Wolfe (1988a), groups of 20 male and 20 female Sprague-Dawley rats were
administered m-xylene (99% purity) by gavage in corn oil at doses of 0, 100, 200, or 800 mg/kg-
day for 90 consecutive days. Survival incidences were 20/20, 17/20, 15/20, and 18/20,
respectively, for males and 20/20, 20/20, 16/20, and 16/20, respectively, for females. Mortality
in the mid-dose males and mid- and high-dose females attained statistical significance (p<0.05),
but a significant trend was observed only in females. Mottled lungs and a failure of the lungs to
collapse were observed in all mid- and high-dose animals that died early and in 2/3 of the low-
dose males that died early but were not evident in any of the animals that survived to study
termination. Histopathologic examination of the lungs from animals that died before study
termination revealed foreign material in the alveoli in all but one animal. Therefore, these deaths
were attributed to vehicle and/or compound aspiration.
Clinical signs present throughout the study were limited to high levels of salivation prior
to dosing in high-dose males and females. Body weight gains over the entire study period were
decreased (p<0.05) in mid- and high-dose males (89% and 75%, respectively, of controls') and
high-dose females (85% of controls'). Food consumption was likewise decreased (p<0.05) in
high-dose males during weeks 1-5 (90% of control levels) and in mid- and high-dose males
during weeks 6-9 (92% of control levels for both groups). A thorough histologic examination
revealed no other abnormalities. Other effects noted were not definitively related to treatment
24
-------
and/or were not biologically significant. The NOAEL and LOAEL are identified as 200 and 800
mg/kg-day, respectively, based on decreased body weight.
In a second study by Wolfe (1988b), groups of 20 male and 20 female Sprague-Dawley
rats were administered p-xylene (99% purity) by gavage in corn oil at doses of 0, 100, 200, or
800 mg/kg-day for 90 consecutive days. Survival incidences were 20/20, 19/20, 17/20, and
16/20, respectively, for males and 20/20, 18/20, 18/20, and 17/20, respectively, for females.
Mortality in high-dose males attained statistical significance, and a statistically significant trend
was present in the male groups. As in the Wolfe (1988a) study, mottled lungs and/or a failure of
the lungs to collapse was observed in nearly all treated animals that died early but was not
evident in any of the animals that survived to study termination. It was determined that most of
the unscheduled deaths were the result of test material aspiration, as indicated by the presence of
intra-alveolar foreign material in the lungs that was generally associated with pulmonary
congestion.
Treatment-related clinical signs were limited to increased salivation occurring just prior
to dosing that was resolved by 1-hour post dosing in both high-dose males and females. Body
weight gains at 13 weeks were slightly reduced in high-dose males and females (89% of control
levels, not statistically significant), and high-dose females had significantly increased food
consumption for weeks 10-13 (110%). No treatment-related effects were observed in
hematology or clinical chemistry parameters, ophthalmologic examination, or organ weights.
Histopathology revealed no abnormal findings in any tissue or organ. The NOAEL and LOAEL
are identified as 200 and 800 mg/kg-day, respectively, based on early mortality in male rats that
showed signs of test material aspiration into the lungs.
In a nephrotoxicity screening study by Borriston Laboratories, Inc. (1983), groups of 10
male Fischer 344 rats were dosed with 0.5 or 2.0 g/kg m-xylene or 2.0 g/kg saline by gavage for
5 days/week for 4 weeks. No nephrotoxic effects were observed in the rats dosed with m-xylene
when compared with controls.
4.2.1.2. Prechronic Inhalation Studies
In a study by Carpenter et al. (1975), groups of 25 male rats and 4 male beagle dogs were
exposed to air containing measured concentrations of 180, 460, or 810 ppm mixed xylenes
(65.0% m-xylene, 7.8 % p-xylene, 7.6% o-xylene, 19.3% ethylbenzene) for 6 hours per day, 5
days per week, for 65 days (rats) or 66 days (dogs). Endpoints assessed included changes in
body weight, hematology, clinical chemistry, urinalysis, organ weights, and histologic
examination. Three rats from each dose level were sacrificed for histologic evaluation after 15
and 35 days of exposure. Additional measurements in dogs included food consumption and
initial and terminal electrocardiograms.
No treatment-related effects in any of the measured parameters were observed in exposed
rats or dogs when compared with control animals. Additionally, 10 rats per dose, including a
control group handled similarly to the exposed rats and a group of 10 naive control rats, were
25
-------
challenged with a 4-hour exposure to 6700 ppm xylene (29.0 mg/L) at the termination of the
subchronic exposure period. No difference in the median time to death was noted in rats
exposed to xylenes for 65 days when compared with control rats. Therefore, the highest
exposure level in this study (810 ppm) is a NOAEL for changes in body weight, organ weights,
and histopathology in male rats and male beagle dogs.
Tatrai and Ungvary (1980) exposed groups of 30 male CFY rats to 0 or 3500 ppm o-
xylene (purity not stated) for 8 hours per day for 6 weeks. Following exposure, body weights
and organ weights were measured. Organs of 5 rats in each group were examined by light
microscopy and livers were also examined by electron microscopy. Group means and standard
errors were calculated with Student's one- and two-sample t-test. Despite increased food and
water consumption, terminal body weight at 6 weeks in the xylene-exposed group was
statistically significantly lower (p<0.05) than in controls (427.50 ± 7.08 g vs. 454.00 ± 10.40 g).
Exposed rats had hepatic changes, including increased absolute and relative liver weights, signs
of hepatocellular hypertrophy, increased proportion of smooth and rough endoplasmic reticulum,
decreased glycogen, and increased peroxisomes. Measurements of drug metabolizing enzymes
were not made. The observed changes in organ weight were consistent with an adaptive
response to organic chemical exposure and probably reflected induction of enzymes in the liver.
An explanation for the differences in body weight was not provided. The only exposure level in
this study, 3500 ppm o-xylene, is a LOAEL for statistically significant body weight decreases of
about 6% in male rats showing adaptive, but not adverse, changes in the liver.
To investigate the potential for exposure to xylene to induce hepatotoxic effects, Tatrai et
al. (1981) exposed male CFY rats to air containing 0 or 1090 ppm (4750 mg/m3) o-xylene for 8
hours per day, 7 days per week, for 6 or 12 months. The purity of the o-xylene was not
provided. Biochemical indices of liver metabolic capacity (drug-metabolizing enzyme activities)
and liver damage (serum activities of AST and ALT; reported as "GOT" and "GPT") were
measured, as were body weights and organ weights. Liver tissue samples were examined by
light and electron microscopy. Statistical differences between exposed and control group means
were compared by variance analysis and Dunnett's test.
Exposure to 1090 ppm o-xylene for 6 or 12 months resulted in increased food and water
consumption, decreased body weight gain, increased absolute and relative liver weight, and
induction of enzymes of the hepatic mixed-function oxidase system (increased cytochromes P-
450 and b-5, cytochrome c reductase, alanine hydroxylase, and aminopyrene N-demethylase).
Data were presented in graphical form only. At 1 year, the control mean body weight was about
700 g, compared with an exposed mean of about 600 g. This change represents about a 15%
decrease and was reported to have been statistically significant (p<0.05).
Histologic and histochemical examination of the organs, including the liver, were
reported to have shown no pathological alterations or changes in serum activities of AST or
ALT. Electron microscopy of liver tissue samples revealed moderate proliferation of the smooth
endoplasmic reticulum. The hepatic effects are not considered adverse. The exposure level of
26
-------
1090 ppm o-xylene is a LOAEL for decreased body weight in male rats with no adverse changes
in hepatic endpoints.
Ungvary (1990) exposed groups of male CFY rats to air containing 0, 140, 350, or 920
ppm (0,600, 1500, or 4000 mg/m3) xylenes (10% o-xylene, 50% m-xylene, 20% p-xylene, 20%
ethylbenzene) for 8 hours per day, 7 days per week for 6 weeks and then for 5 days per week for
6 months. Endpoints evaluated and statistical evaluations were the same as those in earlier
reported studies from this investigator's laboratory (Tatrai and Ungvary, 1980; Tatrai et al.,
1981). No statistically significant differences in body weights were observed in any of the
exposed groups when compared with the control values.
Statistically significant changes observed in exposed groups at 6 months (compared with
control group values) included increased relative liver weight (17% in the high-dose group only);
hypertrophy of the centrilobular zone of the liver (high-dose group only); increased nuclear
volume of hepatocytes and proliferation of smooth endoplasmic reticulum (only the high-dose
and control groups were examined); increases in the concentrations of cytochrome P-450 and
cytochrome b5 (mid- and high-dose groups); increases in the activities of NADPH, cytochrome
c-reductase, alanine p-hydroxylase, succinate dehydrogenase and aminopyrine N-demethylase
(mid- and high-dose groups); and decreased hexobarbital sleeping time (mid- and high-dose
groups). In general, maximal effects were achieved by 6 weeks of exposure; control levels
returned after a 4-week solvent-free period following the 6-month exposure.
In further experiments reported by Ungvary (1990), continuous inhalation exposure of
CFY rats to 0, 350, 460, or 1150 ppm (0,1500, 2000, or 5000 mg/m3) for 72 hours or repeated
inhalation exposure of male mice, rats, or rabbits to 0 or 575 ppm (2500 mg/m3) for 8 hours per
day for 6 weeks resulted in effects similar to those reported for the repeated exposure study in
male rats for 6 months. Lastly, continuous exposure to 0 or 690 ppm xylenes (3000 mg/m3) for
72 hours in male CFY rats following partial hepatectomy or bile duct ligation still induced the
biotransformation enzymes but did not appear to potentiate the damage induced by these two
interventions. The authors did not consider the observed effects to be adverse and identified the
highest exposure level (920 ppm) to be a NOAEL. However, the liver effects could be
considered biologically relevant, indicating a LOAEL of 920 ppm and a NOAEL of 350 ppm.
Jenkins et al. (1970) conducted inhalation studies on 12 Sprague-Dawley or Long Evans
rats, 15 NMRL(ASH) Princeton-derived guinea pigs, 2 squirrel monkeys, and 2 beagle dogs in
which the animals were repeatedly exposed to 780 ppm o-xylene for 8 hours per day, 5 days per
week, for a total of 30 exposures over a 6-week period. In another portion of the study, 14 rats,
15 guinea pigs, 2 dogs, and 3 monkeys were exposed to 78 ppm o-xylene continuously for
90-127 days; 14 rats, 15 guinea pigs, 10 dogs, and 12 monkeys were exposed continuously for
90-127 days to control air to serve as reference groups. During the 780 ppm study, two rats died
on the third day of exposure, one rat and one monkey died on the seventh day of exposure, and
one of the dogs exhibited tremors throughout the exposure. The cause of death and any clinical
signs occurring before death were not reported. No changes in body weights, hematology
27
-------
parameters, or histopathology in animals exposed to 78 or 780 ppm were reported. A
LOAEL/NOAEL is not determined because of inadequate data reporting.
In a study by Morvai et al. (1976), 16 CFY rats exposed to xylene (composition not
stated) developed respiratory paralysis preceded by atrial fibrillation, bradyarrhythmia, and
asystole. The conditions of the exposure were not clear; exposure was assumed to be 1400 ppm
(6000 mg/m3) for 6 hours per day.
4.2.2. Chronic Studies and Cancer Assays
4.2.2.1. Oral Studies
In National Toxicology Program toxicology and carcinogenesis studies (NTP, 1986),
groups of 50 male and 50 female Fischer 344 rats and 50 male and 50 female B6C3F1 mice were
administered mixed xylenes (60% m-xylene, 13.6% p-xylene, 9.1% o-xylene, 17.0%
ethylbenzene) in corn oil by gavage at doses of 0, 250, or 500 mg/kg-day (rats) and 0, 500, or
1000 mg/kg-day (mice) for 5 days per week for 103 weeks. Necropsy and histologic
examinations were performed on all animals. Tissues were examined for gross lesions and
masses. The tissues examined included mandibular lymph nodes, salivary gland, femur
(including marrow), thyroid gland, parathyroids, small intestine, colon, liver, prostate/testis or
ovaries/uterus, heart, esophagus, stomach, brain, thymus, trachea, pancreas, spleen, skin, lungs
and mainstem bronchi, kidneys, adrenal glands, urinary bladder, pituitary gland, eyes (if grossly
abnormal), and mammary gland. Hematology and clinical chemistry analyses were not
conducted.
In rats, no statistically significantly increased incidences of nonneoplastic or neoplastic
lesions (that would have been expected to be treatment-related) were found in exposed groups
when compared with controls. The authors noted that a survival-adjusted increased incidence of
interstitial cell tumors was found in the high-dose male rat group relative to controls, but they
did not consider this to be a treatment-related increase. The increase was attributed to high-dose
male rats dying between weeks 62 and 92 and it was noted that incidences for these tumors were
comparable with those of controls during other time intervals and that the overall incidences
were not statistically significantly different between control and exposed groups.
Effects of exposure in rats were limited to decreased body weight and decreased survival
in high-dose (500 mg/kg-day) males. Mean body weights were 5-8% lower in high-dose male
rats than in controls from week 59 to week 97, with body weights at 103 weeks being 4% less in
high-dose males than in controls (statistical significance not reported). Male rat survival rates
after 103 weeks showed a dose-related decrease (36/50, 25/50, and 20/50 for control, low-, and
high-dose males, respectively). A life-table trend test for decreased survival incidence with
increasing dose was statistically significant (p=0.033). Pair-wise comparisons with control
survival incidence indicated that only the high-dose male rat incidence was significantly
decreased (p=0.04). A number of the deaths were attributed to gavage error (3/50, 8/50, and
11/50, respectively). The authors did not record observations of rat behavior during dosing.
28
-------
Based on the available observations, the incidence of treatment-related deaths
demonstrated a dose-related increase (11/50 [22%], 17/50 [34%], and 19/50 [38%]). Based on
increased mortality and decreased body weight in male rats, this study identifies 500 mg/kg-day
as a LOAEL and 250 mg/kg-day as a NOAEL. There was no evidence of carcinogenicity in
male or female rats exposed to doses up to 500 mg/kg-day.
In mice, no statistically significantly increased incidences of nonneoplastic or neoplastic
lesions were found in male or female exposed groups when compared with controls. The only
treatment-related effect observed was hyperactivity, which occurred in all high-dose mice of
each sex 5-30 minutes after dosing. This effect was observed consistently beginning at week 4
and continued until study termination at 103 weeks. The LOAEL is 1000 mg/kg-day and the
NOAEL is 500 mg/kg-day for hyperactivity. There was no evidence of carcinogenicity in male
or female mice exposed to doses up to 1000 mg/kg-day.
Maltoni et al. (1983, 1985) exposed groups of 40 male and 40 female Sprague-Dawley
rats to 500 mg/kg mixed xylenes (unspecified proportions) in olive oil orally by gavage 4-5 days
per week for 104 weeks. The control groups of 50 males and 50 females were treated with olive
oil only. The animals were kept under observation until spontaneous death; all rats died by 141
weeks. The proportion of mice that survived treatment was similar in controls and treated
groups through 92 weeks (Maltoni et al., 1983), but survival data for later periods were not
reported (Maltoni et al., 1985). For example, 50% and 65% of exposed males and females,
respectively, survived at 92 weeks, compared with 58% and 66% of control males and females.
Average body weights (standard errors not reported) at several intervals through 92
weeks appeared similar or higher in exposed versus control groups, but statistical significance
was not evaluated. For example, exposed male rats had average body weights of 611.0 g and
626.25 g at 78 and 92 weeks, respectively, compared with control values of 525.92 g and 490.68
g. Exposed females at 78 and 92 weeks had average body weights of 413.48 g and 437.11 g,
compared with control values of 356.5 g and 389.09 g. Counts of red blood cells and white
blood cells from 4-5 rats per group were measured from blood collected at 84 weeks; they did
not appear to be affected by exposure (Maltoni et al., 1983). Mean counts (± standard deviation)
for red blood cells were 9.05 ± 0.13 for males and 7.58 ± 0.61 for females in the exposed group,
compared with respective means of 8.72 ± 0.57 and 7.70 ± 0.50 for controls. Respective total
white blood cell counts were 10.65 ± 0.66 and 9.00 ± 1.82 in exposed rats, compared with 2.92 ±
2.29 and 10.80 ± 2.74 in controls.
Only limited information regarding tumor incidences at specific tissue sites was provided
with no information provided on nonneoplastic lesions or tumor pathology. Final (i.e., 141-
week) tumor incidence data were reported only for rats with hemolymphoreticular neoplasias
(thymomas, others, and total) and for rats with malignant tumors at any site (Maltoni et al.,
1985). Incidences for thymomas were 1/34 and 0/36 in exposed males and females, compared
with 0/45 and 0/49 in controls. Incidences of rats with other hemolymphoreticular neoplasias
(not otherwise specified) were 4/34 and 3/36 in exposed males and females, compared with 3/45
29
-------
and 1/45 in controls.1 Fisher exact tests (performed by Syracuse Research Corporation)
indicated no significant differences between groups in the incidences for hemolymphoreticular
neoplasias (including the combined incidence for thymomas and "others").
The study authors also reported an increase in the total number of exposed rats with
malignant tumors (of unspecified type): 14/38 and 22/40 for exposed males and females,
compared with 11/45 and 10/49 for controls.2 The exposed female total malignant tumor
incidence was statistically significantly increased when compared with controls by the Fisher
Exact test. Because of the incomplete reporting of site-specific tumor incidence data and
pathology, the study by Maltoni et al. (1983, 1985) is of limited use in evaluating the
carcinogenicity of xylenes. For noncancer effects, the study identifies an apparent NOAEL of
500 mg/kg for changes in body weight and counts of red and white blood cells in male and
female rats.
4.2.2.2. Inhalation Studies
No chronic toxicity or cancer studies in animals exposed by inhalation to xylenes are
available.
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES
4.3.1. Reproductive Studies
4.3.1.1. Oral Reproductive Studies
No studies of the reproductive toxicity of xylenes following oral exposure are available.
4.3.1.2. Inhalation Reproductive Studies
In a one-generation reproductive toxicity study (Bio/dynamics Inc., 1983), groups of
male and female CD rats were exposed to 0, 60, 250, or 500 ppm mixed xylenes (groups I, II, III,
and IV, respectively; technical-grade xylene: 2.4% toluene, 12.8% ethylbenzene, 20.3% p-
xylene, 44.2% m-xylene, 20.4% o-xylene) by inhalation for 6 hours per day, 5 days per week, for
131 days prior to mating, with exposure continued in females on gestation days (GDs) 1-20 and
lactation days 5-20. Two additional 500-ppm groups were similarly exposed, except that only
the F0 males were exposed in group V, and only the F0 females were exposed in group VI.
1 Denominators are the number of rats reported to have been alive at 58 weeks when the first
hemolymphoreticular neoplasia was observed.
The denominators are the reported numbers of rats alive at 33 weeks when the first malignant tumor was
observed.
30
-------
In-life parameters evaluated in adults included pre-mating body weights, observations for
mortality and clinical signs, detailed weekly physical examination, maternal body weights, and
maternal food consumption and food efficiency. One-half of all F0 males were sacrificed after
the mating period for gross postmortem examination; the remaining half were sacrificed and
examined 21 days later. One-half of the group IF0 females and group IV F0 females were
sacrificed on GD 21 for developmental toxicity evaluation; the results of this evaluation are
described in Section 4.3.2.2. The remaining F0 females were allowed to deliver litters.
Litters were standardized by pooling all pups within each treatment group on lactation
day 4 and redistributing four males and four females from this pool to each dam. However, on
some days the pups could not be pooled if only one litter was available. In this case, litters were
culled to four males and four females when possible. Pups were weighed, sexed, and given a
gross external examination on lactation days 1, 4, and 21. Randomly selected pups from each
group (one/sex/litter) and all remaining F0 females with litters were sacrificed on day 21 of
lactation and subjected to gross necropsy. The remaining pups were maintained for the post-
weaning interval of 28-49 days and weighed and sacrificed on day 49. Randomly selected pups
from each group (one/sex/litter) were given a complete gross postmortem examination.
No adverse effects were noted in F0 adults. No differences were observed in testes
weights or histologic examination of reproductive tissues in xylene-exposed males sacrificed
after mating when compared with control males. Although the female mating index in group III
and group VI was significantly lower than for controls (85 and 85%, respectively, vs. 100% for
controls), the decreases were not considered by the authors to be chemically related because a
similar effect was not observed in group IV (500 ppm-exposed males and females) and also
because the decreases were compared to an unusually high mating performance in the controls.
The male mating index, pregnancy rate, and fertility index in exposed animals were comparable
to control values. Thus, the highest exposure level in this study, 500 ppm, for 6 hours per day, 5
days per week, for 131 days before mating and continuing through weaning at day 20, is a
NOAEL for reproductive endpoints evaluated in the parental generation. Developmental
endpoints evaluated in this study are discussed in Section 4.3.2.2.
In a study of the possible influence of xylene or toluene co-exposure with n-hexane on
testicular endpoints and reproductive function in male Sprague-Dawley rats (Nylen et al., 1989),
groups of rats were also exposed to 0 or 1000 ppm xylene alone for 18 hours per day, 7 days per
week, for 61 days. The xylene tested in this study was referred to as a fluid solvent obtained
from GR Merck, but the composition of the material was not specified. In another study by this
group of investigators (Nylen and Hagman, 1994), the test material was xylene solvent from GR
Merck and was reported to contain 1.5% o-xylene, 65% m-xylene, 32% p-xylene, and 2.5%
ethylbenzene.
The means and ranges for a number of male reproductive tissue variables were essentially
the same for exposed and control rats (six rats per group) evaluated at 2 weeks or 10 months
following exposure. Endpoints evaluated were percentages of intact spermatozoa, percentages
of spermatozoa with normal heads and tails, testis weight, ventral prostate weight, and
31
-------
noradrenaline concentration in vas deferens. In addition, three rats exposed to xylene alone were
reported to have been fertile when tested 14 months after cessation of exposure. This study
identifies 1000 ppm xylene (presumably mixed xylenes) as a NOAEL for testicular effects and
fertility in male rats.
4.3.2. Developmental Studies
4.3.2.1. Oral Developmental Studies
Pregnant CD-I mice were administered mixed xylenes (60.2% m-xylene, 9.1% o-xylene,
13.6% p-xylene, 17% ethylbenzene) by gavage in cottonseed oil three times daily at doses of 0,
520, 1030, 2060, 2580, 3100, or 4130 mg/kg-day during GDs 6-15 (Marks et al., 1982)3. Mice
were sacrificed on GD 18 and maternal and fetal endpoints were assessed. The litter was the
experimental unit for statistical analysis of the data. A/> value of <0.05 was selected as the level
of significance. Differences between groups were evaluated with the Mann-Whitney U-test or
Fisher Exact test. Dose-response relationships were measured by Jonckheere's trend test.
The highest dose (4130 mg/kg-day) was lethal to 15/15 dams, and the 3100 mg/kg-day
dose, resulted in the death of 12/38 dams and decreased maternal body weight gain (49% of
controls' for GDs 1-18). Maternal body weight gains were not significantly affected at doses
<2580 mg/kg-day. Average gravid uterine weight was statistically significantly decreased at
doses >2060 mg/kg-day, compared with control values. Average gravid uterine weight also
showed a statistically significant trend for decreasing effects with increasing dose.
The average fetal weight per litter (stunted and dead fetuses were not included) was
significantly decreased in the groups treated with doses >2060 mg/kg-day. Expressed as
percentages of the control value, the average fetal weights were 100, 93, 88, 80, and 72% for the
520, 1030, 2060, 2580, and 3100 mg/kg-day groups, respectively. The trend for decreasing fetal
weight with increasing dose was statistically significant. The percent resorptions of total number
of implants was significantly different from that of controls in only the 3100 mg/kg-day group
(62.3% vs. 11.2%). Average percentages of fetuses with malformations4, consisting primarily of
cleft palate, were statistically significantly increased in groups treated with doses >2060 mg/kg-
day. This variable showed a statistically significant trend for increasing malformations with
increasing dose. For the control through 3100 mg/kg-day groups, the percentages were 0.3, 1.0,
1.0, 3.4, 7.8, and 9.1%, respectively.
3 Marks et al. (1982) noted that xylene dissolved in cottonseed oil at concentrations (v/v) of 0 (vehicle
control), 2, 3, 8, 10, 12, and 16% were administered by gavage in individual doses three times a day and that the
daily doses were 0, 0.6, 1.2, 2.4, 3.0, 3.6, and 4.8 mL/kg-day. It was noted that a density value of 0.86 g/mL was
used to convert these to units of mg/kg-day, but the actual numbers in the report should be in units of g/kg-day. This
was an apparent typographical error. The mg/kg-day doses cited herein appear to be the correct administered doses.
4 100 x X (# of malformed fetuses in litter/total # of litters) / total # of fetuses in litter
32
-------
Using decreased gravid uterine weight as a measure of maternal toxicity, the maternal
LOAEL is 2060 mg/kg-day and the NOAEL is 1030 mg/kg-day. Maternal mortality and
decreased weight gain occurred at doses >3100 mg/kg-day. The developmental LOAEL is 2060
mg/kg-day, based on decreased fetal body weight and increased percentage of fetuses with
malformations (primarily cleft palate), and the NOAEL is 1030 mg/kg-day.
In another study reported only as an abstract (Nawrot and Staples, 1980), xylene isomers
(m-, o-, or p-) were administered by gavage to pregnant CD-I mice at apparent doses of 0, 0.3,
0.75, or 1.00 mL/kg5 three times daily on GDs 6-15 or 1.0 mL/kg three times daily on GDs
12-15. Assuming a density of 860 mg/mL, these doses correspond to 0, 774, 1935, and 2580
mg/kg-day. The abstract mentioned a control group but did not specify the number of dams in
the groups. Exposure to the mid and high dose of o- or p-xylene and the high dose of m-xylene
during GDs 6-15 resulted in overt maternal toxicity (not otherwise specified) and a significantly
increased incidence of resorptions. The mid- and high-dose o- or p- xylene-exposed groups
(GDs 6-15) additionally had an increased incidence of cleft palate (the actual incidence was not
cited). Exposure to 2580 mg/kg-day of any isomer during GDs 12-15 resulted in a significant
increase in maternal lethality, with exposure to p- or m-xylene additionally resulting in a
significant increase in the incidence of malformations, consisting mostly of cleft palate.
Subsequent studies on the developmental effects of m-xylene were conducted: mice were
administered m-xylene at doses of 1935 or 2580 mg/kg-day during GDs 12-15 or 6-15.
Although exposure to m-xylene during GDs 12-15 did not result in overt toxicity at the low dose
and did not significantly increase the incidence of malformations, exposure during GDs 6-15 did
result in a low (4.4% vs. 0.0% in vehicle controls) but statistically significant increase in cleft
palate in only the high-dose group, in the absence of overt maternal toxicity. The abstract report
does not identify reliable NOAELs and LOAELs for maternal toxicity and developmental
toxicity because of incomplete reporting.
4.3.2.2. Inhalation Developmental Studies
Litton Bionetics (1978a) exposed groups of pregnant CRL:COBS CD (SD) BR rats to 0,
100, or 400 ppm xylene (52% m-xylene, 11% o-xylene, 0.31% p-xylene, 36% ethylbenzene) for
6 hours per day on GDs 6-15. On GD 20, the dams were sacrificed and each uterus was
removed and examined. The number of implantation sites, live and dead fetuses, and resorption
sites were recorded. The fetuses were removed, examined externally for abnormalities, and
weighed. One third of fetuses in each litter were fixed in Bouin's fluid and later examined for
changes in soft tissues of the head and thoracic and visceral organs. The remaining fetuses of
each litter were examined for skeletal abnormalities following staining with Alizarin Red S.
Mean body weight and food consumption were not statistically significantly affected in the
5 The abstract (Nawrot and Staples, 1980) noted that doses were administered three times daily ("t.i.d.")
from day 6-15 of gestation at 0.30, 0.75, and 1.00 mg/kg/dose and from days 12-15 of gestation at 1 mg/kg/dose.
The IARC (1989) Working Group noted that the doses were incorrectly expressed as mg, rather than mL, in the
abstract.
33
-------
exposed dams when compared with controls. When compared with control values, no
statistically significant exposure-related changes, were found in the number of live litters,
number of implantation sites, number or percentage of litters with resorptions, litters with dead
fetuses, mean liver litter size, or the average fetal body weight.
No exposure-related malformations were found in the fetuses, but some skeletal changes
indicative of retarded bone ossification were reported in the 400 ppm group. The incidence of
fetuses with retarded bone ossification was statistically significantly elevated relative to controls
in the 400 ppm group but not in the 100-ppm group. The reported incidences for total fetuses
examined with retarded bone ossification were 19/196 (9.7%), 24/197 (12.2%), and 37/201
(18.4%) for the control, 100, and 400 ppm groups, respectively. However, a Wilcoxon Rank
Sum test (based on the number of abnormal fetuses within a litter) indicated no significant
difference between the 400 ppm group and the control group. The majority of affected fetuses
(27/37) were in three litters in which all of the fetuses were small. The authors interpreted the
difference in retarded bone ossification between the 400 ppm and control groups to be not
related to treatment. The highest exposure level in this study, 400 ppm, is judged to be a
NOAEL for maternal and developmental toxicity.
In the one-generation reproductive study of CD rats described in Section 4.3.1.2
(Bio/dynamics Inc., 1983), one-half of the group I F0 pregnant dams (20 females; control group)
and group IV F0 pregnant dams (12 females exposed to 500 ppm mixed xylenes by inhalation for
6 hours per day, 5 days per week, during a premating period and during gestation) were
sacrificed on GD 21 for developmental toxicity evaluation. Gross necropsy was conducted on
each animal. Maternal exposure to 500 ppm mixed xylenes did not adversely affect maternal
body weights, food consumption, food utilization, or the results of postmortem examination.
Corrected terminal body weights (corrected for gravid uterine weights) for exposed females were
statistically significantly increased when compared with those of controls, but the increases were
not considered to be biologically significant (106% of controls). Although absolute kidney
weights were statistically increased in group IV females (110% of controls'), kidney weights
relative to body weights were comparable to those of controls. The increase in absolute kidney
weights in the exposed females was, therefore, attributed to the higher body weights.
No statistically significant differences were noted between treated (group IV) and control
groups for mean number of corpora lutea, implantations, live fetuses, mean percentage of live
fetuses/implants, or fetal sex ratios. Although the exposed group had an increased mean number
of resorption sites (1.6 vs. 1.2 for controls) and mean percentage of resorptions to implants
(16.2% vs. 9.9% for controls), the increases were not statistically significant. No dams had
whole litter resorption. No definitive treatment-related external, visceral, or skeletal
malformations or variations were observed. The exposed group had a slightly higher incidence
of unossified sternebrae and incompletely ossified cervical vertebral transverse processes, but the
incidences were provided in terms of fetal incidence instead of litter incidence. Mean fetal body
weights on GD 21 were marginally but statistically significantly decreased in exposed female
fetuses (93% of controls'); however, male fetal weights were comparable to those of controls.
The decrease is judged to not be biologically significant.
34
-------
The assessment of developmental effects in the fetuses at GD 21 in this phase of the
Bio/dynamics Inc. (1983) study identified 500 ppm as a NOAEL for maternal and developmental
toxicity (including assessment of external, visceral, and skeletal fetal malformations or
variations).
Dams in the other exposed groups (II, 60 ppm; III, 250 ppm; V, only males exposed to
500 ppm; VI, only females exposed to 500 ppm) and the other half of the dams in group IV
(males and females exposed premating F0, 500 ppm) and the control group I were allowed to
deliver their litters. No gross malformations were reported among the pups. Pups were weighed
on lactation days 1, 4, 14, 21, and 49. No statistically significant decrease in mean pup body
weights were observed in the exposed versus control groups at days 1 and 14. On lactation day
4, mean pup weights were statistically significantly decreased in groups II (60 ppm), III (250
ppm), and IV (500 ppm) (post-pooling) when compared with controls, but the decreases (about
8%) were not of a biologically significant magnitude. The decreased weights may have been the
consequence of an elevated mean pup weight in the control group potentially caused by a smaller
mean litter size (mean number of live pups per litter: 9.6, 11.8, 12.5, 12.4, 10.8, and 11.8 for
groups I-VI, respectively).
Pups from group IV had statistically significant decreased mean pup weights on lactation
day 21 (90% of controls') and statistically significant decreased terminal body weights at 49
days of age (as a percentage of controls': males, 92%; females, 93%). However, despite the
marginal decreases observed in mean pup weights in group IV, no decreases in body weights
were observed in pups from group VI, in which dams were exposed to the same concentration of
xylene (500 ppm) for the same period of time as were dams in group IV. Due to the
inconsistency between the fetal body weights in groups VI and IV and the small mean litter size
in the control group, the marginal decreases observed in mean pup weights from group IV are
not considered to be an adverse effect of treatment.
Female pups from the mid- and high-dose groups (groups III and IV) also had
statistically significant decreased absolute (76 and 78% of controls', respectively) and relative
(80% and 84% of controls', respectively) ovary weights at 21 days of age, but the decreases were
not concentration related and were not observed at 49 days of age. In addition, decreases in
ovary weights were also not observed in group VI pups.
In summary, the highest exposure level (500 ppm mixed xylene for 6 hours per day for
131 days prior to mating and continuing through lactation) in this study is interpreted as a
systemic and reproductive toxicity NOAEL for adult CD rats. The study also reliably identified
500 ppm mixed xylenes as a NOAEL for maternal toxicity and developmental toxicity.
Hudak and Ungvary (1978) exposed groups of pregnant CFY rats to 0 or 230 ppm (1000
mg/m3) xylenes (10% o-xylene, 50% m-xylene, 20% p-xylene, 20% ethyl benzene) for 24 hours
per day during GDs 9-14. The exposed and control groups had 20 and 28 dams, respectively.
Dams were sacrificed on GD 21, and fetuses were weighed and prepared for examination for
gross, visceral, and skeletal malformations and variations. Differences between the groups were
35
-------
statistically assessed with t-tests or the Mann Whitney U test, usingp<0.05 as the level of
significance. No statistically significant differences in maternal body weights; fetal deaths;
mean fetal or placental weights; or external, visceral, or skeletal malformations or signs of
skeletal retardation were noted in the exposed group when compared with the control group.
Exposed group fetuses showed increases for skeletal anomalies. Compared with controls,
the exposed group showed statistically significant increased incidences of fetuses with fused
sternebrae (7/213 vs. 1/315) and extra ribs (22/213 vs. 2/315). The incidence was based on the
number of affected fetuses rather than the affected litters; litter-specific information was not
provided. The interpretation of the observed statistically significant increases in the incidences
of fetal skeletal anomalies (extra ribs or fused sternebrae) is confounded by the lack of ability to
adjust for possible litter size covariation. The study identified 230 ppm as a NOAEL for
maternal and developmental toxicity.
In a study of the potential developmental toxicity of xylene isomers (Ungvary et al.,
1980), groups of 15-30 pregnant CFY rats were exposed by inhalation to air containing
measured concentrations of 0, 35, 350, or 700 ppm (0, 150, 1500, or 3000 mg/m3) of o-, m-, or p-
xylene (analytical purity; actual purity not provided) for 24 hours per day on GDs 7-14. Dams
were sacrificed on GD 21. Four dams in the 700 ppm m-xylene group died. Necropsy revealed
hyperaemia and hemorrhage in several organs, pulmonary edema, and distention of the gut and
urinary bladder. The authors stated that maternal food consumption was considerably less in the
350 and 700 ppm o-xylene or p-xylene groups during the exposure period (GDs 7-14), but
returned to normal when exposure was discontinued (data were not provided).
Maternal body weight gain exhibited a concentration-related decrease during exposure to
all three isomers (data not provided) but was comparable to controls' by GD 21 except for the
group exposed to 700 ppm m-xylene (body weight gain 73% of controls' for GDs 0-21;/><0.05).
Dams exposed to 350 or 700 ppm o-xylene had slightly elevated but statistically significant
liver-to-body weight ratios (109 and 108% of controls', respectively) and had an increase in the
rough endoplasmic reticulum profile and smooth endoplasmic vesicles as compared with
controls. No other findings in the dams were reported.
Exposure to 700 ppm m-xylene—but not o- or p-xylene—resulted in a small but
statistically significant decreased number of mean implantations per dam (11.44 vs. 13.52 in
controls). In contrast, 700 ppm p-xylene—but not o- or m-xylene—resulted in a marked
postimplantation loss (69% vs. 4% in controls) and a corresponding decreased mean litter size
(8.5 vs. 12.6 in controls). Mean fetal body weights were statistically significantly decreased in
the 350 and 700 ppm o-xylene groups (91 and 92% of controls', respectively) and in the 700
ppm p- and m-xylene groups (88 and 91% of the respective controls'). There was a
corresponding increase in the number of weight-retarded fetuses (< 3.3 g) in these same groups.
Histochemical analyses of fetuses from the 700 ppm o- and p-xylene groups revealed
decreased staining of alkaline phosphatase in the proximal convoluted tubules and of succinic
hydrogenase, acid phosphatase, and glucose-6-phosphatase in the renal nephron. Additionally,
36
-------
decreased activities of succinic dehydrogenase and glucose-6-phosphatase were observed in the
liver and thymus cells in fetuses from the 700 ppm m-, p-, and o-xylene groups. No treatment-
related changes were observed following histopathologic or electron microscopic evaluation of
organs in fetuses from exposed dams.
Fetuses were examined for external, visceral, and skeletal anomalies, but only fetal
incidence rates were reported; litter-specific information was not provided. No statistically
significant exposure-related changes were reported for incidences of fetuses with external,
visceral, or skeletal malformations in any of the groups exposed to any of the isomers.
Statistically significant increases in incidence of fetuses with extra ribs were reported at 700 ppm
m-xylene (8/196 [4.1%] vs. 2/284 [0.1%] for controls) or at 700 ppm p-xylene (10/51 [19.6%]
vs. 6/226 [2.7%] for controls). Statistically significant increases in incidences of fetuses with
skeletal retardation were also observed at 35, 350, and 700 ppm p-xylene (24/157 [ 15.3%],
38/197 [19.3%], and 29.4% [29.4%], respectively, vs. 16/226 [7.1%] for controls), and at 700
ppm o-xylene (48/234 [20.5%] vs. 13/168 [7.7%] for controls).
The interpretation of the observed statistically significant increases in the incidences of
fetal skeletal retardation or anomalies (extra ribs) is confounded by the lack of ability to adjust
for possible litter size covariation; therefore, this study identified 700 ppm o-xylene, p-xylene, or
m-xylene (for 24 hours per day on GDs 7-14) as a LOAEL and 350 ppm as the highest NOAEL
for maternal toxicity (decreased body weight). For p-xylene, 700 ppm is a developmental
LOAEL and 350 ppm is a NOAEL for postimplantation loss, decreased litter size, and decreased
fetal body weight. For m-xylene, 700 ppm and 350 ppm are a developmental LOAEL and
NOAEL, respectively, for decreased fetal body weight. Finally, for o-xylene, 350 ppm is the
developmental LOAEL and 35 ppm is the NOAEL for decreased fetal body weight.
In a subsequent study with rats, mice, and New Zealand rabbits, Ungvary and Tatrai
(1985) exposed groups of pregnant CFY rats (19-23) to 0, 60, 440, or 780 ppm (0, 250, 1900, or
3400 mg/m3) xylenes (presumably mixed xylenes, but the composition was not specified) for 24
hours per day during GDs 7-15. The animals were sacrificed on GD 21. Groups of pregnant
CFLP mice (17-18) were exposed to 0, 115, or 230 ppm (0, 500 or 1000 mg/m3) mixed xylenes
or to 115 ppm (500 mg/m3) o-xylene, m-xylene, or p-xylene for three 4-hour periods daily on
GDs 6-15 or 7-20, respectively. Groups of 10 pregnant New Zealand rabbits were exposed to 0,
115, or 230 ppm mixed xylenes (0, 500, or 1000 mg/m3) or 115 ppm o-, m-, or p-xylene for 24
hours per day on GDs 7-20. The mice and rabbits were sacrificed on GDs 18 and 30,
respectively. Data from this study were reported in terms of percentage of fetuses affected in the
four groups; litter-specific data were not reported. Maternal toxic effects were reported to be
moderate and dose-dependent but were not otherwise specified.
In rats exposed to 780 ppm mixed xylenes, statistically significant findings were
increased proportions of dead or resorbed fetuses (13% vs. 5% in controls), weight-retarded (not
otherwise specified) fetuses (13% vs. 2%), skeletal retarded fetuses (31% vs. 13%); and fetuses
with an extra rib (9% vs. 0%). In rats at 60 and 440 ppm, statistically significant findings were
restricted to increased proportion skeletal retarded fetuses (32 and 33% vs. 13% in controls).
37
-------
In mice exposed to mixed xylenes, statistically significant findings were restricted to the
230 ppm group (three 4-hour periods daily) and consisted of increased proportions of weight-
retarded fetuses (30% vs. 7% in controls) and skeletal-retarded fetuses (13% vs. 5%). In mice
exposed to 115 ppm o-, m-, or p-xylene, statistically significant findings were restricted to
increased proportions of skeletal retarded fetuses (11, 11, or 12%, respectively, vs. 5% in
controls).
In rabbits exposed to 230 ppm mixed xylenes for 24 hours per day, no live fetuses were
produced: three dams died, six aborted, and the remaining dam showed total resorption. Another
group of eight rabbits exposed to 230 ppm p-xylene for 24 hours per day also produced no live
fetuses: one dam died, three aborted, and four showed total resorptions. These findings are
indicative of severe maternal toxicity at this exposure level. The report did not clearly specify
whether or not groups of pregnant rabbits were exposed to the o- or to the m-xylene isomers at
230 ppm. In rabbits exposed to 115 ppm mixed xylene, no increases were observed, compared
with controls, in percentages of fetuses with skeletal retardation, minor anomalies, or skeletal,
internal, or external malformations, but the average female (but not male) fetal body weight was
statistically significantly decreased by about 10% (29.4 g vs. 32.7 g in controls). Given the small
magnitude of this decrease and the inconsistency across sexes, this effect on body weight is not
likely to be biologically significant.
In rabbits exposed to 115 ppm o-, m-, or p-xylene for 24 hours per day, no statistically
significant differences from controls were reported in the number of abortions, number of live
fetuses, proportions of dead or resorbed fetuses or fetuses with skeletal malformations, minor
anomalies, or skeletal, internal, or external malformations, with the exception that an increased
percentage of dead or resorbed fetuses was reported for the m-xylene groups (12.8% vs. 5.2% in
controls).
As with the earlier study by Ungvary et al. (1980), interpretation of statistically
significant findings for increased incidence of fetuses with retarded skeletal ossification reported
by Ungvary and Tatrai (1985) is difficult, given the inability to adjust for possible litter size
covariation and the relatively small magnitude of the increased incidences. For rats and mice,
maternal body weight data are not reported in a manner that enables identification of maternal
toxicity NOAELs or LOAELs. For rats in this study, 780 ppm mixed xylenes (for 24 hours per
day on GDs 7-15) is the apparent developmental LOAEL for increased percentage of dead or
resorbed fetuses, and 440 ppm is the highest NOAEL for biologically significant developmental
effects. For mice, the highest exposure level—230 ppm mixed xylenes (three 4-hour periods
daily on GDs 6-15)—is the apparent NOAEL for developmental effects, including skeletal,
visceral, and external malformations and variations.
For mice exposed to o-, p, or m-xylene, the study identified 115 ppm (three 4-hour
periods daily) as a NOAEL for biologically significant effects on fetal survival or fetal
malformations or variations. For pregnant rabbits, 230 ppm mixed xylenes (for 24 hours per day
on GDs 7-20) completely impaired the dams' ability to deliver live fetuses, indicating a severe
maternal toxicity that was not apparent at 115 ppm; 115 ppm was also a NOAEL for
38
-------
developmental toxicity. For rabbits exposed to o-, p-, or m-xylene, the study identified 115 ppm
as a NOAEL for effects on fetal survival and fetal malformations or variations; exposure to p-
xylene at 230 ppm produced severe maternal toxicity and no live fetuses were produced.
To evaluate the effects of prenatal exposure on postnatal neurologic development, Hass
et al. (1995) exposed pregnant rats (Mol:Wist) to 0 or 500 ppm xylenes (19% o-xylene, 45% m-
xylene, 20% p-xylene, 15% ethylbenzene) by inhalation for 6 hours per day on GDs 7-20 and
allowed them to litter. Litter size was not standardized, but litters with fewer than six pups were
not used. From each litter (n = 13-15 litters), two males and two females were kept for
behavioral testing. One male and one female from each litter were kept in standardized housing
until 3 months, when they underwent the Morris water maze test (finding a hidden platform
while swimming in a maze). The remaining male and female from each litter were kept in
enriched housing (cages contained various toys) and tested for rotarod (the ability to remain on a
rotating rod for 30 seconds), open field, and Morris maze performance at about 3 months of age.
The results were evaluated by analysis of variance using a repeated measures design and a
significance level ofp=0.05.
Exposure to xylenes did not affect maternal clinical signs, body weight gain, or food
consumption. Control and exposed groups had a similar gestation period, number of pups per
litter, and sex distribution per litter. The number of litters available for evaluation in the control
and exposed groups was 13 and 15, respectively. Exposed litters had a slight decrease in mean
birth weight (5%) and a trend toward lower body weight during the postnatal followup, but the
differences did not achieve statistical significance. When the data were combined for males and
females, absolute brain weights were statistically significantly decreased on postnatal day (PND)
28, but statistically significant decreases were not observed in absolute or relative brain weights
when considering males or females separately or for relative brain weights of males and females
combined. The air-righting reflex was statistically significantly delayed by 1 day in exposed
litters due to the ability of only four pups to right themselves. No differences were observed in
open field performance, and the observed decrease in rotarod performance in exposed female
pups was not statistically significant.
Offspring from xylene-exposed rats that were raised in the enriched environment showed
no difference in the Morris maze test when compared with controls. Offspring from exposed rats
that were raised in the standard housing, however, had impaired performance. Testing at 12
weeks showed a nonstatistically significant trend (p=0.059) for increased latency for finding the
platform at the beginning of the learning test. At 16 weeks, exposed offspring took statistically
significantly more time to find a platform hidden in the center of the pool. Further analysis
revealed that the effect was limited to the female offspring from the standard housing (not the
enriched housing) and that these females had an increase in swimming length, but swim speed
was unaffected. In four consecutive trials conducted at 16 weeks, the mean time (i.e., latency) to
find the hidden platform was consistently greater in the standard housing exposed females than
in the control female offspring.
39
-------
Female offspring from the standard housing conditions were also evaluated at 28 and 52
weeks (Hass et al., 1997). At 28 weeks, an increased latency for finding a platform that was
moved to a new position was observed in the female offspring only during the first trial of a
three-trial testing block, whereas the next two trials resulted in similar latencies between exposed
and control rats. The increased latency again corresponded with increased swimming length.
No other statistically significant differences were observed for other testing situations in the
Morris maze test. At 55 weeks, no statistically significant differences were observed between
groups.
The Hass et al. (1995, 1997) studies found that prenatal exposure to 500 ppm xylenes, 6
hours per day on GDs 7-20 affected the performance of standard housing female rats in the
Morris water maze test: it took the female offspring longer to find a hidden platform while
swimming in a water maze. Although swim length (i.e., the distance covered before finding the
platform) was increased, swim speed was unaffected, indicating a cognitive rather than a motor
effect. This study is limited, however, in that only one concentration was tested. Overall, the
results suggest that gestational exposure to 500 ppm produced a minimally adverse effect on
neurologic development, and it appeared to be reversible.
The statistically significant differences between exposed and control female offspring in
time to find a hidden platform that were observed at 16 and 28 weeks were not observed at 55
weeks, and the difference in latency between the exposed and control females in finding a hidden
platform at 28 weeks was only observed in the first of three consecutive trials. In addition, no
clear effects were observed in the other neurological tests (rotarod performance and open field
activity) or in offspring housed in an enriched environment. Thus, this study identifies 500 ppm
on GDs 7-20 as a developmental LOAEL for reversibly impaired Morris water maze
performance in rat offspring; 500 ppm is a NOAEL for maternal toxicity.
Hass and Jakobsen (1993) exposed groups of 36 pregnant Wistar rats to air containing 0
or 200 ppm technical xylene (composition not provided) for 6 hours per day during GDs 6-20.
On GD 21, two-thirds of the rats were sacrificed and were used to assess developmental toxicity.
One-third of the rats were allowed to litter. Developmental milestones and rotarod performance
were assessed in eight offspring (four males and four females) from each litter. No maternal
toxicity was observed in the exposed dams. The only effect noted in fetuses from exposed dams
was an increased incidence of delayed ossification of os maxcillare in the skull, with 18/26
exposed litters affected versus 2/22 control litters. In the postnatal study, statistically
significantly decreased rotarod performance was observed in female pups on PNDs 22 and 23,
and in male pups on PND 23.
This study is limited in that only one exposure concentration was tested and only a
limited battery of behavioral tests were used. In a later report, Hass et al. (1995) questioned the
significance of the effect on rotarod performance, because the testers were not blind to the
exposure status of the animals. In addition, the later study did not find a clear effect of 500 ppm
technical xylene (for 6 hours per day on GDs 7-20) on rotarod performance. Thus, Hass and
Jakobsen (1993) did not identify a reliable developmental NOAEL or LOAEL.
40
-------
Rosen et al. (1986) exposed 18 to 21 pregnant Sprague-Dawley rats to 0, 800, or 1600
ppm p-xylene (0, 3500, or 7000 mg/m3; 99% pure) on GDs 7-16. The treatment did not affect
litter size or the weight of pups at birth or on PND 3; CNS development, as measured by the
acoustic startle response on PNDs 13, 17, 21, and 63, or the figure-8 maze activity, evaluated on
PND 22 and 65; or the growth rate of the pups. The only effect of exposure was a statistically
significant decrease in maternal body weight gain in the 1600 ppm dams (74% of controls'). The
maternal toxicity LOAEL is 1600 ppm, based on decreased body weight gain, and the NOAEL is
800 ppm. The developmental neurotoxicity NOAEL is 1600 ppm.
Kiikner et al. (1997/98) investigated the effect of xylene inhalation on the liver of
pregnant and nonpregnant rats and pups of exposed litters by exposing pregnant Wistar rats to 0
or 2600 ppm xylenes (0 or 11,300 mg/m3) (purity and composition not stated) for 8 hours per day
on GD 6 until term (GD 21). Nonpregnant rats were exposed to 2600 ppm xylenes for the same
period, and a control group of pregnant rats inhaled clean air (not stated whether nonpregnant
controls were also included). Biochemical analysis of the livers from pregnant rats exposed to
xylene revealed increases in AST (18%), ALT (19%), alkaline phosphatase (17%), and arginase
(63%). Electron microscopic evaluation of pregnant and nonpregnant rat liver tissue revealed
mitochondria that concentrated near the periphery of hepatocytes and nuclei as increased number
of lysosomes and expanded smooth endoplasmic reticulum. In fetal livers from exposed litters,
findings included expanded smooth endoplasmic reticulum, structurally deformed mitochondria,
and granular endoplasmic reticulum. No structural defects were observed in the kidneys or
pancreas of exposed pregnant or nonpregnant rats or of fetuses from exposed litters.
Mirkova et al. (1983) exposed groups of pregnant white Wistar rats to air containing 0, 3,
12, or 110 ppm (0, 14, 53, or 468 mg/m3) xylene isomers (composition not provided) for 6 hours
per day, 5 days per week, during GDs 1-21. On GD 21, a number of the animals were sacrificed
for intrauterine toxicity evaluation, and the remainder were allowed to deliver for postnatal
evaluations of pups. The pregnancy rates were 29/36, 11/18, 18/27, and 11/15 for the 0, 3, 12,
and 110 ppm groups, respectively.
The authors reported numerous manifestations of toxicity in mid- and high-dose groups,
including a statistically significant increased percentage of post-implantation loss per
implantation (10.7% and 14.9%, respectively, vs. 5.5% for controls); a statistically significant
decreased fetal body weight (3.20 g and 3.17 g, respectively, vs. 3.64 for controls); and a
statistically significant increase in the percentage of hemorrhages in fetuses (46% and 53%,
respectively, vs. 31% for controls). Although the authors reported an increased incidence of
anomalies of the internal organs (including hydrocephalus, microphthalmia, intracerebral
hematomas, and hemorrhages in the liver) and defects in ossification of the sternum and bones of
the skull in fetuses from exposed dams, the incidence rates for these anomalies were not
provided. A statistically significant decrease in pup weight on PNDs 7 and 21 was also reported
for the mid- and high-dose groups, but data were not provided.
The data from this study are limited by numerous factors, including composition and
purity of xylenes not being provided, incomplete description of methods, inadequate litter size
41
-------
for proper fetal evaluations, high incidence of fetal hemorrhages in the control group (calling
into question the health of the animals), and incomplete reporting of results. Therefore, this
study does not identify reliable NOAELs or LOAELS for maternal or developmental toxicity.
4.4. OTHER STUDIES
4.4.1. Neurotoxicity Studies
4.4.1.1. Prechronic Oral Studies
In an acute study (NTP, 1986), groups of five male or five female B6C3F1 mice or
Fischer 344/N rats were administered a single dose of 0, 500, 1000, 2000, 4000, or 6000 mg/kg
mixed xylenes by gavage in corn oil. In mice, mortality was observed before the end of the
study in 3/5 high-dose males and 4/5 high-dose females. Clinical signs reported in mice dosed
with 4000 or 6000 mg/kg xylenes included tremors, prostration, and/or slowed breathing. In
rats, mortality was observed within 48 hours of dosing in 5/5 high-dose males or females and in
3/5 males dosed with 4000 mg/kg. Clinical signs observed in rats dosed with 4000 or 6000
mg/kg xylenes included lack of coordination, prostration, loss of hindleg movement, and
hunched posture within 24 hours of dosing, and rough coats were observed in 2000 mg/kg dose-
groups. Surviving animals did not exhibit any clinical signs by the end of week 1.
In a subacute study (NTP, 1986), groups of five male or female Fischer 344 rats were
dosed with 0, 125, 250, 500, 1000, or 2000 mg/kg mixed xylenes orally by gavage for 14
consecutive days. Treatment-related mortality was observed in 3/5 high-dose males and 5/5
high-dose females. High-dose male and female rats exhibited shallow labored breathing and
prostration immediately after dosing. Additionally, body weight gains were reduced by 23-29%
in males dosed with 250, 500, or 1000 mg/kg when compared with controls, whereas females
dosed with 125 or 1000 mg/kg had body weight gains 17% and 26% lower than those of
controls.
Also in NTP (1986), groups of 10 male and 10 female B6C3F1 mice were administered
mixed xylenes (60% m-xylene, 13.6% p-xylene, 17.0% ethylbenzene, 9.1% o-xylene) in corn oil
by gavage at doses of 0, 125, 250, 500, 1000, or 2000 mg/kg-day for 5 days/week for 13 weeks.
At termination of the study, necropsy was performed on all animals, and comprehensive
histologic examinations (27 organs) were performed on vehicle and high-dose group animals.
Effects noted at the high dose included the death of two female mice; clinical signs of lethargy,
short and shallow breathing, unsteadiness, tremors, and paresis occurring 5-10 minutes post-
dosing and lasting for 15-60 minutes in male and female mice; and decreased body weight gain
in male and female mice (7% and 17%, respectively, less than controls'). No treatment-related
gross or microscopic pathologic lesions were observed. No adverse effects were reported in
mice dosed with 125, 250, 500, or 1000 mg/kg-day. The LOAEL is 2000 mg/kg-day, based on
obvious clinical signs of nervous system depression, and theNOAEL is 1000 mg/kg-day.
42
-------
4.4.1.2. Prechronic Inhalation Studies
Korsak et al. (1992) exposed groups of 12 male Wistar rats to toluene, m-xylene, or a 1:1
mixture for 6 hours per day, 5 days per week, at a concentration of 0 or 100 ppm for 6 months or
1000 ppm for 3 months. Rotarod performance and spontaneous motor activity were assayed 24
hours after termination of the exposure periods. The rotarod test was used as a measure of motor
coordination disturbances from exposure to m-xylene. The rotarod test involves placing the
subject animals on a rotating rod and evaluating their ability to remain on the rod for a period of
2 minutes. The animals were trained to perform the task, exposed to chemical or control gas,
and evaluated at defined intervals. By the time interval after exposure, considerable proportions
of absorbed xylenes would be expected to have been eliminated from the body (see Section 3.4
and Appendix B). Body weights and weights of seven organs were measured; only data for
animals sacrificed after 3 months of exposure was reported (controls and 1000 ppm rats).
At 3 and 6 months, blood samples were collected 24 hours after termination of exposure
for measurement of serum chemistry variables (e.g., ALT, AST, sorbitol dehydrogenase, alkaline
phosphatase, and total protein) and hematologic variables (erythrocyte counts, hemoglobin
concentration, hematocrit, leukocyte count, and differential leukocyte counts). Serum chemistry
and hematologic results were reported only for rats exposed to 1000 ppm for 3 months.
Statistical evaluations (using ap=0.05 level of significance) of collected data included analysis
of variance, Dunnet's test, and the Fisher exact test.
Rats exposed to m-xylene alone exhibited statistically significantly decreased rotarod
performance and decreased spontaneous activity, as measured 24 hours after termination of the
exposures, when compared with controls. The percentages of failures in the rotarod test were
roughly 60% in rats exposed to 1000 ppm for 3 months, 35% in rats exposed to 100 ppm for 6
months, and 0% for controls at either time period. The mean spontaneous motor activity in rats
exposed to 100 ppm for 6 months was about 400 movements per hour, compared with about 800
movements per hour for controls. Spontaneous motor activity data for rats exposed to 1000 ppm
m-xylene for 3 months were not presented in the report.
No statistically significant exposure-related changes in body weight, absolute or relative
organ weights, or clinical chemistry or hematology variables were noted in rats exposed to 1000
ppm m-xylene for 3 months, with the exception of decreased differential counts (percentage of
white blood cells counted) of lymphocytes (45.5 ± 9.5 vs. 60.8 ± 6.4 for controls; 25% decrease)
and increased counts of monocytes (16.3 ± 8.9 vs. 8.3 ± 4.2 for controls; 96% increase).
However, total counts of white blood cells (in units of cells per mm3 of blood) were not
statistically significantly changed by exposure. The LOAEL is 100 ppm, based on decreased
rotarod performance and decreased spontaneous motor activity. No NOAEL was identified.
In a second study, Korsak et al. (1994) exposed groups of 12 male Wistar rats to 0, 50, or
100 ppm m-xylene or n-butyl alcohol or a 1:1 mixture (purity of chemicals not provided) for 6
hours per day, 5 days per week, for 3 months and evaluated similar endpoints as in the earlier
study. Blood for clinical biochemistry and hematologic analysis was collected 24 hours after
43
-------
termination of the inhalation exposure. The report does not specify the timing of the neurologic
examinations, but it appears reasonable to assume that it was the same as in the earlier study by
the same investigators, that is, 24 hours after termination of exposure. Statistical evaluations
(using ap=0.05 level of significance) of the collected data included analysis of variance,
Dunnet's test, and the Fisher exact test.
No statistically significant exposure-related changes were noted in body weight gain,
absolute or relative organ weights, hepatic activities of microsomal monooxygenases, lipid
peroxidation, or levels of triglycerides in the liver. Statistically significant decreases in
erythrocyte number were seen in animals exposed to 50 ppm (93% of controls') or 100 ppm
(80.5% of controls') of m-xylene alone. Similarly, decreased levels of hemoglobin were
reported in both groups (92% of controls' for both groups). At 100 ppm, a statistically
significant increase in leukocyte number (35% increase over controls') was reported. Exposure
to 50 or 100 ppm m-xylene alone also resulted in decreased rotarod performance starting at 1
month of exposure and continuing until the end of the 3-month exposure. Decreases were
statistically significant in the 100 ppm group when compared with controls. The results were
presented in graphical form; the actual numerical data are not provided. Decreases in
performance were roughly 8% and 33% for the 50 and 100 ppm groups, respectively, versus 0%
for the controls.
Sensitivity to pain was assessed using the hot plate behavior test, in which the animals
are placed on a hot (54°C) surface and the time interval between being placed on the plate and
licking of the paws is measured. Rats exposed to 50 or 100 ppm m-xylene alone had statistically
significant increased sensitivity to pain at the end of the 3-month exposure (latency of the paw-
lick response was 8.7 and 8.6 seconds, respectively, vs. 12.2 seconds for the controls). The
LOAEL is 100 ppm, based on decreased rotarod performance and decreased latency in the paw-
lick response in the hot-plate test, and the NOAEL is 50 ppm.
To evaluate whether xylene influences aging of the CNS or induces persistent changes in
radial maze performance, Gralewicz et al. (1995) exposed 8-month-old male LOD-Wistar rats
(20 per dose level) to air containing 0, 100, or 1000 ppm "pure" m-xylene (exact purity not
provided) for 6 hours per day, 5 days per week, for 3 months. One-hour electroencephalograph
(EEG) recordings were performed on days 28 and 56 of exposure and on days 14, 28, 56, and 84
after exposure. The number and duration of spontaneous neocortical spike and wave discharges
(SWDs) from the EEG were taken as electrophysiological indices of the biological age of the
brain. As rats age, SWDs increase in number and become longer. Because of large
interindividual variation in number and duration of SWDs within each group, these variables
were normalized to a percentage of the initial values. Exposed rats were not subjected to the
daily exposure protocol when EEG recordings were made on days 28 and 56 during the exposure
period.
Tests of spatial learning in an 8-arm radial maze were also conducted for a 2-week period
starting from day 70 after exposure to day 83. During the first adaptation stage of the test (five
consecutive daily training periods), rats were familiarized with the maze. The second stage (five
44
-------
consecutive daily trials) measured effectiveness of finding water in the maze (e.g., duration of
trial, number of entries into the arms, number of omission and preservation errors). One-way or
two-way parametric analysis of variance was applied to the collected data, and effects were
regarded as statistically significant at/><0.05. Body weights were also measured during and
after the exposure period at various intervals, but statistically significant differences among the
groups were not found.
The analysis of variance indicated no group effect on the normalized number and
cumulative-duration SWD variables. However, a statistically significant group x successive
recording period effect was indicated. In control rats, these variables were increased to a
statistically significant degree over those in the exposed groups only on day 84 after exposure.
The mean cumulative SWD duration (expressed in percentage) on day 84 was about 300 for the
control, compared with means of about 150 in each of the exposed groups. The authors
hypothesized that these exposure-related changes in the spontaneous, age-related changes in
cortical SWD activity may have been related to cortical excitability or an increase in
catecholaminergic transmissions. Unlike the controls, rats exposed to 100 or 1000 ppm m-
xylene did not exhibit a statistically significant shortening of the time needed to complete a trial
in the radial maze with successive daily trials.
These results indicate a learning deficit in the exposed rats. For example, on the fifth
consecutive trial, the mean trial durations in each of the exposed groups were about 240-250
seconds, compared with a mean of about 150 seconds for the control group. In addition, the
exposed groups did not exhibit the statistically significant decrease in omission errors with
successive days in the radial arm maze test that was exhibited by the control group (number of
arms in the maze omitted during a 5-minute period when the rats explored the maze). The mean
number of omission errors in control rats showed a progressive decrease from about 2.75 on the
first trial to 0 on the fourth and fifth successive trials. In contrast, the means on the fifth
consecutive trial were about 1.5 and 2.5 for the 100- and 1000 ppm groups, respectively. The
lowest exposure level in this study, 100 ppm, is designated as a LOAEL for deficits in radial
maze performance.
Gralewicz and Wiaderna (2001) exposed groups of male Wistar rats (10-11 animals per
group) to 0 or 100 ppm of m-xylene for 6 hours per day, 5 days per week, for 4 weeks.
Behavioral testing was performed at various intervals before exposure (radial maze and open-
field evaluations) and after exposure (radial maze [days 14-18], open-field activity [day 25],
passive avoidance [days 39-48], hot plate test [days 50-51], and active avoidance [days 54-60]).
The radial maze and hot plate test protocols were the same as in other studies by this group
(Gralewicz and Wiaderna, 2001; Gralewicz et al., 1995; Korsak et al., 1992) described above.
In the open-field activity test, animals were placed in a 100 cm x 100 cm arena that was
surrounded by 20 cm-high walls and divided into 49 equal squares. The number of square
borders crossed (locomotor activity), number of rearings (exploratory activity), and number of
grooming episodes were recorded. In the passive avoidance test, animals were placed on a
platform above the floor of the cage, and the time until the animal stepped off the platform was
45
-------
recorded in a series of six trials. In the first two trials, the animals were allowed to explore the
cage for 60 seconds after stepping down; in the third trial, they received a series of footshocks
after stepping off the platform. In trials 4, 5, and 6 the animals received no shocks and were
allowed to stay on the floor for 1 minute after stepping off the platform. In the active avoidance
test, the animals were trained to avoid an electric footshock by moving from one compartment of
the cage to another when a sound was played. After successfully displaying avoidance behavior
in four of five trials, the animals were considered to be trained. Postexposure evaluations
determined the frequency of avoidance behavior in response to the same stimulus.
No differences between control and exposed rats were seen in radial maze parameters
(number of arm entries, arms omitted, or arms entered multiple times) either before exposure (7
days prior to exposure) or at 14-18 days after the termination of exposure. Similarly, no
differences between groups were seen in open-field activity, examined on day 8 prior to
exposure and day 25 postexposure, or in active avoidance (number of trials to avoidance
criterion), examined on days 54 and 60 postexposure. Xylene-exposed rats showed a
significantly shorter step-down time (trial 6 only; no difference in trials 1-5) in the passive
avoidance test (examined on days 39-48 postexposure) and a significantly greater paw-lick
latency in the hot plate behavior test (35 seconds vs. 10 seconds in control) (examined on days
50-51 postexposure), identifying 100 ppm as a LOAEL for neurobehavioral effects.
Pry or et al. (1987) conducted studies to examine the potential for xylene to cause
ototoxicity. Groups of 12 weanling male Fischer 344 rats were exposed to air containing 0, 800,
1000, or 1200 ppm mixed xylenes (10% p-xylene, 80% m-xylene, 10% o-xylene) for 14 hours
per day for 6 weeks. Chamber concentrations were measured at least once daily. Hearing loss
was assessed by measuring behavioral auditory thresholds (conditioned avoidance response
task), whereby rats were trained to pull or climb a pole suspended from the ceiling to avoid a
shock following warning tones and by measuring brainstem auditory-evoked response (BAER),
an electrophysiologic measurement of auditory function. The frequencies of the tones tested
were 4, 8, 12, and 20 kHz, with the sound levels (decibels) varying with frequency. Results were
presented only in graphical form, with actual data not provided. Tests were conducted every 2
weeks during exposure and at 2 weeks postexposure.
All xylene-exposed groups had dose-dependent increases in behavioral auditory and
BAER thresholds relative to controls at some frequencies. Behavioral auditory thresholds were
elevated at 12 and 20 kHz in 800 ppm-group rats; at 8, 12, and 20 kHz in 1000 ppm-group rats;
and at all frequencies in 1200 ppm-group rats. BAER thresholds were elevated at 16 kHz in 800
ppm-group rats; at 8 and 16 kHz in 1000 ppm-group rats, and at 4, 8, and 16 kHz in 1200 ppm-
group rats (8kHz not tested for BAER threshold determinations). No other indices of toxicity
were investigated. On the basis of increased behavioral auditory and BAER thresholds, the
LOAEL is 800 ppm and the NOAEL is not determined.
Nylen and Hagman (1994) exposed groups of male Sprague-Dawley rats by inhalation to
air containing 0, or 1000 ppm xylene (1.5% o-xylene, 65% m-xylene, 32% p-xylene, 2.5%
ethylbenzene), 1000 ppm n-hexane, or 1000 ppm of n-hexane mixed with xylene, for 18 hours
46
-------
per day, 7 days per week, for 61 days. Two days following exposure, rats exposed to xylene
alone had statistically significantly decreased body weights and a slight loss in auditory
sensitivity, as recorded by auditory brainstem response, compared with controls. Xylene
exposure did not affect flash-evoked potentials or nerve and muscle action potentials measured
in the tail.
As discussed in Section 4.2.1.2. (Jenkins et al., 1970), one of two dogs exposed to 780
ppm o-xylene for 8 hours per day, 5 days per week, for 6 weeks exhibited tremors throughout the
exposure. No additional information was provided.
In a study conducted by Savolainen et al. (1979), groups of 20 male Wistar rats were
exposed to vapors containing 300 ppm xylenes (85% m-xylene, 15% o- and p-xylene) or control
air for 6 hours per day, 5 days per week, for 5 to 18 weeks with or without concomitant exposure
to ethanol in drinking water. Exposure to xylene alone resulted in an increase in microsomal
superoxide dismutase activity in the brain at the end of the exposure and transient decreased
preening frequency.
Rosengren et al. (1986) exposed groups of four male and four female Mongolian gerbils
by continuous inhalation to xylene at 0, 160, or 320 ppm for a period of 3 months, followed by a
4-month postexposure solvent-free period. Xylene exposure caused regional increases in the
brain concentrations of glial fibrillary acidic protein, a main component of astroglial filaments;
S-100 protein (found in fibrillary astrocytes); and DNA. The authors stated that these findings
were compatible with the presence of astrogliosis. No other evaluations, including a recording
of clinical signs, were described.
Studies have also been conducted to assess the potential for xylene exposure in utero to
result in postnatal neurobehavioral deficits. For more information on these studies, the reader is
referred to Section 4.3.2.
4.4.2. Genotoxicity
The genotoxicity of commercial xylenes and all three individual isomers has been
studied, and the results are, for the most part, negative (IARC, 1989). All studies cited in the
GENE-TOX data base are negative with the exception of one study for which no conclusion was
drawn. Xylenes are not mutagenic in bacterial test systems with Salmonella typhimurium (Bos
et al., 1981; Florin et al., 1980; NTP, 1986) and Escherichia coli (McCarroll et al., 1981) or in
cultured mouse lymphoma cells (Litton Bionetics, 1978b). Xylenes do not induce chromosomal
aberrations or sister chromatid exchanges in Chinese hamster ovary cells (Anderson et al., 1990)
or cultured human lymphocytes (Gerner-Smidt and Friedrich, 1978), chromosomal aberrations in
rat bone marrow (Litton Bionetics, 1978b), micronuclei in mouse bone marrow (Mohtashamipur
et al., 1985), or sperm head abnormalities in rats (Washington et al., 1983). Technical grade
xylenes—but not o- and m-xylene—are weakly mutagenic in Drosophila recessive lethal tests
(Donner et al., 1980). No increase in the frequency of sister chromatid exchanges was observed
47
-------
in peripheral lymphocytes in individuals exposed to xylenes in an occupational setting (Haglund
et al., 1980; Pap and Varga, 1987) or an experimental setting (Richer et al., 1993).
4.4.3. Comparison of the Toxicity of Individual Xylene Isomers
Technical-grade mixed xylenes, the form most commonly used as a commercial solvent,
is a blend of three isomers (o-, m-, and p-xylene), and it frequently contains a significant portion
of ethylbenzene. Humans are most likely exposed to a mixture of xylenes rather than to
individual isomers, and it is the solvent used most frequently in toxicity studies. It should be
noted, however, that the composition of the mixture (relative amounts of the individual isomers
and ethylbenzene) varies considerably depending on its source. Results from studies comparing
the toxicity of individual xylene isomers indicate that differences, when they occur, are specific
to the endpoint under consideration.
For oral exposure, minimal comparison studies on the toxicity of individual xylene
isomers are available. Condie et al. (1988) reported no significant differences in body weight
changes in male Sprague Dawley rats given 10 consecutive oral doses of 2000 mg/kg-day of
each of the individual isomers. Furthermore, although the 90-day studies by Wolfe et al. (1988a,
b) reported decreased body weights at 800 mg/kg-day of both m-xylene and p-xylene, Condie et
al. (1988) did not report decreased body weight at 750 mg/kg-day when mixed xylenes were
administered to the same strain of rats for 90 days.
For inhalation exposure, various patterns of toxicity among xylene isomers have been
reported. In rats exposed by inhalation for 30 minutes, EC50s (the concentrations producing half-
maximal decreases in response rate) for effects on an operant behavior test showed a relative
toxicity order of o-xylene > p-xylene > m-xylene, whereas EC50s for a motor performance test
showed a toxicity order of p-xylene > o-xylene = m-xylene. The range of EC50 values among the
isomers was not considered large (Moser et al., 1985). In contrast, p-xylene—but not o-xylene
or m-xylene—altered audiometric variables in rats exposed to 1800 ppm 6 hours per day, 5 days
per week, for 13 weeks (Gagnaire et al., 2001).
A different order of toxicity has been described for effects on motor coordination (rotarod
performance) in rats following 6 hours exposure to concentrations of 3000 ppm xylenes: o-
xylene > m-xylene > p-xylene (Korsak et al., 1990). All three isomers caused decreased fetal
body weights in rats exposed for 24 hours per day on GDs 7-14, but o-xylene caused the effect
at a lower concentration (350 ppm) than did either p-xylene or m-xylene (700 ppm) (Ungvary et
al., 1980). No available studies compare the potency of isomers for affecting neurological
endpoints following subchronic or chronic inhalation exposure.
Some of the studies that have examined the toxicity of individual xylene isomers are
described in more detail below.
48
-------
Condie et al. (1988) did not find any significant differences in the toxicity of the
individual isomers in an experiment in which groups of 10 male and 10 female Sprague-Dawley
rats were administered m-, o-, or p-xylene orally by gavage in corn oil for 10 consecutive days at
doses of 0, 250, 1000, or 2000 mg/kg-day. Two female rats receiving the high dose of p-xylene
died; the deaths were attributed to treatment. Male rats receiving 2000 mg/kg-day of each
isomer had statistically lower body weights (88-94% of controls'), whereas body weights of
high-dose females were not affected. Males and females receiving 2000 mg/kg-day of each
isomer had statistically elevated liver weights and/or liver to body weight ratios (ranging from
128 to 148% of controls'). The authors concluded that there are no significant differences in the
toxicity of the individual isomers.
Moser et al. (1985) evaluated the effects of the individual xylene isomers and a
commercial xylene mixture on operant responding and motor performance in CD-I male albino
mice following 30-minute static inhalation exposures. The minimally effective concentration for
disruption of operant performance (lever-pressing behavior) was 1400 ppm for all isomers, with
an EC50 of 6200, 5200, or 5600 ppm for m-xylene, o-xylene, and p-xylene, respectively. The
minimally effective concentrations for the inverted screen test (mice falling off the screen or
unable to travel the 13 cm distance in 60 seconds) were 3000 ppm for m-xylene and o-xylene
and 2000 ppm for p-xylene, and the EC50 values for performance on the inverted screen test were
3790, 3640, and 2676 ppm for m-xylene, o-xylene, and p-xylene, respectively. There were no
consistent, significant differences in the potency of the individual isomers. Although o-xylene
exhibited a more potent effect on operant behavior, p-xylene more severely affected motor
performance.
Gagnaire et al. (2001) exposed groups of male Sprague Dawley rats (n = 16) to 0, 450,
900, or 1800 ppm of o-, m-, or p-xylene for 6 hours per day, 5 days per week, for 13 weeks.
Audiometric measurements were made using implanted electrodes at weeks 0, 4, 8, and 13 of
exposure and at weeks 4 and 8 postexposure. Neither o-xylene or m-xylene resulted in
detectable changes in audiometric measurements, either during exposure or during the 8 week
postexposure recovery period. In contrast, exposure to 1800 ppm of p-xylene caused significant
decrements in brainstem auditory-evoked potential (2, 4, 8, and 16 KHz) and cochleograms (total
cell count) at every time point examined. No changes were seen in animals exposed to 450 or
900 ppm at any time examined.
In a study by Molnar et al. (1986), motility was assessed in groups of eight CFY white
male rats following exposure by inhalation for 4 hours to at least six concentrations each of m-
xylene, o-xylene, or p-xylene (individual concentrations not provided). Exposure to 130 to 1500
ppm m-xylene and 400 to 1500 ppm p-xylene resulted in a dose-dependent increase in group
motility, whereas exposure to 150 to 1800 ppm o-xylene resulted in a slight depression of
activity. At higher concentrations, however, activity was decreased in all groups, with the
minimum narcotic concentration for the three isomers reported as 2180 ppm for o-xylene, 2100
ppm for m-xylene, and 1940 ppm for p-xylene.
49
-------
Korsak et al. (1990) found that o-xylene, in comparison with other isomers, more
severely affected motor performance. Groups of 10 male Wistar rats were exposed to
approximately 3000 ppm o-, m-, or p-xylene for 6 hours, with rotarod performance measured
before exposure and immediately after termination of the exposure. The results of the testing,
given in terms of the number of failures per number of tested animals, were as follows: for 19/20
for o-xylene at an average concentration of 3027 ppm, 6/20 for m-xylene at an average
concentration of 3093 ppm, and 1/20 for p-xylene at an average concentration of 3065 ppm.
To address the potential for xylene isomers to cause maternal or developmental toxicity,
Ungvary et al. (1980) exposed groups of 15-30 pregnant CFY rats to air containing measured
concentrations of 35, 350, or 700 ppm of o-, m-, or p-xylene continuously during GDs 7-14.
Dams were sacrificed on GD 21. The general conclusion of the study was that exposure to m-
xylene was the most toxic to the dams, whereas fetal toxicity varied with the isomer. m-Xylene
exposure (700 ppm) resulted in a decreased number of mean implantations per dam. p-Xylene
exposure (700 ppm) resulted in increased post-implantation loss and a correspondingly
decreased litter size. All concentrations of p-xylene and the highest concentration of o-xylene
resulted in increased fetal incidence of skeletal retardation. Finally, all isomers caused decreased
fetal body weight: o-xylene at 350 and 700 ppm and m-xylene and p-xylene at 700 ppm.
Fang et al. (1996) determined the Minimum Alveolar Concentration (MAC) (the
concentration that produces anesthesia, i.e., lack of movement, in 50% of those exposed) of the
individual isomers in rats. The MACs of o-, m-, and p-xylene were 0.00118 ± 0.00009, 0.00139
± 0.00010, and 00.00151 ± 0.0007 atm, respectively, with a difference in MAC values of less
than 30% among the isomers.
In summary, although differences in the toxicity of the xylene isomers have been
detected, no consistent pattern following oral or inhalation exposure has been identified.
4.5. SYNTHESIS AND EVALUATION OF MAJOR NONCANCER EFFECTS AND
MODE OF ACTION—ORAL AND INHALATION
4.5.1. Oral Exposure
Human studies following oral exposure to xylenes are not available. Results from several
subchronic studies in rats and mice (Condie et al., 1988; NTP, 1986; Wolfe et al., 1988a, b) and
one chronic study in male rats (NTP, 1986) identify decreased body weight as a potential health
hazard from repeated oral exposure to dose levels generally greater than 500-800 mg/kg-day.
Xylene-induced body weight decreases have been observed most consistently in male rats
(Condie et al., 1988; NTP, 1986; Wolfe et al., 1988a, b). There have been observations of mild
body weight decreases in female rats exposed for 90 days to 800 mg/kg-day m-xylene or p-
xylene (Wolfe et al., 1988a, b). Decreased body weights (93% of controls') were also observed
in mice exposed for 13 weeks to 2000 mg mixed xylenes/kg-day (NTP, 1986). Decreased body
weights were not observed in female rats or in male and female mice exposed for 2 years to
50
-------
doses of xylenes of up to 500 or 1000 mg/kg-day, respectively (NTP, 1986). The mode of action
responsible for this effect has not been studied.
Studies of orally exposed animals provide some evidence for other noncancer effects at a
few sites, but these appear to occur either at dose levels above the lowest levels inducing body
weight changes or inconsistently across studies. Several subchronic studies (Condie et al., 1988;
Wolfe et al., 1988a, b) reported nonneoplastic and organ weight findings. Increased liver
weights, in the absence of histopathologic changes, were observed in one study (Condie et al.,
1988) in male and female rats exposed for 90 days to mixed xylenes at doses > 150 and 750
mg/kg-day, respectively. However, similar liver weight changes were not observed in other
subchronic studies of rats (NTP, 1986; Wolfe et al., 1988a, b) and mice (NTP, 1986) exposed to
doses as high as 1000 and 2000 mg/kg-day, respectively, or in the only available chronic study
of rats and mice exposed to doses as high as 500 and 1000 mg/kg-day, respectively (NTP, 1986).
Increased kidney weights, with associated signs of nephropathy, were observed in female but not
male rats exposed for 90 days to mixed xylenes (Condie et al., 1988), but similar kidney effects
were not observed in the other subchronic studies (NTP, 1986; Wolfe et al., 1988a, b).
Neurological effects from oral exposure to xylenes are a consideration, based on the reports
of neurological symptoms in workers exposed to airborne xylenes. Animal data on oral exposure
to support this assumption are limited and indicate that gross clinical signs and symptoms of
neurological impairment occur only at exposure levels above the lowest levels associated with
body weight changes. In 13-week studies (NTP, 1986), weakness, lethargy, short and shallow
breathing, unsteadiness, tremors, and paresis were observed in mice exposed to 2000 mg/kg-day
mixed xylenes. These symptoms were observed 5-10 minutes after dosing and lasted for 15-60
minutes. Similar transient neurological symptoms were not observed or were not reported in
mice at lower exposure levels or in rats exposed to 1000 mg/kg-day. In 2-year studies (NTP,
1986), hyperactivity 5-30 minutes after dose administration was observed in mice at 1000
mg/kg-day, but not at 500 mg/kg-day. Other subchronic rat studies (Condie et al., 1988; Wolfe
et al., 1988a, b) reported no gross clinical or histological findings indicative of neurological
impairment. The oral data base is limited in that comprehensive examinations of persistent
changes in neurobehavior (e.g., a functional observational battery) following acute or repeated
oral exposure to xylenes have not been conducted.
Results from two animal studies indicate that developmental effects are a potential hazard
from oral exposure to xylenes, but the effects occur at dose levels greater than those inducing
body weight changes. Cleft palate formation was reported in the fetuses of mice exposed to
2060 mg/kg-day—but not 1030 mg/kg-day—on GDs 6-15 (Marks et al., 1982). Another
gestational exposure study of mice (Nawrot and Staples, 1980) (reported only as an abstract) also
reported cleft palate formation at 1960 mg/kg-day but not at 780 mg/kg-day. Maternal toxicity
was not sufficiently evaluated in either of these studies to identify maternal NOAELs and
LOAELs. No available studies have examined the potential effects of oral exposure of xylenes
on reproductive endpoints.
4.5.2. Inhalation Exposure
51
-------
The weight of evidence from limited human data and more extensive animal data indicates
neurological impairment and developmental effects as potential health hazards from repeated
inhalation exposure to xylenes. Reversible symptoms of neurological impairment and irritation
of the eyes and throat are well-known health hazards from acute inhalation exposure to xylenes
and other aromatic solvents. In general, these acute effects are expected to involve reversible
molecular interactions of the solvent itself (not metabolites) with membranes of the affected
tissues, including neuronal membranes, and are most pronounced at high exposure levels (in
excess of 1000 ppm). At lower concentrations, more subtle effects may occur. Human
volunteers exposed under controlled conditions to xylene concentrations in the range of 200-400
ppm for short time periods (15 minutes to 4 hours) reported symptoms of irritation (e.g.,
watering eyes and sore throat) or neurological impairment (e.g., mild nausea, headache)
(Carpenter et al., 1975; Gamberale et al., 1978).
In other studies involving single or multiple 4-hour exposures of human volunteers to 200
ppm xylene (Laine et al., 1993; Savolainen and Linnavuo, 1979; Savolainen et al., 1984),
reversible effects on balance and reaction times were reported. However, other studies of 4-hour
exposures to 200 ppm did not find impaired performance in tests of simple reaction time, short-
term memory, and choice reaction time (Olson et al. 1985) or changes in visually evoked brain
potentials (Seppalainen et al., 1983) or EEG patterns (Seppalainen et al., 1991). Impaired
performance on tests of memory and reaction times was reported for subjects exposed to 100
ppm xylene for 4 hours (Dudek et al., 1990).
The available controlled-exposure human studies indicate that concentrations around
100-200 ppm are close to the threshold level for short-term reversible neurological and irritation
effects from xylenes. Available human data alone provide inadequate evidence for neurological
impairment from repeated exposure to xylene concentrations less than or equal to 200 ppm.
Aside from the controlled-exposure studies reviewed above, most of the human data associating
xylene exposure with neurological impairment are case reports involving acute high-level
exposures (800-10,000 ppm) (e.g., Goldie, 1960; Hipolito, 1980; Klaucke et al., 1982).
Epidemiologic studies are restricted to a cross-sectional health evaluation study (Uchida et
al., 1993) that noted increased prevalence of self-reported neurological symptoms and
irritation—but no apparent changes in serum enzymes indicative of liver or kidney damage—in a
group of Chinese workers. The workers were from a boot manufacturing plant that used a
xylene-containing glue and two other plants that used mixed xylenes as a solvent in wire
production or printing. The measured time-weighted-average (TWA) mean concentration of
airborne xylenes in these workplaces was 21 (± 21) ppm. The study has several limitations,
including a lack of reporting on the duration of exposure, co-exposure to other chemicals, no
clear demonstration of relationships between response and dose or duration, and the inherent bias
presented by self-reporting of symptoms.
Although the human evidence for persistent effects on the nervous system or other persistent
effects from repeated inhalation exposure to xylenes is inadequate, results from animal studies
52
-------
more clearly identify potential persistent neurological impairment and possible developmental
effects as potential health hazards from repeated inhalation exposure.
A number of subchronic studies in animals provide evidence for neurological effects
following repeated inhalation exposure to xylenes (Table 2). The lowest exposure level that
produced changes in a number of neurological endpoints was identified in several studies of rats
exposed to 100 ppm m-xylene, 6 hours per day, 5 days per week, for 3 months. These studies
observed statistically significant changes in neurobehavioral tests conducted at least 24 hours
following cessation of exposure: decreased rotarod performance indicative of impaired motor
coordination (Korsak et al., 1992, 1994), decreased spontaneous motor activity (Korsak et al.,
1992), impaired radial maze performance indicative of a learning deficit (Gralewicz et al., 1995),
and decreased latency to paw lick in the hot plate test, indicating increased sensitivity to pain
(Korsak et al., 1994).
At a lower exposure level (50 ppm by the same exposure protocol), rats showed statistically
significantly decreased latency in the paw lick response but no statistically significant effects on
rotarod performance (Korsak et al., 1994). Rats exposed to 100 ppm m-xylene by the same daily
protocol for a shorter duration (4 weeks) displayed no statistically significant changes in tests of
radial maze performance, open-field activity, or active avoidance, but paw lick latency was
increased in the hot plate test and step-down time was shortened in 1/6 trials in the passive
avoidance test (Gralewicz and Wiaderna, 2001).
Table 2. Neurological effects of xylenes following subchronic inhalation
exposure of adult male rats
Study
Korsak etal. (1994)
Korsak etal. (1992)
Duration
6 hrs/day,
5 days/wk,
3 months
6 hrs/day,
5 days/wk,
6 months
NOAEL
(ppm)
50
None
LOAEL
(ppm)
100
100
Effect
Impaired rotarod
performance, decreased
latency of paw-lick
response
Impaired rotarod
performance, decreased
motor activity
53
-------
Gralewicz et al.
(1995)
Gralewicz and
Wiaderna(2001)
Pryoretal. (1987)
Nylen and Hagman
(1994)
6 hrs/day,
5 days/wk,
3 months
6 hrs/day,
5 days/wk,
4 weeks
7 hrs/day,
7 days/wk,
6 weeks
18 hrs/day,
7 days/wk,
61 days
None
None
None
None
100
100
800
1000
Altered radial maze
performance
Altered passive avoidance
test, increased latency of
paw-lick response
Hearing loss
Decreased auditory
sensitivity
Overall results from these rat studies provide evidence that repeated exposure to m-xylene
at concentrations > 100 ppm (6 hours per day, 5 days per week) may produce persistent changes
in several neurologic endpoints in adult rats. The mode of action of these changes has not been
studied. Supporting evidence for persistent neurologic effects from xylenes exposure includes
reports of changes in indices of hearing loss in rats exposed to >800 ppm mixed xylenes 14
hours per day for 6 weeks (Pryor et al., 1987) and in rats exposed to 1000 ppm mixed xylenes 18
hours per day, 7 days per week, for 61 days (Nylen and Hagman, 1994).
There are no available studies on the possible developmental toxicity of inhaled xylenes in
humans, but a number of studies have examined standard developmental toxicity endpoints and
neurobehavioral endpoints in offspring of animals exposed to mixed xylenes or individual xylene
isomers. Pertinent developmental effect levels in animals exposed by inhalation to xylenes
during gestation are summarized in Table 3.
-------
Table 3. Pertinent developmental effect levels in animals exposed by inhalation to xylenes
during gestation
Species
Rat
Effect/Endpoint
Impaired cognitive (but not motor) performance in Morris water
maze by female (but not male) offspring at 500 ppm
No effects on rotarod performance by offspring at 500 ppm
No effects on acoustic startle response or figure-8 maze activity
in offspring, PNDs 22 and 65 at concentrations up to 1600 ppm
Effects on fetal skeletal and visceral malformations or
variations, fetal BW, or maternal BW gain at concentrations up
to 400 ppnf
No effects on fetal skeletal and visceral malformations or
variations, fetal BW, maternal BW, or maternal fertility at 500
ppnf
Decreased maternal BW gain at 1600 ppm; no effects on litter
size, pup birth weight, or postnatal growth rate
Maternal toxicity (decreased BW) at 700 but not 350 ppm;
decreased fetal BW at 350 and 700 ppm but not 35 ppm; effects
on fetal skeletal and visceral malformations or variations"
Maternal toxicity (decreased BW) at 700 but not 350 ppm;
decreased fetal BW at 700 ppm but not 350 or 35 ppm; effects
on fetal skeletal and visceral malformations or variations8
Maternal toxicity (decreased BW) at 700 but not 350 ppm; post-
implantation loss, decreased litter size, and decreased fetal BW
at 700 ppm but not 350 or 35 ppm; effects on fetal skeletal and
visceral malformations or variations'1
NOAEL
(ppm)
None
identified
500
1600
400
500
800
35
350
350
LOAEL
(ppm)
500
None
identified
None
identified
None
identified
None
identified
1600
350
700
700
Exposure
0, 500 ppm mixed xylenes, 6
hrs/day, CDs 7-20
0, 500 ppm mixed xylene, 6
hrs/day, CDs 7-20
0, 800, 1600 ppm p-xylene, 6
hrs/day, CDs 7-16
0, 100, 400 ppm mixed
xylenes 6 hrs/day, GDs 6-15
0, 500 ppm mixed xylenes 6
hrs/day for 131 days prior to
mating and continuing to GD
2 lor through lactation
0, 800, 1600 ppm p-xylene, 6
hrs/day, GDs 7-16
0,35, 350, 700 ppm o-
xylene, 24 hrs/day GDs 7-14
0,35, 350, 700 ppm m-
xylene, 24 hrs/day GDs 7-14
0,35, 350, 700 ppm p-
xylene, 24 hrs/day GDs 7-14
Reference
Hass et al.
(1995, 1997)
Hass et al.
(1995)
Rosen et al.
(1986)
Litton
Bionetics
(1978a)
Bio/dynamics
Inc. (1983)
Rosen et al.
(1986)
Ungvary et al.
(1980)
Ungvary et al.
(1980)
Ungvary et al.
(1980)
-------
Table 3. Pertinent developmental effect levels in animals exposed by inhalation to xylenes
during gestation (continued)
Species
Rat
(cont'd)
Mouse
Rabbit
Effect/Endpoint
Maternal toxicity not clearly reported; increased proportion
dead or resorbed fetuses at 780 ppm; effects on fetal skeletal
and visceral malformations or variations3
Maternal toxicity not clearly reported; effects on fetal skeletal or
visceral malformations or variations'1
Maternal toxicity not clearly reported; effects on fetal skeletal or
visceral malformations or variations3
No live fetuses produced at 230 ppm and three dams died,
indicating severe maternal toxicity; no statistically significant
effects on abortions, live fetus numbers, or fetal skeletal or
visceral malformations or variations at 115 ppm; mean female
(but not male) fetal BW decreased by about 10% when
compared with controls at 1 15 ppm
No effects on fetal survival or fetal skeletal or visceral
malformations or variations at 1 15 ppnf ; at 230 ppm p-xylene,
no live fetuses produced and one dam died, indicating severe
maternal toxicity; report did not specify whether rabbits were
exposed to o- or m-xylene at 230 ppm
NOAEL
(ppm)
440
230
115
None
identified
115
LOAEL
(ppm)
780 ppm
None
identified
None
identified
115 (body
weight)
230 (dead
fetuses)
230
Exposure
0, 60, 440, 780 ppm mixed
xylenes, 24 hrs/day GDs
7-15
0, 115, 230 ppm mixed
xylenes, three 4-hour periods
daily on GDs 6-15
0, 1 15 ppm o-, p-, or m-
xylene, three 4-hour periods
daily on GDs 6-15
0, 115, 230 ppm mixed
xylenes, 24 hrs/day GDs
7-20
0, 115 ppmo-, p, orm-
xylene, 24 hrs/day GDs
7-20; 230 ppm p-xylene 24
hrs/day GDs 7-20
Reference
Ungvary and
Tatrai (1985)
Ungvary and
Tatrai (1985)
Ungvary and
Tatrai (1985)
Ungvary and
Tatrai (1985)
Ungvary and
Tatrai (1985)
a Statistically significant increases in incidences of fetuses with retarded skeletal ossification or with extra ribs were reported as compared with controls.
However, for all of the studies except Litton Bionetics (1978a), the incidences were reported on a total-fetus-per-exposure-group basis; no litter-specific
data were reported. These effects were judged to be difficult to interpret, given the inability to adjust for possible litter size covariation. For example, after
adjustment for covariance with litter size, the incidences of fetuses with delayed ossification in rats exposed to 400 ppm 6 hrs/day in Litton Bionetics
(1978a) were no longer significant. The statistically significantly elevated incidences of skeletal anomalies or retardation were not considered adverse in
determining developmental NOAELs and LOAELs in these studies. See Section 4.3.2 for more discussion.
-------
Table 3. Pertinent developmental effect levels in animals exposed by inhalation to xylenes
during gestation (continued)
PND = postnatal day
BW = body weight
-------
Evidence exists for impaired neurological development in rat offspring following
gestational exposure to inhaled xylenes, but it is not strong. Changes in neurobehavioral
variables reported for offspring of animals exposed during gestation are restricted to impaired
cognitive (but not motor) performance in the Morris water maze test in female but not male
offspring of rats exposed to 500 ppm mixed xylenes 6 hours per day on GDs 7-20 (Hass et al.,
1995, 1997) and decreased rotarod performance in offspring of rats exposed to 200 ppm
"technical" xylene 6 hours per day on GDs 6-20 (Hass and Jakobsen, 1993). Deficits in the
water maze test were observed only in female rat offspring raised in standard housing and not in
female rats raised in "enriched" housing with various toys (Hass et al., 1995).
Although decreased rotarod performance by offspring was observed in the study by Hass
and Jakobsen (1993), it was not observed in the later study by the same group of investigators
(Hass et al., 1995). The reported effect on rotarod performance in the earlier study was
questioned by Hass et al. (1995) because it was not conducted by experimenters who were blind
to the exposure status of the rats. In addition, offspring of rats exposed to 800 or 1600 ppm p-
xylene, 6 hours per day on GDs 7-16 performed similarly to offspring of nonexposed rats in tests
of CNS development—an acoustic startle response test on PNDs 13, 17, 21, and 63 and a figure-
8 maze activity test on PNDs 22 and 65 (Rosen et al., 1986).
Several other inhalation studies have examined standard developmental toxicity endpoints
in rats (Litton Bionetics, 1978a; Bio/dynamics Inc., 1983; Rosen et al., 1986; Ungvary et al.,
1980; Ungvary and Tatrai, 1985) and mice and rabbits (Ungvary and Tatrai, 1985) following
gestational exposure to xylenes (Table 3). These studies identified maternally toxic levels for
decreased body weight gain in pregnant rats at doses greater than or equal to 700 ppm o-, p-, or
m-xylene 24 hours per day (Ungvary et al., 1980) or 1600 ppm p-xylene 6 hours per day (Rosen
et al., 1986) and for maternal death and abortions in pregnant rabbits exposed to 230 ppm (but
not 115 ppm) mixed xylenes or p-xylene for 24 hours per day (Ungvary and Tatrai, 1985).
In rats, effects on fetal skeletal and visceral malformations (such as cleft palate) and
variations (such as retarded skeletal ossification or extra ribs) were reported at concentrations of
up to 700 ppm o-, m-, or p-xylene 24 hours per day (Ungvary et al., 1980) or 780 ppm mixed
xylenes 24 hours per day. Statistically significant increased incidences of fetuses with retarded
skeletal ossification or extra ribs were reported in these studies, but on an exposure-group basis.
No litter-specific information was provided except in the Litton Bionetics (1978a) study.
The most significant effects on developmental endpoints were decreased fetal body weight
or fetal survival in rats at xylene isomer doses of 350 or 700 ppm 24 hours per day (Ungvary et
al., 1980) or at mixed xylenes concentration of 780 ppm 24 hours per day (Ungvary and Tatrai,
1985) and increased abortions in rabbits exposed to 230 ppm 24 hours per day (Ungvary and
Tatrai, 1985). These effects, although of concern, occurred at concentrations above those at
which neurobehavioral effects were reported in adult animals (see Table 2).
Although the mechanism by which xylenes exert their toxic effects on the nervous system
and developing fetus are not completely understood, some theories exist. The CNS toxicity
58
-------
observed during exposure to high concentrations has been attributed to the liposolubility of
xylenes in the neuronal membrane (Desi et al., 1967; Savolainen and Pfaffli, 1980; Tahti, 1992).
It has been suggested that xylenes disturb the actions of proteins essential to normal neuronal
function. Changes in levels of various neurotransmitters and lipid composition have been
observed in several brain areas following acute- and intermediate-duration exposure to xylenes
(Andersson et al., 1981; Honma et al., 1983). It is unclear whether these represent direct effects
of xylenes or are secondary changes resulting from nonspecific CNS depression. Some authors
have also suggested that metabolic intermediates such as arene oxides or methylbenzaldehyde
may be responsible for the toxic effects of xylenes (Savolainen and Pfaffli, 1980).
The mechanism by which xylenes produce toxic effects in fetuses has not been fully
investigated. The p-xylene-induced delayed fetal development found in studies with rats
(Ungvary et al., 1981) may have been caused by decreased levels of progesterone and estradiol.
The decreased hormonal levels may have been due to increased microsomal activity and
increased hormonal catabolism.
In summary, human data are suggestive of neurological effects and irritation of the eyes and
respiratory tract following inhalation exposure to xylenes. Animal studies have demonstrated
that neurological effects are the most sensitive effects of xylenes inhalation exposure, with
measurable effects in several neurobehavioral endpoints beginning at concentrations as low as
100 ppm following subchronic exposure (Gralewicz et al., 1995; Korsak et al., 1992, 1994;
Nylen and Hagman, 1994; Pryor et al., 1987). At higher exposure levels, changes in body
weight have been reported in some studies (Tatrai and Ungvary, 1980; Tatrai et al., 1981) but not
in others (Carpenter et al., 1975; Jenkins et al., 1970; Ungvary, 1990).
Similarly, high-dose exposure to xylenes has resulted in changes in liver morphology,
weight, and enzymatic functions (Tatrai and Ungvary, 1980; Tatrai et al., 1981; Ungvary, 1990).
Gestational exposure of animals to xylenes has resulted in neurodevelopmental effects (Hass et
al., 1995; 1997; Hass and Jakobsen, 1993) and other possible developmental effects (Ungvary et
al., 1980; Ungvary and Tatrai, 1985), but only at levels above those at which neurobehavioral
effects in adult male rats were reported.
4.6. WEIGHT-OF-EVIDENCE EVALUATION AND CANCER CHARACTERIZATION
Under the Draft Revised Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1999) data
are inadequate for an assessment of the carcinogenic potential of xylenes. Adequate human
data on the carcinogenicity of xylenes are not available, and the available animal data are
inconclusive as to the ability of xylenes to cause a carcinogenic response. Evaluations of the
genotoxic effects of xylenes have consistently provided negative results.
Data on the carcinogenicity of xylenes following inhalation exposure are limited. A number
of human occupational studies have suggested possible carcinogenic effects of chronic inhalation
exposure. However, in each case co-exposure to other chemicals was a major confounder,
59
-------
leading to an inability to adequately assess the potential effects of chronic exposure. Animal
data on the carcinogenicity of xylenes following inhalation exposure are not available.
Examinations of the carcinogenicity of xylenes following oral exposure in humans are not
available. NTP (1986) conducted a 2-year oral cancer bioassay in male and female Fischer 344
rats and male and female B6C3F1 mice. Rats were exposed to 0, 250, or 500 mg/kg-day of
mixed xylenes by gavage for 5 days per week for 103 weeks. No evidence of carcinogenesis
was seen in male or female rats. Similarly, mice exposed to 0, 500, or 1000 mg/kg-day for 2
years did not show evidence of carcinogenic effects (NTP, 1986). However, a study by Maltoni
et al. (1983, 1985) reported an increase in the overall number of malignant tumors in male and
female rats treated by gavage with 0 or 500 mg/kg mixed xylenes for 4 days per week for 104
weeks. However, only total tumor incidence was reported; descriptions of target organs and
tumor types were not included in the report. In the absence of additional information, and
because only one dose was used, the Maltoni et al. (1983, 1985) study does not provide
sufficient evidence of the carcinogenicity of xylenes in animals.
4.7. SUSCEPTIBLE POPULATIONS
No definitive data addressing susceptible populations are available.
Available human data are not adequate to determine whether a gender-related difference in
susceptibility to xylene exists. Differences in sensitivity between males and females in animal
studies have been inconsistent across species or endpoint. For example, male but not female rats
consistently showed decreased body weight changes following repeated oral exposure to mixed
xylenes in several studies (NTP, 1986; Condie et al., 1988), whereas body weight changes were
observed in both male and female mice exposed to 2000 mg/kg/day mixed xylenes for 13 weeks
(NTP, 1986). In contrast, female but not male offspring of rats exposed by inhalation to 500
ppm mixed xylenes 6 hours per day on GDs 7-20 showed reversible impairment of learning in
the Morris water maze test (Hass et al., 1995; 1997). However, the adversity of this effect
appears to be minimal, because only one group of female offspring (those housed under standard
conditions without various toys) showed the effect and the impairment appeared to diminish with
time. In addition, other neurological endpoints (acoustic startle response or figure-8 maze
activity) were not affected in either male or female offspring of rats exposed to up to 1600 ppm
p-toluene 6 hours per day on GDs 6-15 (Rosen et al., 1986).
For other relevant endpoints, possible differences between genders have not been
adequately studied. For example, other critical neurological endpoints associated with repeated
inhalation exposure of adult animals to xylenes have been studied only in male rats (Pryor et al.,
1987; Korsak et al., 1992, 1994; Gralewicz et al., 1995; Gralewicz and Wiaderna, 2001), and
acute controlled exposure studies with human subjects have involved only men (e.g., Gamberale
et al., 1978; Savolainen andLinnavuo, 1979; Rhiihimaki and Savolainen, 1980; Savolainen et
al., 1984; Olson et al., 1985; Dudek et al., 1990; Laine et al., 1993).
60
-------
As discussed in Sections 4.5.1 and 4.5.2, available studies of the developmental toxicity of
xylenes do not suggest that the developing organism is more sensitive than the adult to xylene
exposure. Similarly, no studies are available that have examined whether children may be more
susceptible than adults to the effects of xylenes exposure.
5. DOSE-RESPONSE ASSESSMENTS
5.1. ORAL REFERENCE DOSE (RfD)
5.1.1. Choice of Principal Study and Critical Effect
No studies of the toxicity of xylenes in humans following subchronic or chronic oral
exposure are available. Available animal studies do not identify a single sensitive health effect
other than changes in body weight, which have been seen consistently in male rats across several
studies. Decreased body weights (5-8% decrease relative to controls) were observed in male but
not female rats exposed to 500 mg/kg-day of mixed xylenes by gavage for 2 years (NTP, 1986).
Subchronic studies have also reported decreased body weights: an 11% decrease (compared with
controls) in male but not female rats exposed to 1000 mg/kg-day of mixed xylenes for 13 weeks
(NTP, 1986); a 6% decrease in male but not female rats exposed to 1500 mg/kg-day of mixed
xylenes for 90 days (Condie et al. 1988); a 15% decrease in male rats exposed to 800 mg/kg-day
m-xylene for 90 days (Wolfe 1988a); and a 7% decrease in mice exposed to 2000 mg/kg-day
mixed xylenes for 13 weeks (NTP, 1986). No changes in body weight were reported in mice
exposed for 2 years to mixed xylenes at doses as high as 1000 mg/kg-day (NTP, 1986).
Mortality was dose-related and statistically significantly increased (p=0.04) in high-dose
male rats following chronic exposure to mixed xylenes in the NTP (1986) study. Total survival
rates (treatment- and gavage-related) were vehicle-control, 36/50; 250 mg/kg-day, 26/50; and
500 mg/kg-day, 20/50. A number of deaths were reported as gavage related. Treatment-related
mortality rates were vehicle-control, 11/50; 250 mg/kg-day, 17/50; and 500 mg/kg-day, 19/50.
The authors noted that behavioral effects were not recorded during gavage. Survival rates of
female rats and both sexes of mice were not statistically significantly different from those of
controls; no treatment-related trends were evident.
Increased mortality was also observed in rats in two subchronic studies following exposure
to m-xylene (Wolfe et al., 1988a) and p-xylene (Wolfe et al., 1988b). In the first study,
statistically significantly increased mortality was observed in the mid-dose (200 mg/kg-day)
male and mid- and high-dose (800 mg/kg-day) female rats. In the second study, mortality in
high-dose male rats (800 mg/kg-day) attained statistical significance, and a statistically
significant trend was present in the male dose groups. In both studies, early mortality may have
resulted from treatment-related aspiration of the test material, as evidenced by lung congestion.
61
-------
Other effects of oral exposure to xylenes appear to occur at greater exposure levels than
those associated with decreased body weight and increased mortality. Adverse kidney effects
were noted at 750 mg/kg-day but not at 150 mg/kg-day in the 90-day study by Condie et al.
(1988). Kidney effects were not found in the NTP (1986) bioassay with Fisher 344/N rats or
B6C3F1 mice exposed to xylenes for 13 weeks or 2 years. Likewise, no nephropathy was
reported in a nephrotoxicity screening assay in male Fischer 344/N rats exposed to 2000 mg/kg
m-xylene for 5 days per week for 4 weeks (Borriston Laboratories, Inc., 1983). In addition, no
kidney effects were found in Sprague-Dawley rats exposed for 90 days to m-xylene or p-xylene
at doses as high as 800 mg/kg-day (Wolfe et al., 1988a, b). Thus, the available data do not
consistently identify the kidney as a sensitive target of xylenes in animals. Likewise, the
available data do not consistently identify the liver as a sensitive target of xylenes in animals
(NTP, 1986; Wolfe et al., 1988a, b; Condie et al., 1988).
The only neurobehavioral effects noted in any of the available studies occurred in the 2-year
NTP (1986) study in mice, which noted hyperactivity immediately following gavage in animals
exposed to 1000 mg/kg-day but not 500 mg/kg-day. Developmental effects (decreased fetal
body weight and an increase in the incidence of cleft palate) occurred at exposure levels of
-2000 mg/kg-day (Marks et al., 1982; Nawrot and Staples, 1980). An adequate examination of
the effects of oral xylenes exposure on reproductive endpoints has not been reported.
The 2-year study in rats conducted by NTP (1986) was selected as the principal study for
the derivation of the RfD for xylenes because it is the only oral animal study of chronic duration
and because some effects (decreased body weight and possible increased mortality) were evident
at doses lower than those that produced effects seen in other studies. The body weight decrease
(5-8% of controls') is considered to be of marginal biological significance, but there was a
statistically significant trend for decreased survival in male rats with increasing exposure levels,
and survival in the high-dose males was statistically significantly decreased when compared with
controls. Given the possibility of treatment-related frank toxicity, it is not considered prudent to
discount the only other observed effect, that is, decreased body weight. Thus, the highest dose in
the study, 500 mg/kg-day, is considered a LOAEL for changes in body weight and mortality.
In the study selected for the principal study, animals were exposed to mixed xylenes. As
discussed in Section 4.4.3., although some differences in toxicity were apparent among the
individual isomers, there is no consistent evidence for quantitative differences in the potency of
the isomers following oral or inhalation exposure.
The selection of the principal study is the same as reported in the previous IRIS assessment.
The co-critical effects are the same with the exception of hyperactivity in mice at the high dose
of 1000 mg/kg-day. This effect has not been included as a critical effect in the current
assessment due to the higher dose required to achieve the effect.
5.1.2. Methods of Analysis
62
-------
The data were analyzed using the NOAEL/LOAEL approach. Because neither individual
body weight data nor group means and standard deviations for body weights were reported in the
NTP (1986) report, a benchmark dose (BMD) approach was not feasible. However, a BMD
analysis was applied to the body weight effects seen in the subchronic study by Wolfe (1988a)
for comparison purposes, resulting in an RfD similar to one derived from the NOAEL/LOAEL
approach using the NTP (1986) chronic rat data (see Appendix C). The NOAEL/LOAEL
approach using the chronic data and the comparison BMD analysis using the subchronic data are
described in Section 5.1.3.
5.1.3. Oral Reference Dose Derivation
The NTP (1986) study identified a NOAEL of 250 mg/kg-day for decreased body weight in
male rats exposed by gavage 5 days per week for 2 years. The dose was duration adjusted as
follows:
NOAEL[ADJ] = NOAEL x 5 days/7days
= 250 mg/kg-day x 5/7
= 179 mg/kg-day
To the duration-adjusted NOAEL of 179 mg/kg-day, a total uncertainty factor (UF) of
1000 was applied to derive the RfD:
RfD = NOAEL[ADJ] - (UF x MF)
= 179 mg/kg-day -(1000x1)
= 0.2 mg/kg-day
The total UF of 1000 (lOx lOx 10) was derived by applying a UF of 10 to account for
laboratory animal-to-human interspecies differences. No information is available to support a
change from default.
A UF of 10 was applied for intraspecies uncertainty to account for human variability and
sensitive populations. This factor accounts for humans who may be more sensitive than the
general population to exposure to xylenes.
A UF of 10 was used to account for data base uncertainty. The available oral data base
for xylenes includes chronic and subchronic gavage toxicity studies in mice and rats and a
developmental toxicity study. None of these studies indicate that additional data would result in
a lower RfD. However, the data base lacks adequate studies of the oral neurotoxicity of xylenes
as well as multigenerational reproductive toxicity and developmental neurotoxicity studies.
Given the identification of neurological impairment as a critical health hazard from inhalation
exposure to xylenes, the lack of comprehensive neurotoxicity testing following chronic oral
exposure is of particular concern. It should be noted that transient neurotoxic effects (e.g.,
lethargy, tremors and unsteadiness) were reported in mice following oral exposure to xylenes for
63
-------
13 weeks (NTP, 1986). There are no toxicokinetic data identifying oral dose levels at which
first-pass hepatic metabolism of xylenes becomes saturated in animals or humans; such data
could decrease uncertainty regarding whether or not neurological impairment may occur at dose
levels below those causing body weight decreases and mortality in rats. It is uncertain whether
the availability of comprehensive oral neurotoxicity data would result in a lower RfD.
An additional uncertainty associated with the oral data base is that the majority of studies
examined mixed xylenes, which are known to contain ethylbenzene. The present IRIS
assessment for ethylbenzene (U.S. EPA, 2002), which was entered on the data base in 1987, cites
effects on the liver and kidney as the most sensitive endpoints following oral exposure. As
discussed above, effects on the liver and kidney have been reported following oral exposure to
mixed xylenes, but the most sensitive effect reported in animal bioassays is decreased body
weight and increased mortality, as identified by the principal study (NTP, 1986). However,
because the mechanism behind the critical effect has not been clearly elucidated, a possible
contribution of ethylbenzene to the toxicity of mixed xylenes cannot be entirely eliminated.
Additional studies comparing the toxicity of mixed xylenes with that of the individual isomers
would better inform the data base.
The RfD is based on a NOAEL from a chronic study, which obviates the need for a UF
due to LOAEL to NOAEL extrapolation or subchronic extrapolation.
An RfD was derived for comparison purposes by applying the BMD methodology to the
data from the subchronic rat study of Wolfe et al. (1988a) (see Appendix C). Modeling the body
weight data with a linear model for continuous data results in a BMDL (lower bound of the
benchmark dose) of 440 mg/kg-day, using a 10% change as the benchmark response. Derivation
of an RfD from this point of departure using the same UFs as above but with an additional UF of
3 for extrapolation from subchronic to chronic duration (total UF of 3000) would result in a
value of 0.1 mg/kg-day, which is similar to the proposed RfD.
The RfD generated in this assessment differs from the previous RfD (2 mg/kg-day). The
difference is accounted for by the addition of a data base UF, which was not considered in the
previous assessment.
5.2. INHALATION REFERENCE CONCENTRATION (RfC)
5.2.1. Choice of Principal Study and Critical Effect
As discussed in Section 4.5.2., the weight of evidence from limited human data and more
extensive animal data identify mild neurological impairment as a sensitive potential health
hazard from repeated inhalation exposure to xylenes in the concentration range of 100 to 200
ppm. Evidence for biologically significant developmental toxic effects comes from studies of
animals exposed to higher concentrations. Toxicity studies in animals (e.g., Carpenter et al.,
1975; Jenkins et al., 1970) have not found consistent evidence for other effects such as changes
in body weight or hepatic, hematologic, or renal toxicity endpoints following exposure to
64
-------
concentrations as high as 800-1000 ppm for 6 hours per day, 5 days per week. In addition, no
effects on reproductive performance or histology of reproductive organs were found in CD rats
exposed to 500 ppm mixed xylenes for 6 hours per day, 5 days per week, prior to mating and
continuing through the end of lactation (Bio/dynamics Inc., 1983).
The discussion of the available data (predominately from animal studies) in Section 4.5.2
illustrates that the weight of evidence for mild neurological impairments following repeated
inhalation exposure to xylene concentrations below 200 ppm is greater than the weight of
evidence for developmental effects.
Because available human data are insufficient for derivation of an RfC and chronic
animal inhalation data are lacking, the subchronic study by Korsak et al. (1994) was selected as
the principal study. Neurological effects (impaired motor coordination) were selected as the
critical effect for derivation of the RfC. Two neurological endpoints were evaluated in Korsak et
al. (1994). Rotarod performance was statistically significantly decreased (33% from controls') at
100 ppm, whereas a statistically significant decreased sensitivity to pain was observed at 50 and
100 ppm (8.6 and 8.7 seconds, respectively, vs. 12.2 seconds for controls; measurements were
made 24 hours postexposure). Gralewicz and Wiaderna (2001) also measured the effect of m-
xylene exposure (6 hours per day, 5 days per week, for 4 weeks; neurological endpoints
measured on post-exposure day 50) on pain sensitivity. In this study, a statistically significant
increase in pain sensitivity (35 seconds vs. 10 seconds in control) was found at the 100 ppm
dose, the lowest dose tested. The variation in the response to m-xylene in these two studies
decreases the confidence in using the pain sensitivity endpoint as the critical effect.
A number of statistically significant neurological effects have been noted in male rats at a
dose of 100 ppm m-xylene in other supporting studies: decreased rotarod performance and
spontaneous movement activity following exposure for 6 hours per day, 5 days per week, for 6
months (Korsak et al., 1992); decreased radial maze performance following exposure for 6 hours
per day, 5 days per week for 3 months (Gralewicz et al., 1995); and shortened step-down time in
the passive avoidance test following exposure for 6 hours per day, 5 days per week, for 4 weeks.
All the studies measured neurological endpoints 24 hours postexposure, with the exception of
Gralewicz and Wiaderna (2001), which measured effects at postexposure day 50. A NOAEL of
50 ppm and a LOAEL of 100 ppm are identified for neurological effects (impaired motor
coordination).
The principal study (Korsak et al., 1994) reported no statistically significant exposure-
related changes in body weight gain, absolute or relative organ weights, hepatic activities of
monoxygenases or lipid peroxidation, or levels of triglycerides in the liver. Compared with
controls, exposed rats showed statistically significant changes in red blood cell counts (7-20%
decrease), hemoglobin levels (-8% decrease), and white blood cell counts (35% increase).
Effects in red blood cell counts and hemoglobin levels were observed at 50 ppm; however, these
changes were not observed in another study from the same laboratory in rats exposed to 1000
ppm m-xylene (Korsak et al., 1992). Furthermore, effects on erythrocytes were not found at
concentrations of 78-810 ppm in other studies (Carpenter et al., 1975; Jenkins, 1970).
65
-------
There is some uncertainty associated with selecting a principal study for xylenes that
involved exposure to m-xylene alone, but m-xylene is generally the predominant isomer in
commercial mixtures. In addition, although there are no studies comparing the effects of xylene
isomers on critical neurological endpoints following subchronic or chronic inhalation exposure,
the potencies of individual xylene isomers were similar in affecting neurobehavior, as shown in a
study of rats following acute exposures (Moser et al., 1985).
5.2.2. Methods of Analysis
A NOAEL/LOAEL approach was used to derive the RfC. A benchmark concentration
(BMC) was not derived because the principal study reported the neurobehavioral data only as
group means; standard deviations or standard errors for groups or for the individual animal data
were not presented. In the absence of numerical and statistical information, BMC modeling
cannot be performed. Validated PBPK models for xylene inhalation for rats and humans have
been developed (Haddad et al., 1999; Tardif et al., 1991, 1992, 1993 a, b, 1995) and were used
for comparison purposes in the calculation of a NOAEL[HEC/PK]. The results of the calculations
are presented in Section 5.2.3.2 and Appendix B.
5.2.3. Inhalation Reference Concentration Derivation
5.2.3.1. Principal RfC Derivation
The NOAEL of 50 ppm (217 mg/m3) was duration adjusted as follows:
The NOAEL[ADJ] was used to derive a human equivalent concentration (HEC), as
NOAEL, Am] = NOAEL x
[ADJ] 7 days 2
"
,,56
= 217mg/m3 x — x —
7 24
/day
= 3 9 mg/m3
described in U.S. EPA (1994b). Xylene is considered a category 3 gas because of its low water
solubility and its potential for accumulation in blood during exposure and because its most
sensitive effect is an extrarespiratory effect. The NOAEL^q was calculated using the equation
66
-------
(HV,)H
46.0
26.4
NOAEL|HEC] = NOAEL|ADJ] x
Kl
KL
= 39mg/m3 x 1
= 39mg/m3
RfC = NOAEL[HFrl -T- (UF x MF)
[HEC]
= 39mg/m3^(300x l)
= O.lmg/m3
NOAEL[HEC] = NOAEL[ADJ] x
where Hb/g = blood/gas partition coefficient for the species in question, animal (A) or human (H)
Tardif et al. (1995) reported an (Hb/g)H of 26.4 for m-xylene, and an earlier study from the
same group (Tardif et al., 1993a) reported an (Hb/g)A of 46.0 for m-xylene in the rat. It follows
that
However, when (Hb/g)A > (Hb/g)H, a value of 1 is used for the ratio (U.S. EPA, 1994b). Therefore,
To the NOAEL[HEC] of 39 mg/m3, a total UF of 300 was applied to derive the RfC as
follows:
The total UF of 300 (101/2 x 10 x 101/2x 101/2) was derived by applying a UF of 3 to
account for laboratory animal-to-human interspecies differences. A factor of 3 was applied
because default NOAEL^q dosimetric adjustments were used to calculate a HEC, reducing the
uncertainty involved with the extrapolation from the results of an animal study to a human
67
-------
exposure scenario (i.e., the toxicokinetic portion of the UF is 1; the toxicodynamic portion is 3).
(See Section 5.2.3.2 for further discussion of the toxicokinetic area of uncertainty.)
A UF of 10 was applied for intraspecies uncertainty to account for human variability and
sensitive populations. The degree of human variance in abilities to absorb or dispose of xylenes
is unknown, as is the degree of human variance in responding to xylenes neurotoxicity. Results
from developmental toxicity studies of rats exposed by inhalation during gestation indicate that
adverse developmental effects occur only at doses higher than chronic doses producing the
critical effects observed in adult male rats in the principal and supporting studies, suggesting that
the developing fetus is not at special risk from low-level exposure to xylenes. However, as with
oral exposure, the effects of inhaled xylenes in other potentially sensitive populations such as
newborns or young children or animals have not been assessed.
A UF of 3 was applied for extrapolation from subchronic to chronic duration. A factor of
10 was not used because the changes in rotarod performance did not increase with time from 1 to
3 months, and they were similar to those described in a separate study of 6 months duration
(Korsak et al., 1992).
A UF of 3 was applied for uncertainties in the data base. The inhalation data base
includes some human studies, subchronic studies in rats and dogs, neurotoxicity studies, a one-
generation reproductive toxicity study, developmental toxicity studies, and developmental
neurotoxicity studies. Although the available developmental toxicity studies are confounded by
a lack of litter incidence reporting, the data reported for fetal incidences do not indicate effects at
levels lower than the level found to induce neurologic impairment in several endpoints in male
rats. The data base is lacking a two-generation reproductive toxicity study.
5.2.3.2. PBPK Model Applications
Rat and human PBPK models for xylene inhalation (Tardif et al., 1991, 1992, 1993a, b,
1995; Haddad et al., 1999) were applied to the rat NOAEL of 50 ppm (217 mg/m3) to explore
how the use of these models may influence the derivation of the RfC.
If the NOAEL[ADJ] of 39 mg/m3 (24 hours per day) is used as the exposure concentration
for the rat PBPK model, the model predicts a steady-state pooled venous blood concentration of
0.144 mg/L. This value was then used in the human PBPK model, resulting in an estimated
NOAEL[HEC/PK] (continuous inhaled concentration in humans that would result in a steady-state
pooled venous blood concentration of 0.144 mg/L) of 41 mg/m3. This supports the NOAEL^q
of 39 mg/m3 calculated by the standard inhalation dosimetric methods.
Alternatively, the unadjusted NOAEL of 217 mg/m3 (50 ppm) and the actual exposure
protocol used in the Korsak et al. (1994) rat study (6 hours per day, 5 days per week, for 3
months) were used in the rat PBPK model to predict arterial blood concentration in rats as a
function of time up to 13 weeks. The results show a daily rise and fall of xylene concentrations,
68
-------
consistent with rapid elimination from the blood (see Appendix B, Figure 4). The use of three
different dose surrogates was explored in extrapolating to HECs with the human PBPK model:
• an overall TWA blood concentration (0.198 mg/L, averaged over 1-hour intervals
across 13 weeks),
• the maximum (MAX) blood concentration attained on any given day during exposure
(1.09 mg/L, essentially a constant over 13 weeks), and
• the mid-point (MID) concentration between the maximum (1.09 mg/L) and the
minimum (0 mg/L) concentration on any given day during exposure (0.55 mg/L).
Using these values as potential dose surrogates in the human model, the model predicted
air concentrations that would produce these steady-state concentrations in human blood with
continuous exposure. As shown in Figures 5, 6, and 7 in Appendix B, air concentrations
predicted to attain these steady-state blood concentrations in humans with continuous exposure
were 10.5 ppm (46 mg/m3) for the TWA surrogate, 27.4 ppm (106 mg/m3 ) for the MID, and 49.8
ppm (216 mg/m3) for the MAX.
The rat model predicts that blood concentrations would be essentially zero when the
critical effects on rotarod performance were measured (24 hours after cessation of exposure).
This supports the idea that the observed effects are not dependent on the concurrent presence of
xylenes in the blood and that they may be persistent neurological effects. A better dose
surrogate for use in the model would be brain concentrations, but the model has not been
developed in that regard.
The TWA dose surrogate is likely to provide a more accurate description of exposure to
the rats in the study than the MID or MAX dose surrogates, especially since the effects were
measured after m-xylene had completely cleared from the blood. Thus, using the TWA as the
dose surrogate in extrapolating from the rats in the Korsak et al. (1994) study to humans, the
model predicts a HEC of 46.5 mg/m3. This is very similar, although not identical, to the HEC
concentration (39 mg/m3) predicted using the default NOAEL[HEC] dosimetry methodology and
provides support for the RfC derived in Section 5.2.3.1.
5.3 CANCER ASSESSMENT
Human epidemiological studies have found statistically increased incidence of cancer,
but these studies are limited by the number of subjects in the cohort and the low number of
incidence reported, and the results are confounded by exposures to other solvents. The animal
carcinogenicity data base for xylenes is limited to an NTP (1986) oral bioassay in rats and mice
and an oral bioassay in rats reported by Maltoni et al. (1983, 1985). NTP (1986) found no
evidence of carcinogenicity of xylenes in rats or mice of either sex. In contrast, Maltoni et al.
(1983, 1985) reported an increase in total malignant tumors in female but not male rats following
exposure to xylenes. However, the increased incidence was calculated by combining different
69
-------
types of tumors across tissue sites. Incomplete reporting of site-specific tumor incidence data
and pathology by Maltoni et al. limit the usefulness of this bioassay in evaluating the
carcinogenicity of xylenes. Results from genotoxicity studies have been consistently negative.
Under the Draft Revised Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1999)
data are inadequate for an assessment of human carcinogenic potential of xylenes due to
inadequate evidence of carcinogenicity in humans and animals. Because the available human
and animal studies provide inadequate evidence of carcinogenicity, no estimates of dose-
response relationships can be made.
5.3.1. Oral Exposure
Not applicable.
5.3.2. Inhalation Exposure
Not applicable.
6. MAJOR CONCLUSIONS IN CHARACTERIZATION OF
HAZARD AND DOSE-RESPONSE
6.1. HUMAN HAZARD POTENTIAL
Xylenes (CASRN 1330-20-7) have the chemical formula C8H10 [structural formula
(CH3)2C6H4] and a molecular weight of 106.17. Although liquid at room temperature, xylenes
have a low vapor pressure, resulting in extensive volatilization into the air. Commercial or
mixed xylenes are comprised of three isomers: weto-xylene (m-xylene), ort/zo-xylene (o-xylene),
and/>ara-xylene (p-xylene), of which the m-isomer usually predominates (44-70% of the
mixture). The exact composition of the isomers is commonly dependent on the source. Mixed
xylenes are used in the production of the individual isomers or ethylbenzene, as a solvent, in
paints and coatings, or as a blend in gasoline.
Data on the effects of xylenes in humans following oral exposure are not available.
Results from several subchronic studies in rats and mice and one chronic study in male rats
identify decreased body weight and increased mortality as potential health hazards from repeated
oral exposure to doses generally greater than 500-800 mg/kg-day. Studies of orally exposed
animals provide some evidence for other noncancer effects at a few sites, but these appear to
occur either at doses above the lowest levels inducing body weight changes or inconsistently
across studies. Results from two animal studies indicate that developmental effects (e.g., cleft
palate) are a potential hazard from oral exposure to xylenes but that they occur at doses greater
than those inducing body weight changes.
70
-------
Reversible symptoms of neurological impairment and irritation of the eyes and throat are
well-known health hazards from acute inhalation exposure to xylenes and other aromatic
solvents in humans. In general, these acute effects are most pronounced at high exposure levels,
in excess of 1000 ppm; at lower concentrations, more subtle effects may occur. Animal studies
more clearly identify neurological effects as sensitive effects of repeated inhalation exposure to
xylenes. Several studies involving subchronic inhalation exposure of rats to m-xylene identified
100 ppm for 6 hours per day as an exposure level that produced statistically significant changes
in several neurologic endpoints, including impaired rotarod performance and decreased motor
activity, which are indicative of motor coordination; altered radial maze performance, which is
indicative of possible spatial learning impairment; and increased sensitivity to pain. Tests of
these endpoints were conducted at least 24 hours after exposure ceased, thus providing some
evidence that the changes were persistent.
A number of studies have examined the potential developmental effects of airborne
mixed xylenes or xylene isomers in animals, but adverse effects have been reported only at
exposure levels greater than those at which neurological effects have been reported.
6.2. DOSE-RESPONSE
6.2.1. Noncancer/Oral
No studies of the toxicity of xylenes in humans following subchronic or chronic oral
exposure are available. No thorough chronic studies of the noncancer toxicity of xylene in
animals are available in the literature. Noncarcinogenic endpoints reported for a lifetime oral
carcinogenicity bioassay in rats (Maltoni et al., 1983, 1985) were restricted to body weight and
hematologic variables, which were reported to have been without effect in rats exposed to 500
mg/kg-day xylenes. In contrast, in the NTP (1986) 2-year carcinogenicity bioassay in rats and
mice, male but not female rats exposed by gavage to 500 mg/kg-day for 5 days per week showed
a statistically significant increase in mortality and a statistically significant decrease (5-8%) in
mean body weight.
Subchronic studies have also reported decreased body weights: an 11% decrease
(compared with controls) in male but not female rats exposed to 1000 mg/kg-day of mixed
xylenes for 13 weeks (NTP, 1986), a 6% decrease in male but not female rats exposed to 1500
mg/kg-day of mixed xylenes for 90 days (Condie et al. 1988), a 15% decrease in male rats
exposed to 800 mg/kg-day m-xylene for 90 days (Wolfe 1988a), and a 7% decrease in mice
exposed to 2000 mg/kg-day mixed xylenes for 13 weeks (NTP, 1986). No changes in body
weight were reported in mice exposed for 2 years to mixed xylenes at doses as high as
1000 mg/kg-day (NTP, 1986).
Changes in body weight and increased mortality in rats in the chronic NTP (1986)
bioassay were selected as the critical effect for derivation of an RfD. The NOAEL of 250
mg/kg-day was duration adjusted to 179 mg/kg-day and divided by a total UF of 1000 (10 x 10 x
10) (10 for animal-to-human extrapolation, 10 for intrahuman variability, and 10 for deficiencies
71
-------
in the data base, including a lack of studies examining reproductive and neurotoxic effects) to
derive the RfD of 0.2 mg/kg-day.
6.2.2. Noncancer/Inhalation
Although adequate subchronic or chronic data in humans exposed to xylene by inhalation
are lacking, acute controlled-exposure studies have identified self-reported symptoms of
irritation (e.g., watering eyes and sore throat) or neurological impairment (e.g., mild nausea,
headache, altered reaction time, altered balance) as potential effects of xylene following
inhalation exposure in humans. However, results from subchronic animal studies identify
neurological impairment and possible developmental effects as potential health hazards from
repeated inhalation exposure. Scattered reports of body weight changes and adaptive liver
changes in animals are available, but the results do not consistently identify these effects as
potential health hazards. Several studies have examined the potential developmental toxicity of
mixed xylenes and xylene isomers, but they have identified adverse effects only at levels
considerably greater than those at which neurological effects have been reported.
The subchronic rat study by Korsak et al. (1994) was selected as the principal study for
derivation of the RfC. A NOAEL of 50 ppm and a LOAEL of 100 ppm were identified for
decreased rotarod performance (impaired motor coordination). This neurologic test was
administered 24 hours after termination of the exposure period, when xylenes would be expected
to have been eliminated from the body. Other subchronic rat studies provide support for the
finding that 100 ppm exposure produces statistically significant changes in a number of
neurological endpoints: decreased rotarod performance (Korsak et al., 1992), decreased
spontaneous motor activity (Korsak et al., 1992), and impaired radial maze performance
indicative of a learning deficit (Gralewicz et al., 1995).
Additional support for 100 ppm as an exposure level that may produce mild neurologic
deficits comes from the report (Gralewicz and Wiaderna, 2001) that rats exposed to 100 ppm
m-xylene for 4 weeks showed shortened step-down time in 1/6 trials in the passive avoidance
test 50 days postexposure. These studies collectively identify 100 ppm as the lowest reliable
subchronic animal LOAEL and 50 ppm as a NOAEL for deficits in neurologic endpoints.
The NOAEL of 50 ppm (217 mg/m3) was duration adjusted to 39 mg/m3, and a
NOAEL[HEC] of 39 mg/m3 was calculated on the basis of species differences in blood/gas
partition coefficients. The NOAEL[HEC] of 39 mg/m3 was divided by a total UF of 300 (101/2 x 10
x 101/2x 101/2; 3 for animal-to-human extrapolation using default dosimetric adjustments, 10 for
intrahuman variability, 3 for extrapolation from subchronic to chronic duration, and 3 for
deficiencies in the data base, including lack of studies in two species [available studies are
predominantly in male rats] and a two-generation reproductive toxicity study) to give the RfC of
0.1 mg/m3. Alternative approaches using available PBPK models to extrapolate rat exposure
concentrations to HECs arrived at similar points of departure for the RfC as the NOAEL^q of
39 mg/m3.
72
-------
6.2.3. Cancer/Oral and Inhalation
Data in both humans and animals are inadequate to evaluate potential associations
between xylene exposure and cancer. A number of human occupational studies have suggested
possible carcinogenic effects of chronic inhalation exposure to xylene. However, in each case
co-exposure to other chemicals was a major confounding factor, leading to an inability to
adequately assess the potential effects of chronic exposure to xylene. Animal data on the
carcinogenicity of xylene following inhalation exposure are not available. Human data on the
carcinogenic effects of oral exposure to xylenes are not available, and animal data provide
inadequate evidence of the carcinogenicity of xylenes. Under the Draft Revised Guidelines for
Carcinogen Risk Assessment (U.S. EPA, 1999) data are inadequate for an assessment of the
carcinogenic potential of xylenes.
73
-------
7. REFERENCES
Abu-Al-Ragheb, S; Salhab, AS; Amr, SS. (1986) Suicide by xylene ingestion: A case report and
review of literature. Am J Forensic Med Path. 7:327-329.
ACGIH (American Conference of Governmental Industrial Hygienists). (1991) Documentation
of the threshold limit values and biological exposure indices: Xylene (o-, m-, and p-isomers).
Sixth ed. Cincinnati, OH; pp. 1732-1740.
Anderson, BE; Zeiger, E; Shelby, MD; et al. (1990) Chromosome aberration and sister
chromatid exchange test results with 42 chemicals. Environ Mol Mutagen. 16: Suppl.
18:55-137.
Andersson, K; Fuxe, K; Nilsen, OG; et al. (1981) Production of discrete changes in dopamine
and noradrenaline levels and turnover in various parts of the rat brain following exposure to
xylene, ortho-, meta-, and/>ara-xylene, and ethylbenzene. Toxicol Appl Pharmacol.
60:535-548.
Arms, AD; Travis, CC. (1988) Reference parameters in pharmacokinetic modeling. (NTIS PB
88-196019). Washington, DC.
Arp, EW, Jr.; Wolf, PH; Checkoway, H. (1983) Lymphocytic leukemia and exposures to
benzene and other solvents in the rubber industry. J Occup Med. 25:598-602.
Arthur, LJ; Curnock, DA. (1982) Xylene-induced epilepsy following innocent glue sniffing. Br
Med 1284:1787.
Astrand, I; Engstrom, J; Ovrum, P. (1978) Exposure to xylene and ethylbenzene. I. Uptake,
distribution, and elimination in man. Scand J Work Environ Health. 4:185-194.
ATSDR (Agency for Toxic Substances and Disease Registry). (1995) Toxicological profile for
xylene (update). Chamblee, GA.; p. 270.
Bergman, R. (1983) Application and results of whole-body autoradiography in distribution
studies of organic solvents. CRC Crit Rev Toxicol. 12:59-118.
Bio/dynamics Inc. (1983) Parental and fetal reproduction toxicity study in rats with mixed
xylene. EPA/OTS public files. Bio/dynamics Inc., East Millstone, NJ; Document # FYI-AX-
0983-0209.
Borriston Laboratories, Inc. (1983) Four-week oral nephrotoxicity screening study in male F-344
rats phases I and II pathology report. FYI submission AX-1283-0280. Submitted by the
American Petroleum Institute to the U.S. Environmental Protection Agency, Washington, DC.
74
-------
Bos, RP; Brouns, RME; van Doom, R; et al. (1981) Non-mutagenicity of toluene, o-, m-, and p-
xylene, o-methylbenzylalcohol and o-methylbenzylsulfate in the Ames assay. Mutat Res.
88:273-279.
Bowers, DE, Jr.; Cannon, S; Jones, DH. (1982) Ultrastructural changes in livers of young and
aging rats exposed to methylated benzenes. Am J Vet Res. 43:679-683.
Bray, HG; Humphris, BG; Thorpe, WV. (1949) Metabolism of derivatives of toluene: 3. o-, m-,
and p-Xylenes. Biochem J. 45:241-244.
Budavari, S; O'Neil, M J; Smith, A; et al. (eds.). (1996) The Merck Index. Merck & Co., Inc.,
Whitehouse Station, NJ; pp. 1722-1723.
Carlsson, A. (1981) Distribution and elimination of 14C-xylene in rat. Scand J Work Environ
Health. 7:51-55.
Carpenter, CP; Kinkead, ER; Geary, DL Jr; et al. (1975) Petroleum hydrocarbon toxicity studies.
V. Animal and human response to vapors of mixed xylene. Toxicol Appl Pharmacol.
33:543-58.
Condie, LW; Hill, JR; Borzelleca, JF. (1988) Oral toxicology studies with xylene isomers and
mixed xylene. Drug Chem Toxicol. 11:329-354.
David, A; Flek, J; Frantik, E; et al. (1979) Influence of phenobarbital on xylene metabolism in
man and rats. Int Arch Occup Environ Health. 44:117-125.
Desi, I; Kovacs, F; Zahumenszky, Z; et al. (1967) Maze learning in rats exposed to xylene
intoxication. Psychopharmacologia. 11:224-230.
Donner, M; Maki-Paakkanen, J; Norppa, H; et al. (1980) Genetic toxicology of xylenes. Mutat
Res. 74:171-172.
Dudek, B; Gralewicz, K; Jakubowski, M; et al. (1990) Neurobehavioral effects of experimental
exposure to toluene, xylene and their mixture. Polish J Occup Med. 3:109-116.
Engstrom, J; Bjurstrom, R. (1978) Exposure to xylene and ethylbenzene. II. Concentration in
subcutaneous adipose tissue. Scand J Work Environ Health. 4:195-203.
Fang, Z ; Sonner, J; Laster, MJ; et al. (1996) Anesthetic and convulsant properties of aromatic
compounds and cycloalkanes: Implications for mechanisms of narcosis. Anesth Analg.
83:1097-1104.
Fishbein, L. (1988) Xylenes: uses, occurrence and exposure. IARC Sci Publ. 85:109-120.
75
-------
Florin, I; Rutberg, L; Curvall, M; et al. (1980) Screening of tobacco smoke constituents for
mutagenicity using the Ames' test. Toxicology. 15:219-232.
Gagnaire, F; Marignac B; Langlais C; et al. (2001) Ototoxicity in rats exposed to ortho-, meta-,
and para-xylene vapours for 13 weeks. Pharm Toxicol. 89:6-14.
Gamberale, F; Annwall, G; Hultengren, M. (1978) Exposure to xylene and ethylbenzene. III.
Effects on central nervous functions. Scand J Work Environ Health. 4:204-211.
Gargas, ML; Burgess RJ; Voisard DE; et al. (1989) Partition coefficient of low molecular weight
volatile chemicals in various tissues and liquids. Toxicol. Appl. Pharmacol. 98: 87-99.
Gerin, M; Siemiatycki, J; Desy, M; et al. (1998) Associations between several cites of cancer
and occupational exposure to benzene, toluene, xylene, and styrene: results of a case-control
study in Montreal. Am J Ind Med. 34:144-156.
Gerner-Smidt, P; Friedrich, U. (1978) The mutagenic effect of benzene, toluene, and xylene
studied by the SCE technique. Mutat Res. 58:313-316.
Goldie, I. (1960) Can xylene (xylol) provoke convulsive seizures? Ind Med Surg. 29:33-35.
Gralewicz, S; Wiaderna, D. (2001) Behavioral effects following subacute inhalation exposure to
m-xylene or trimethylbenzene in the rat: a comparative study. Neurotoxicology. 22:79-89.
Gralewicz, S; Wiaderna, D; Tomas, T. (1995) Development of spontaneous, age-related
nonconvulsive seizure electrocortical activity and radial-maze learning after exposure to m-
xylene in rats. Int J Occup Med Environ Health. 8:347-360.
Haddad, S; Tardif, R; Viau, C; et al. (1999) A modeling approach to account for toxicokinetic
interactions in the calculation of biological hazard index for chemical mixtures. Toxicol Lett.
108:303-308.
Haglund, U; Lundberg, I; Zech, L. (1980) Chromosome aberrations and sister chromatid
exchanges in Swedish paint industry workers. Scand J Work Environ Health. 6:291-298.
Hass, U; Jakobsen, BM. (1993) Prenatal toxicity of xylene inhalation in the rat: a teratogenicity
and postnatal study. Pharmacol Toxicol. 73:20-23.
Hass, U; Lund, SP; Simonsen, L; et al. (1995) Effects of prenatal exposure to xylene on postnatal
development and behavior in rats. Neurotoxicol Teratol. 17:341-349.
Hass, U; Lund, SP; Simonsen, L. (1997) Long-lasting neurobehavioral effects of prenatal
exposure to xylene in rats. Neurotoxicology. 18:547-551.
76
-------
Hipolito, RN. (1980) Xylene poisoning in laboratory workers: case reports and discussion. Lab
Med. 11:593-595.
Honma, T; Sudo, A; Miyagawa, M; et al. (1983) Significant changes in the amounts of
neurotransmitter and related substances in rat brain induced by subacute exposure to low levels
of toluene and xylene. Indust Health. 21:143-151.
Hudak, A; Ungvary, G. (1978) Embryotoxic effects of benzene and its methyl derivatives:
toluene, xylene. Toxicology. 11:55-63.
IARC (International Agency for Research on Cancer). (1989) IARC Monographs on the
Evaluation of Carcinogenic Risks to Humans. Vol. 47. Some Organic Solvents, Resin
Monomers and Related Compounds, Pigments and Occupational Exposures in Paint
Manufacture and Painting. Lyon, France; pp. 125-156.
Jenkins, LJ, Jr.; Jones, RA; Siegel, J. (1970) Long-term inhalation screening studies of benzene,
toluene, o-xylene, and cumene on experimental animals. Toxicol Appl Pharmacol. 16:818-823.
Kaneko, T; Endoh, K; Sato, A. (1991) Biological monitoring of exposure to organic solvent
vapors. I. A physiological simulation model of m-xylene pharmacokinetics in man. Yamanashi
Med J. 6:127-135.
Kaneko, T; Wang, P; Sato, A. (1993) Enzyme induction by ethanol consumption affects the
pharmacokinetics of inhaled m-xylene only at high levels of exposure. Arch Toxicol. 67:
473-477.
Kaneko, T; Wang, P; Tsukada, H; et al. (1995) m-Xylene toxicokinetics in phenobarbital-treated
rats: comparison among inhalation exposure, oral administration, and intraperitoneal
administration. Toxicol Appl Pharmacol. 131:13-20.
Kaneko, T; Horicuhi, J; Sato, A (2000) Development of a physiologically based
pharmacokinetic model of organic solvent in rats. Pharmacol Res. 42(5):
http://www.idealibrary.com.
Klaucke, DN; Johansen, M; Vogt, R (1982) An outbreak of xylene intoxication in a hospital.
AmJIndMed. 3:173-178.
Korsak, Z; Sokal, JA; Wasiela, T; et al. (1990) Toxic effects of acute exposure to particular
xylene isomers in animals. Pol J Occup Med. 3:221-226.
Korsak, Z; Sokal, JA; Gorny, R. (1992) Toxic effects of combined exposure to toluene and m-
xylene in animals. III. Subchronic inhalation study. Polish J Occup Med Environ Health.
5:27-33.
77
-------
Korsak, Z; Wisniewska-Knypl, J; Swiercz, R. (1994) Toxic effects of subchronic combined
exposure to n-butyl alcohol and m-xylene in rats. Int J Occup Med Environ Health. 7:155-166.
Kiikner, A; Canpolat, L; Ozan, E; et al. (1997/98) The effect of xylene inhalation on the rat liver.
Acta Physiologica Hungarica. 85:231-241.
Kumarathasan, P; Otson, R; Chu, I. (1997) Measurement of the distribution of m-xylene in rat
tissues by head space gas chromatography. Arh Hig Rada Toksikol. 48:378-382.
Kumarathasan, P; Otson, R; Chu, I. (1998) Application of an automated HS-GC method in
partition coefficient determination for xylenes and ethylbenzene in rat tissues. Chemosphere.
37:159-178.
Laine, A; Savolainen, K; Riihimaki, V; et al. (1993) Acute effects of m-xylene inhalation on
body sway, reaction times, and sleep in man. Int Arch Occup Environ Health. 65:179-188.
Litton Bionetics. (1978a) Teratology study in rats - xylene. Final report EPA/OTS Public Files.
Litton Bionetics, Kensington, MD. Document 878210350.
Litton Bionetics. (1978b) Mutagenicity evaluation of xylene. EPA/OTS Public Files. Litton
Bionetics, Kensington, MD. Document 878210347.
Maltoni, C; Conti, B; Cotti, G. (1983) Benzene: a multipotential carcinogen. Results of long-
term bioassays performed at the Bologna Institute of Oncology. Am J Ind Med. 4:589-630.
Maltoni, C; Conti, B; Cotti, G; et al. (1985) Experimental studies on benzene carcinogenicity at
the Bologna Institute of Oncology: current results and ongoing research. Am J Ind Med.
7:415-446.
Marks, TA; Ledoux, TA; Moore, JA. (1982) Teratogenicity of a commercial xylene mixture in
the mouse. J Toxicol Environ Health. 9:97-105.
McCarroll, NE; Piper, CE; Keech, BH. (1981) AnE. coli microsuspension assay for the
detection of DNA damage induced by direct-acting and promutagens. Environ Mutagen.
3:429-444.
Mirkova, E; Zaikov, C; Antov, G; et al. (1983) Prenatal toxicity of xylene. J Hyg Epi Microbiol.
27:337-343.
Mohtashamipur, E; Norpoth, K; Woelke, U; et al. (1985) Effects of ethylbenzene, toluene, and
xylene on the induction of micronuclei in bone marrow polychromatic erythrocytes of mice.
Arch Toxicol. 58:106-109.
78
-------
Molnar, J; Paksy, KA.; Naray, M. (1986) Changes in the rat's motor behavior during 4-hr
inhalation exposure to prenarcotic concentrations of benzene and its derivatives. Acta Physiol
Hung. 67:349-354.
Morley, R; Eccleston, DW; Douglas, CP; et al. (1970) Xylene poisoning: a report on one fatal
case and two cases of recovery after prolonged unconsciousness. Brit Med J. 3:442-443.
Morvai, V; Hudak, A; Ungvary, G. (1976) ECG changes in benzene, toluene, and xylene
poisoned rats. Acta Med Acad Sci Hung. 33:275-286.
Moser, VC; Coggeshall, EM; Balster, RL. (1985) Effects of xylene isomers on operant
responding and motor performance in mice. Toxicol Appl Pharmacol. 80:293-298.
NRC (National Research Council). (1983) Risk Assessment in the Federal Government:
Managing the Process. Committee on the Institutional Means for Assessment of Risks to Public
Health, Commission on Life Sciences. Washington, DC.
NRC Emergency and Continuous Exposure Limits for Selected Airborne Contaminants. (1984)
Committee on Toxicology, Board on Toxicology and Environmental Health Hazards,
Commission on Life Sciences. Washington, DC.
Nawrot, PS; Staples, RE. (1980) Embryotoxicity and teratogenicity of isomers of xylene in the
mouse. Soc Toxicol Abst. PAP 19th: A22, 65.
NTP (National Toxicology Program). (1986) Technical Report on the Toxicology and
Carcinogenesis of Xylenes (mixed) (60% m-xylene, 13.6% p-xylene, 17.0% ethylbenzene, and
9.1% o-xylene) in F344/N Rats and B6C3F1 mice (gavage studies). Research Triangle Park,
NC. NTP TR 327, NIH Publ. No. 86-2583.
Nylen, P; Hagman, M. (1994) Function of the auditory and visual systems, and of peripheral
nerve, in rats after long-term combined exposure to n-hexane and methylated benzene
derivatives. II. Xylene. Pharmacol Toxicol. 74:124-129.
Nylen, P; Ebendal T; Eriksdotter-Nilsson M; et al. (1989) Testicular atrophy and loss of nerve
growth factor-immunoreactive germ cell line in rats exposed to n-hexane and a protective effect
of simultaneous exposure to toluene or xylene. Arch Toxicol. 63:296-307.
Ogata, M; Tomokuni, K; Takatsuka, Y. (1970) Urinary excretion of hippuric acid and m- or p-
methylhippuric acid in the urine of persons exposed to vapours of toluene and m- or p-xylene as
a test of exposure. Brit J Indust Med. 27:43-50.
Ogata, M; Yamazaki, Y; Sugihara, R; et al. (1980) Quantitation of urinary o-xylene metabolites
of rats and human beings by high-performance liquid chromatography. Int Arch Occup Environ
Health. 46:127-139.
79
-------
Olson, BA; Gamberale, F; Inegren, A. (1985) Coexposure to toluene and p-xylene in man:
central nervous functions. Br J Ind Med. 42:117-122.
Pap, M; Varga, C. (1987) Sister-chromatid exchanges in peripheral lymphocytes of workers
occupationally exposed to xylene. MutatRes. 187:223-225.
Pryor, GT; Rebert, CS; Howd, RA. (1987) Hearing loss in rats caused by inhalation of mixed
xylene and styrene. J Appl Toxicol. 7:55-61.
Purcell, KJ; Cason GH; Gargas ML; et al. (1990) In vivo metabolic interactions between benzene
and toluene. Toxicol Lett. 52:141-152.
Recchia, G; Perbelline, L; Prati, GF; et al. (1985) Coma da probabile ingestione accidentale di
xilene: trattamento mediante emoperfusione con carbone attivato. Med Lav. 76:67-73.
Richer, CL; Chakrabarti, S; Senecal-Quevillon, M; et al. (1993) Cytogenic effects of low-level
exposure to toluene, xylene, and their mixture on human blood lymphocytes. Arch Occup
Environ Health. 64:581-585.
Riihimaki, V. (1979) Conjugation and urinary excretion of toluene and m-xylene metabolites in a
man. Scand J Work Environ Health. 5:135-142.
Riihimaki, V; Savolainen, K. (1980) Human exposure to m-xylene. Kinetics and acute effects
on the central nervous system. Ann Occup Hyg. 23:411-422.
Rosen, MB; Crofton, KM; Chernoff, N. (1986) Postnatal evaluation of prenatal exposure to p-
xylene in the rat. Toxicol Let. 34:223-229.
Rosengren, LE; Kjellstrand, P ; Aurell, A; et al. (1986) Irreversible effects of xylene on the brain
after long term exposure: a quantitative study of DNA and the glial cell marker proteins S-100
and GFA. Neurotoxicology. 7:121-136.
Sato, A; Nakajima, T. (1979) Partition coefficients of some aromatic hydrocarbons and ketones
in water, blood and oil. Br J Ind Med. 36:231-234.
Savolainen, K; Linnavuo, M. (1979) Effects of m-xylene on human equilibrium measured with a
quantitative method. Acta Pharmacol Toxicol. 44:315-318.
Savolainen, K; Pfaffli, P. (1980) Dose-dependent neurochemical changes during short-term
inhalation exposure to m-xylene. Arch. Toxicol. 45:117-122.
Savolainen, H; Pfaffli, P; Helojoki, M; et al. (1979) Neurochemical and behavioral effects of
long-term intermittent inhalation of xylene vapour and simultaneous ethanol intake. Acta
Pharmacol et Toxicol. 44:200-207.
80
-------
Savolainen, K; Kekoni, J; Riihimaki, V; et al. (1984) Immediate effects of m-xylene on the
human central nervous system. Arch Toxicol Suppl. 7:412-417.
Sedivec, V; Flek, J. (1976a) The absorption, metabolism, and excretion of xylenes in man. Int
Arch Occup Environ Health. 37:205-217.
Sedivec, V; Flek, J. (1976b) Exposure test for xylenes. Int Arch Occup Environ Health.
37:219-232.
Senczuk, W; Orlowski, J. (1978) Absorption of m-xylene vapours through the respiratory tract
and excretion. Br J Ind Med. 35:50-55.
Seppalainen, AM; Salmi, T; Savolainen, K; et al. (1983) Visual evoked potentials in short-term
exposure of human subjects to m-xylene andl,l,l-trichloroethane. Appl Behav Pharmacol
Toxicol. 1983:349-352.
Seppalainen, AM; Laine, A; Salmi, T; et al. (1991) Electroencephalographic findings during
experimental human exposure to m-xylene. Arch Environ Health. 46:16-24.
Sevcik, P; Hep, A; Peslova, M. (1992) Intravenous xylene poisoning. Intensive Care Med.
18:377-378.
Spirtas, R; Stewart, PA; Lee, JS; et al. (1991) Retrospective cohort mortality study of workers at
an aircraft maintenance facility. I. Epidemiological results. BrJ Ind Med. 48:515-530.
Tahti, H. (1992) The neurotoxicity of organic solvents, studied with in vitro models. Altern Lab
Anim. 20:290-296.
Tardif, R; Lapare, S; Plaa, GL; et al. (1991) Effect of simultaneous exposure to toluene and
xylene on their respective biological indices in humans. Int Arch Occup Environ Health.
63:279-284.
Tardif; Plaa, GL; Brodeur, J. (1992) Influence of various mixtures of inhaled toluene and xylene
on the biological monitoring of exposure to these solvents in rats. Can J Physiol Pharmacol.
70:385-393.
Tardif, R; Lapare, S; Krishnan, K; et al. (1993a) Physiologically based modeling of the
toxicokinetic interaction between toluene and m-xylene in the rat. Toxicol Appl Pharmacol.
120:266-273.
Tardif, R; Lapare, S; Krishnan, K; et al. (1993b) A descriptive and mechanistic study of the
interaction between toluene and xylene in humans. Int Arch Occup Environ Health. 65:
S135-S137.
81
-------
Tardif, R; Lapare, S; Charest-Tardif, G; et al. (1995) Physiologically-based pharmacokinetic
modeling of a mixture of toluene and xylene in humans. Risk Anal. 15:335-342.
Tardif, R; Charest-Tardif, G; Brodeur, J; et al. (1997) Physiologically-based pharmacokinetic
modeling of a ternary mixture of alkyl benzenes in rats and humans. Toxicol Appl Pharmacol.
144:120-134.
Taskinen, H; Antilla, A; Lindbohm, ML; et al. (1989) Spontaneous abortions and congenital
malformations among the wives of men occupationally exposed to organic solvents. Scand. J.
Work Environ. Health. 15:345-352.
Taskinen, H; Lindbohm, ML; Hemminki, K. (1986) Spontaneous abortions among women
working in the pharmaceutical industry. Br J Ind Med. 43:199-205.
Taskinen, H; Kyyronen, P; Hoikkala, M; et al. (1994) Laboratory work and pregnancy outcome.
JOccup Med. 36:311-319.
Tatrai, E; Ungvary, G. (1980) Changes induced by o-xylene inhalation in the rat liver. Acta Med
Acad Sci Hung. 37:211-216.
Tatrai, E; Ungvary, G; Cseh, IR; et al. (1981) The effect of long-term inhalation of o-xylene on
the liver. Ind Environ Xenobiotics, Proc Int Conf; pp. 293-300.
Turkall, RM; Skowronski, GA; Kadry, ARM; et al. (1992) Sex differences in the bioavailability
of soil-adsorbed m-xylene in orally exposed rats. Toxicol Let. 63:57-67.
Uchida, Y; Nakatsuka, H; Ukai, H; et al. (1993) Symptoms and signs in workers exposed
predominantly to xylene. Int Arch Occup Environ Health. 64:597-605.
Ungvary, G. (1990) The effect of xylene exposure on the liver. Acta Morphol Hung.
38:245-258.
Ungvary, G; Tatrai, E. (1985) On the embryotoxic effects of benzene and its alkyl derivatives in
mice, rats and rabbits. Arch Toxicol., Suppl. 8:425-430.
Ungvary, G; Tatrai, E; Hudak, A; et al. (1980) Studies on the embryotoxic effects of ortho-,
meta-and para-xylene. Toxicology. 18:61-74.
Ungvary, G; Varga, B; Horvath, E; et al. (1981) Study on the role of maternal sex steroid
production and metabolism in the embryotoxicity of para-xylene. Toxicology. 19:263-268.
U.S. EPA (U.S. Environmental Protection Agency). (1986a) Guidelines for the Health Risk
Assessment of Chemical Mixtures. Federal Register. 51 (185):34014-34025.
82
-------
U.S. EPA. (1986b) Guidelines for Mutagenicity Risk Assessment. Federal Register. 51
(185):34006-34012.
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.
U.S. EPA. (1991) Guidelines for Developmental Toxicity Risk Assessment. Federal Register
56:63798-63826.
U.S. EPA. (1994a) Interim Policy for Particle Size and Limit Concentration Issues in Inhalation
Toxicity: Notice of Availability. Federal Register 59(206):53799.
U.S. EPA. (1994b) Methods for Derivation of Inhalation Reference Concentrations and
Application of Inhalation Dosimetry. EPA/600/8-90/066F.
U.S. EPA. (1994c) Peer Review and Peer Involvement at the U.S. Environmental Protection
Agency. Signed by Administrator Carol Browner, June 7, 1994.
U.S. EPA. (1995) Use of the Benchmark Dose Approach in Health Risk Assessment.
EPA/630/R-94/007.
U.S. EPA. (1996) Guidelines for Reproductive Toxicity Risk Assessment. Federal Register 61
(212):56274-56322.
U.S. EPA. (1998a) Guidelines for Neurotoxicity Risk Assessment. Federal Register 63
(93):26926-26954.
U.S. EPA. (1998b) Science policy council handbook: peer review. Office of Science Policy,
Office of Research and Development, Washington, DC. EPA/100/B-98/001.
U.S. EPA. (1998c) Aggregated Inventory Update. Electronic Database. Office of Pollution
Prevention and Toxics. Washington, D.C.
U.S. EPA. (1999) Guidelines for carcinogen risk assessment. Review Draft. NCEA-F-0644,
July. Risk Assessment Forum, Washington, DC. Available at:
http://www.epa.gov/ncea/raf/cancer.htm.
U.S. EPA. (2000a) Science policy council handbook: peer review. Second edition. Office of
Science Policy, Office of Research and Development, Washington, DC. EPA 100-B-OO-OOl.
U.S. EPA. (2000b) Science policy council handbook: risk characterization. Office of Science
Policy, Office of Research and Development, Washington, DC. EPA 100-B-00-002.
83
-------
U.S. EPA. (2000c) Benchmark Dose Technical Support Document. External Review Draft,
EPA/630/R-00/001. Office of Research and Development, Risk Assessment Forum,
Washington, DC.
U.S. EPA. (2000d) Supplementary Guidance for Conducting Health Risk Assessment of
Chemical Mixtures. Office of Research and Development, Risk Assessment Forum,
Washington, DC. EPA/630/R-00/002.
U.S EPA (2001) Agency requests National Academy of Sciences input on consideration of
certain human toxicity studies; announces interim policy [press release with attached
memorandum]. Washington, DC: December 14.
U.S. EPA. (2002) IRIS Summary of Ethylbenzene (CASRN 100-41-4) National Center for
Environmental Assessment, Washington, DC. Available at: http://www.epa.gov/iris.
Washington, WJ; Murthy, RC; Doye, A; et al. (1983) Induction of morphologically abnormal
sperm in rats exposed to o-xylene. Arch Andrology. 11:233-237.
Wilcosky, TC; Checkoway, H; Marshall, EG; et al. (1984) Cancer mortality and solvent
exposures in the rubber industry. Am Ind Hyg Assoc J. 45:809-811.
Wolfe, GW. (1988a) Subchronic toxicity study in rats with m-xylene. Report by Hazleton
Laboratories America, Inc., sponsored by Dynamac Corporation, Rockville, MD. Project No.
2399-108.
Wolfe, GW. (1988b) Subchronic toxicity study in rats with p-xylene. Report by Hazleton
Laboratories America, Inc. sponsored by Dynamac Corporation, Rockville, MD. Project No.
2399-110.
84
-------
APPENDIX A. EXTERNAL PEER RE VIEW—SUMMARY OF
COMMENTS AND DISPOSITION
The support document and IRIS summary for xylenes have undergone both internal peer
review by scientists within EPA and a more formal external review by scientists in accordance
with EPA guidance on peer review (U.S. EPA, 1994c). Comments made by the internal
reviewers were addressed prior to submitting the documents for external review and are not part
of this appendix. 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 significant comments made by the external
reviewers and EPA's response to these comments follows. EPA also received scientific
comments from the public. These comments and EPA's response are included in a separate
section.
Scientific Comments from the External Peer Review
(1) General Comments
A. Editorial/grammatical suggestions
Comment: All three reviewers made editorial comments on the documents. A number of
comments related to a lack of sufficient detail in study descriptions.
Response: Where possible, the suggested editorial changes have been implemented. Study
descriptions have been augmented throughout the document.
B. Rationale for health assessments
Comment: All three reviewers commented that the rationale for selection of principal studies
and critical effects as well as the weight of evidence sections were inadequate.
Response: The appropriate sections have been revised in accordance with the reviewers'
comments to clarify the selection of principal studies and critical effects and to more clearly
describe the weight of evidence for noncancer health effects from oral and inhalation exposure.
C. Lack of concordance between oral and inhalation reference values
Comment: One reviewer commented extensively on the lack of a direct concordance between
the RfD and the RfC, noting that the RfC may be too low. In order to check on the
appropriateness of the RfC, the reviewer calculated a daily dose, based on the RfC of 0.1 mg/m3,
of 0.017 mg/kg-day for a 70 kg human, assuming a 60% absorption, and compared that with the
RfD proposed in the external peer review draft of 0.7 mg/kg-day. The reviewer noted that ratio
of the RfD to the estimated average daily absorbed dose delivered by the RfC (40:1) indicated
85
-------
that xylenes may be 40 times more toxic by inhalation than by oral administration and expressed
the opinion that this is likely to be "grossly incorrect." The reviewer noted that the apparent
discrepancy could be due to the duration adjustment that was made in calculating the
NOAEL[HEC] of 39 mg/m3 from the rat experimental concentration of 217 mg/m3, 6 hours per day,
5 days per week, pointing out that such an adjustment would not be justified if the
neurobehavioral effects noted in the principal rat study were of an acute rather than a persistent
nature. The reviewer suggested that PBPK analysis may help to rectify this apparent
discrepancy but noted that this difference should be either justified or corrected.
Response: Several changes were made in the document in response to the reviewer's comments.
Text describing and evaluating the rat and human PBPK models for xylenes is included in
Section 3.5 and Appendix B. The new text notes that existing models are for inhalation and do
not have an oral portal of entry. As such, they are not useful for route-to-route extrapolation.
Application of the existing models to the derivation of the RfC was explored, and the results of
this analysis are presented in Appendix B and in Sections 5 and 6. The analysis indicated that
the NOAEL[HEC] derived with the default inhalation dosimetry methodology for category 3 gases
was similar to the HEC predicted by the PBPK models, providing support for the RfC.
In addition, the new text specifies that the neurobehavioral tests in the principal and
supporting studies for the RfC appeared to have been administered 24 hours after exposure
ceased, providing support that the observed effects may be persistent rather than acute effects
that are dependent on the presence of xylenes in the blood. The revised document notes that
blood concentrations were used as dose surrogates in applying the models and that brain
concentrations would be a better dose surrogate. However, models have not been developed to
predict brain concentrations of xylenes. Other changes were made (principally in Section 3.3) to
more completely discuss what is known concerning the first-pass metabolic effect associated
with oral exposure to xylenes.
In relation to the potential discordance between the RfD and RfC identified by the
reviewer, it may be more appropriate to compare the two points of departure (POD) using
physiological data for the rat. Using the reviewers' simple calculation, this would involve
multiplying the duration adjusted inhalation POD (39 mg/m3) times the rat minute volume (0.33
m3/day) times the presumed absorption of 0.6 (from human data), and then dividing by the rat
body weight of 0.180 kg. This yields an estimate of systemic intake of 42.9 mg/kg-day, which is
only 1/4 the oral POD (200 mg/kg-day) chosen for the RfD. Whether fortuitous or having
scientific basis, the implication is that of concordance of the oral and inhalation data, particularly
given the knowledge that first pass metabolism by the oral route occurs.
Furthermore, the RfD on which the reviewer based his calculations has been replaced
with a new RfD, which was derived in response to external peer review comments. The
principal study, critical effect, and point of departure for the derivation of the RfD have changed
as noted. The updated value of the RfD is 0.2 mg/kg-day versus the value of 0.7 mg/kg-day
proposed in the external peer review draft. The ratio of the current RfD and the reviewer's
86
-------
estimated daily absorbed dose associated with exposure to the RfC is now 11 rather than the
previous value of 40.
D. Reproductive and developmental data
Comment: One reviewer recommended additional consideration and descriptions of
developmental/reproductive toxicity studies. This reviewer indicated that the weight of evidence
shows consistent developmental effects across studies.
Response: The reproductive and developmental data have been re-evaluated, and expanded
study descriptions are included in Section 4.3. The text in Section 4.5.2. includes a more
complete rationale and explanation of the weight of evidence indicating that neurobehavioral
effects are a more sensitive health effect than are effects on the developing organism following
inhalation exposure to xylene.
E. Other comments
Comment: One reviewer recommended the inclusion of tables summarizing the
NOAELs/LOAELs for the oral and inhalation studies.
Response: Two tables, one summarizing the available neurobehavioral studies in adult male rats
and one summarizing the pertinent developmental studies (and their NOAELs and LOAELs),
have been added to Section 4.5.2, outlining the weight of evidence for the inhalation data. The
oral data base was not as extensive as the inhalation data base and was therefore discussed in
paragraph form.
Comment: One reviewer noted that the contribution of ethylbenzene to the toxicity of mixed
xylenes should be discussed.
Response: A discussion of the possible role of ethylbenzene in the oral effects of mixed xylenes
has been added to Section 5.1.3, in the discussion of uncertainties in the data base. Additional
discussion of the potential contribution of ethylbenzene to the oral effects of xylenes is included
in Sections 4.5.1. and 4.4.3. An equivalent section was not added to Section 5.2.3 because the
inhalation RfC is not based on a study that used mixed xylenes. The critical effects for the
current inhalation RfC for ethylbenzene are mild developmental effects and skeletal variants
observed in rats exposed to 1000 ppm ethylbenzene (U.S. EPA, 2002). Although developmental
effects are an endpoint of concern for inhalation of xylenes, the available data suggest that
neurological effects are a more sensitive endpoint.
Comment: Two reviewers suggested that a summary paragraph be added to the toxicokinetic
section of the document.
Response: A brief summary is included at the beginning of Section 3.
87
-------
(2) Specific Charge Questions
Question 1. Are you aware of any other data/studies that are relevant (i.e., useful for hazard
identification or dose-response assessment) for the assessment of the adverse health effects, both
cancer and noncancer, of this chemical?
Comments: External reviewers identified a number of studies that were not cited,
including a male rat reproductive toxicity study (Nylen et al., 1989), several toxicokinetic studies
in animals (Turkall et al., 1992; Kaneko et al., 1993, 1995), several case-control studies of
spontaneous abortions in solvent-exposed female workers (Taskinen et al., 1989), and several
PBPK modeling studies (Tardif et al., 1991, 1992, 1993a, b; Haddad et al., 1999).
Response to Comments: In response to the reviewers' comments, discussions of these
studies were incorporated into the document. A review and evaluation of the male rat
reproductive toxicity study by Nylen et al. (1989) are included in Section 4.3.1.2. Descriptions
and evaluations of available rat and human PBPK models for inhaled xylene as well as
discussion of their application to the derivation of the RfC are now included in Sections 3.5,
5.2.2., 5.2.3, and Appendix B. Toxicokinetic data published by Turkall et al. (1992) and
Kaneko et al. (1993, 1995) are discussed in detail in Sections 3.1. and 3.5. Case-control studies
by Taskinen et al. (1989, 1994) are evaluated in Section 4.1.2.1.
Question 2. For the RfD and RfC, has the most appropriate critical effect been chosen? Points
relevant to this determination include whether or not the choice follows from the dose-response
assessment, whether the effect is considered adverse, and whether the effect and the species in
which it is observed is a valid model for humans.
Comments: For the RfD, all three reviewers questioned the choice of hyperactivity seen
at 2000 mg/kg-day in the 13-week NTP mouse study as the critical effect, particularly given (a)
the availability of a chronic study, (b) the presence of mortality at 500 mg/kg-day in the chronic
rat bioassay, and (c) the apparent acute nature of the effect (persisted only a short time after
exposure each day). All three reviewers acknowledged the difficulty in selecting an appropriate
endpoint for deriving an RfD. Alternate endpoints that may be appropriate for deriving an RfD
were suggested: the developmental effects (cleft palate and reduced fetal body weight) reported
by Marks et al. (1982) or the reduced body weight found in the NTP (1986) chronic and
subchronic studies. All reviewers discussed the concern that decreased body weight should be
used as a critical effect unless a stronger case could be made to discount this effect.
One reviewer indicated that increased mortality should at least be considered in the dose-
response evaluation but that this effect should not be the critical effect. Another reviewer
suggested using the LOAEL for hypersensitivity in mice from the chronic study (NTP, 1986)
over the LOAEL for subchronic clinical effects (NTP, 1986) in deriving the RfD. This reviewer
stated that using the subchronic study over the chronic study had too many shortcomings. In
addition to being of shorter duration, the subchronic study only used 10 animals/sex/group,
whereas the chronic study used 50 animal s/sex/group. In addition, it was pointed out that
88
-------
comprehensive histopathology was conducted on all animals in the chronic study but only in the
vehicle control and high-dose animals in the subchronic study.
For the RfC, all three reviewers agreed that the selection of the critical effect was
appropriate. However, it was pointed out that another study (Gralewicz et al., 1995) identified
effects at the same dose (i.e., 100 ppm) but was not included as a co-principal study. One
reviewer thought that there was also some evidence of human effects at lower doses (14 ppm), as
identified in the Uchida et al. (1993) study, that should be addressed. Another reviewer thought
that hematological effects were discounted without sufficient explanation.
Response to Comments: As per the reviewers' comments, the critical effect for the RfD
has been re-evaluated and changed to reduced body weight and increased mortality, based on the
evidence from subchronic and chronic rat studies following oral xylenes exposure. A new RfD
has been derived using decreased body weight and mortality as the critical effect. The
developmental NOAEL and LOAEL identified by Marks et al. (1982) were not used because the
available LOAEL values for decreased body weight were lower, indicating a more sensitive
effect. Sections 4.5.1 and 5.1 have been extensively modified, reflecting the new principal study
and critical effect.
The weight of evidence for the inhalation data and the rationale for the RfC have been
updated and enhanced, with additional discussion of the neurotoxicity data, including the
Gralewicz et al. (1995) study and the addition of a later study by the same group (Gralewicz and
Wiaderna, 2001). The limitations of the Uchida et al. (1993) study are more fully discussed in
Section 4.5.2. Hematological effects reported by Korsak et al. (1992, 1994) are more fully
described.
Question 3. Have the cancer and noncancer assessments been based on the most appropriate
studies? These studies should present the critical effect/cancer (tumors or appropriate precursor)
in the clearest dose-response relationship. If not, what other study (or studies) should be chosen
and why?
Comments: For the RfD, none of the reviewers concurred with the selection of the 13-
week mouse study by NTP (1986) as the principal study. One reviewer suggested that the study
of Marks et al. (1982) be selected as the principal study and two others questioned using the
subchronic portion of the NTP (1986) study, which identified a higher LOAEL, over the chronic
portion of the NTP study.
Regarding the selection of the principal study for deriving the RfC, all three reviewers
agreed that the most appropriate principal study had been chosen. All three reviewers agreed
that the cancer classification was appropriate and that the data were not adequate for derivation
of an oral slope factor or inhalation unit risk, although one suggested that a classification of "no
evidence of carcinogenicity in adequately conducted cancer bioassays in two sexes of two
species" be used. Two reviewers suggested that the limitations of the Maltoni et al. (1983, 1985)
study be more fully discussed.
89
-------
Response to Comments: As per the reviewers' comments, the oral data base has been re-
evaluated, and the NTP (1986) chronic study in rats has been selected as the principal study for
derivation of the RfD, using the critical endpoints of decreased body weight and increased
mortality in male rats. As discussed above, the study by Marks et al. (1982) was not selected
because it identified a considerably higher LOAEL than was identified in NTP (1986). Section
5.1 has been extensively modified, reflecting the change in the principal study.
A more complete discussion of the limitations of the Maltoni et al. (1983, 1985) study is
included in Section 4.6, as per the reviewer's suggestion. A classification of "not likely to be
carcinogenic in humans" was not adopted because although the Maltoni et al. (1983, 1985) study
has several limitations, which were outlined by the reviewer, it was considered to add sufficient
doubt to the classification.
Question 4. Studies included in the RfD and RfC sections under the heading
"Supporting/Additional studies" are meant to lend scientific justification for the designation of
critical effect by including any relevant pathogenesis in humans, any applicable mechanistic
information, and any evidence corroborative of the critical effect or to establish the
comprehensiveness of the data base with respect to various endpoints
(reproductive/developmental toxicity studies, for example). Should other studies be included
under the supporting/additional studies category? Should some studies be removed? Do you
agree with the selection of the NOAEL/LOAEL for determining the RfD, given the manner in
which the data are reported?
Comments: For the RfD, all three reviewers commented that the selection of the
LOAEL of 2000 mg/kg-day from the NTP (1986) subchronic mouse study was not the
appropriate selection, mainly due to difficulties with the critical endpoint, as discussed above.
One reviewer suggested that additional studies or findings be presented to support the LOAEL,
and another suggested the possibility of using Marks et al. (1982) as a co-critical study, as it
identified a similar LOAEL. One reviewer noted that insufficient study detail was provided in
both the principal and supporting studies section and the additional studies/comments section.
One reviewer commented that the selection of the LOAEL for the RfC seemed to be a
reasonable choice; the other two reviewers made no comment as to the appropriateness of the
NOAEL and LOAEL selected. No reviewer expressed dissatisfaction with the evidence in
support of the RfC, although one suggested that the study descriptions be clarified and
NOAEL/LOAEL values be clearly identified where appropriate. One reviewer indicated that
Gralewicz et al. (1995) should be included in the IRIS Summary as a supporting study because
effects were noted at the same point of departure as that of the critical effect.
Response to Comments: The RfD derivation has been changed. A new RfD has been
derived on the basis of decreased body weight and increased mortality in the chronic NTP (1986)
study in rats. The majority of the reviewers' comments regarding the RfD have been addressed
in the replacement of the NOAEL and LOAEL in question with more appropriate values.
Additional detail has been added to study descriptions, including the identification of
90
-------
NOAEL/LOAEL values. The discussion of the RfC rationale has also been expanded and
includes the rationale for selection of critical study and supporting studies.
Question 5. For the noncancer assessments, are there other data that should be considered in
developing the uncertainty factors (UFs) or the modifying factor? Do you consider that the data
support the use of different (default) values than those proposed?
Comments: Two reviewers suggested that the application of pharmacokinetic models be
considered to reduce the uncertainty of extrapolation from animals to humans for the RfD and
RfC. One reviewer felt the existing UFs for the RfD were appropriate and one suggested that a
threefold UF for data base insufficiency would be appropriate rather than the existing 10 in the
external peer review draft because of adequate chronic and subchronic studies in two species and
two developmental studies in rats in the oral data base. This reviewer stated that the inhalation
data base further supports the identification of the same endpoint (i.e., neurotoxicity). This
reviewer also stated that a subchronic-to-chronic extrapolation UF of 3 was warranted. The third
reviewer only commented on the interspecies UF as discussed above.
Regarding the UFs for the RfC, one reviewer commented that the existing UFs (3 each
for dosimetric adjustments, data base insufficiency, and duration extrapolation and 10 for
intraspecies variation) were appropriate. Another reviewer suggested that a threefold UF might
not be sufficient for extrapolation from a subchronic-to-chronic study. This reviewer thought
that a combined UF of 10 should account for subchronic-to-chronic extrapolation and data base
deficiencies.
Response to Comments: As a result of other comments, the previous RfD has been
removed and a new value has been derived on the basis of decreased body weight and increased
mortality in the chronic NTP (1986) rat study; therefore, the current UFs differ from those in the
external peer review draft. The current UFs include a 10 for animal-to-human extrapolation, a
10 for intrahuman variability, and a 10 for deficiencies in the data base, for a total UF of 1000.
The UF of 3 for subchronic-to-chronic extrapolation has been removed, as the critical effect is
now based on a chronic study. The data base UF remains at 10. Additional information is
included in Section 5.1.3.
The UFs for the RfC remain as described because the reviewers were generally in
agreement regarding the choice of UFs for the RfC. A discussion of the potential applicability of
the available PBPK models is included above. The available models do not include an oral input
component, making application to the derivation of the RfD not feasible. Application of the
models to extrapolate exposure levels from animal to humans has been performed to support the
RfC derivation in response to peer review comments. The PBPK model calculations do not alter
the final RfC value. This information is discussed in Sections 3.5, 5.2.2, and 5.2.3 and Appendix
B.
Question 6. Do the confidence statements and weight-of-evidence statements present a clear
rationale and accurately reflect the utility of the studies chosen, the relevancy of the effects
91
-------
(cancer and noncancer) to humans, and the comprehensiveness of the data base? Do these
statements make sufficiently apparent all the underlying assumptions and limitations of these
assessments? If not, what needs to be added?
Comments: Two reviewers suggested that the confidence in the RfD be low rather than
low-to-medium, whereas the third did not suggest a change. All reviewers agreed that the
confidence statement concerning the RfC was appropriate.
Response to Comments: Because a new RfD has been derived, a new statement of
confidence has been prepared, reflecting medium confidence in the new RfD. The confidence
level of the RfC remains at medium.
Scientific Comments from the Public
The commentor questioned the application of a 10 for the data base UF for the RfD. The
commentor stated that there are multiple studies by both the inhalation and oral routes of
exposure and that these studies should be sufficient to reduce the data base UF.
EPA Response: The available oral data base for xylenes includes chronic and subchronic
gavage toxicity studies in mice and rats and a developmental toxicity study. The chronic studies
were conducted largely as cancer bioassays. The data base lacks adequate studies of the oral
neurotoxicity of xylenes as well as multigenerational reproductive toxicity and developmental
neurotoxicity studies.
Given the identification of neurological impairment as a critical health hazard from inhalation
exposure to xylenes, the lack of comprehensive neurotoxicity testing follow ing repeated oral
exposure is of particular concern. There are no toxicokinetic data identifying oral dose levels at
which first-pass hepatic metabolism of xylenes becomes saturated in animals or humans; such
data could decrease uncertainty regarding whether or not neurological impairment may occur at
dose levels below those causing body weight decreases and mortality in rats. For these reasons,
the data base UF for the RfD remains at 10.
The commentor raised issues about the interpretation of results in the Korsak et al. (1994) study
that was used for the derivation of the RfC. Specific issues included the following: (1) the only
significant finding for m-xylene from this study is the rotarod performance decrement effect; and
(2) there is no mention of whether the rotarod tests were conducted immediately post-exposure
or somewhat later. This significant omission makes it much more difficult to judge whether the
performance decrements observed are transient or permanent. The commentor also stated that the
composite UF applied to the RfC is too high given the "mild" nature of the effects reported in the
critical study.
EPA Response: (1) Justification for selection of the critical effect has been augmented in Section
5.2.1. Briefly, additional discussion of supporting studies in which neurological effects have
been reported including altered pain sensitivity, motor coordination, and cognitive ability has
92
-------
been added. Collectively, these data indicate that xylenes are neurotoxic and that decreased
rotarodperformance is a viable critical effect.
(2) In the Korsak et al. (1992) study measurements were made 24 hours post-exposure: given the
similarity between the two studies, a reasonable assumption was made that measurements in the
Korsak et al. (1994) study were conducted using the same protocol. This information has been
added to the study description and Section 5.2.1. Additionally, Gralewicz and Wiaderna (2001)
reported altered pain sensitivity 50 days post exposure indicating a more permanent effect from
exposure to xylenes. As noted in Section 5.2.3.2, application of the Tardif models (Tardifet al.,
1991, 1992, 1993a, 1993b, 1995; Haddadet al., 1999) for inter species extrapolation indicates
that the blood concentration of xylenes would be essentially zero when responses were observed
24 hours post exposure in the Korsak et al. (1992, 1994) studies also suggesting that effects
could be persistent.
The basis for the choice of uncertainty factors has been augmented in light of additional studies
that were identified during the external review process. The text in Section 5.2.3.1 has been
modified accordingly.
The commentor emphasized that the assessment should incorporate recent developments in
pharmacokinetics for the derivation of reference values for xylenes and specifically identified a
pharmacokinetic model by Pelekis et al (2001) which could be used to justify a reduction in the
intraspecies UF for the RfD and RfC. [Pelekis, M; Gephardt, LA; Lerman, SE. (2001)
Physiological-model-based derivation of the adult and child pharmacokinetic intraspecies
uncertainty factors for volatile compounds. Regul. Toxicol. Pharmacol. 33:12-20].
EPA Response: In response to the general comment concerning the application of
pharmacokinetic information in the assessment, a PBPK model was considered for the
derivation of the RfC. This information is now presented in Section 5.2.3.2. The point of
departure derived from the model is essentially the same as that reported from a direct
application of the rat data.
In brief, the Pelekis model was developed to apply pharmacokinetic information to derive a
chemical-specific intraspecies UF. The result of the effort is an informed quantitation of
"normal" human-to-human and adult-to-child variability. The modeling is specific for the
inhalation route of exposure; thus, consideration for the oral route was not possible.
It should be noted that the model lacks a specific component for the target organ, i.e., the brain.
The concept of a chemical-specific intraspecies UFfor xylenes is intriguing; however, the
Pelekis model is based solely on the pharmacokinetic differences between adults and children.
While it is common to divide the interspecies UF into pharmacokinetic andpharmacodynamic
portions, it is not readily evident that this "simple " apportionment would apply to the
intraspecies UF. In the case of intraspecies variability, the differences in humans may be due to
lifestage (childhood versus advanced age), genetic polymorphisms, decreased renal clearance in
disease states, unknown pharmacodynamic variations in response to xylenes exposure, etc. It is
93
-------
not clear that the variability defined in the Pelekis model accounts for the differences in
pharmacokinetics of these various human states. In addition, it is not clear what additional
contributors to intraspecies variability (both pharmacokinetic andpharmacodynamic) would
need to be quantitated and combined to derive a chemical-specific intraspecies UFfor xylenes.
It is likely that a better understanding of the mode of action would be necessary.
94
-------
APPENDIX B. PBPK MODELS FOR m-XYLENE
B.l. Structures of the models
Both rat (Tardif etal., 1991, 1992, 1993a) and human (Tardif et al., 1993b, 1995; Haddad
et al., 1999) PBPK models for m-xylene inhalation have been developed. These models,
developed both individually and for use in modeling of mixtures of other solvents (e.g., benzene,
toluene, and ethylbenzene), predict blood and tissue concentrations from air concentrations on
the basis of partition coefficients, blood flow rates, and ventilation rates.
Conceptually, the models consist of five dynamic tissue compartments, representing the
lung, adipose tissue, slowly perfused tissues, richly perfused tissues, and the liver. A visual
depiction of the model is provided in Figure B2. Inhalation of xylenes is represented by addition
of xylenes to the system via the lung component. Concentration in arterial blood is predicted on
the basis of the existing venous blood concentration, the rate of xylenes exhalation, the inhaled
xylenes concentration, and the blood/gas partition coefficient. The concentration in each tissue
compartment is predicted on the basis of the existing tissue concentration and the arterial
concentration, using appropriate tissue/blood coefficients. Metabolism is assumed, for purposes
of the model, to occur only in the liver compartment and is described by a series of equations
that assume a saturable process characterized by a Vmax (maximal velocity for metabolism, in
mg/hr) and Km (Michaelis-Menten affinity constant). The pooled venous concentration is
calculated as a mean concentration, based on the blood flow rates from each compartment and
the concentration of blood leaving each compartment.
Physiological parameters used in the application of the PBPK models are shown in
Table Bl. For the human models, cardiac output and alveolar ventilation rate were calculated
using the equation Q=18 L/hr-kg x (body weight)0'70. For the rat models, these parameters were
calculated using Q=15 L/hr-kg x (body weight)075. Other physiological parameters for the
models were obtained from the literature (Arms and Travis, 1988; Gargas et al., 1989; Purcell et
al., 1990; Kaneko et al., 1991). Validation of the models following inhalation exposure in both
rats (Tardif et al., 1993a, 1997) and humans (Tardif et al., 1995, 1997) has been reported. These
models have also been applied to mixtures containing xylenes and other aromatic solvents.
B.2. Application of the models to derive human equivalent concentrations (HECs)
If the NOAEL[ADJ] of 39 mg/m3 is used as the exposure concentration for the rat PBPK
model, the model predicts a steady-state pooled venous blood concentration of 0.144 mg/L. This
value was used in the human PBPK model, resulting in an estimated NOAEL[HEC/PK] (continuous
inhaled concentration in humans that would result in a steady-state pooled venous blood
concentration of 0.144 mg/L) of 41 mg/m3. This supports the NOAEL[HEC] of 39 mg/m3
calculated by the standard inhalation dosimetry methods.
95
-------
o
o
CO
13
O
c
-------
Table Bl. Physiological parameters and partition coefficients used in
the PBPK models
Parameters
Alveolar ventilation rate (L/hr-kg)
Cardiac output (L/hr-kg)
Fraction of cardiac output corresponding to each compartment
Values
Rat Human
15.0 18.0
15.0 18.0
Fat 0.09 0.05
Slowly perfused tissues 0.15 0.25
Richly perfused tissues 0.51 0.44
Liver 0.25 0.26
Fraction of body weight corresponding to each compartment
Fat 0.09 0.19
Slowly perfused tissues 0.72 0.62
Richly perfused tissues 0.05 0.05
Liver 0.049 0.026
Partition coefficients
Blood/air 46.0 26.4
Fat/blood 40.4 77.8
Slowly perfused tissues/blood 0.91 3.0
Richly perfused tissues/blood
Liver/blood
1.97 4.42
1.97 3.02
• the maximum (MAX) blood concentration attained on any given day during exposure
(1.09 mg/L; essentially a constant over 13 weeks as shown in Figure B2);
• the mid-point between the maximum (1.09 mg/L) and the minimum (0/mg/L) (MID)
concentration on any given day during exposure (0.55 mg/L).
Using these values as potential dose surrogates in the human model, the model predicted
air concentrations that would produce these steady-state concentrations in human blood with
continuous exposure. As shown in Figures B3, B4, and B5, air concentrations predicted to attain
these steady-state blood concentrations in humans with continuous exposure are 10.5 ppm (46
mg/m3) for the TWA surrogate, 27.4 ppm (106 mg/m3) for the MID, and 49.8 ppm (216 mg/m3)
for the MAX.
The rat model predicts that blood concentrations were essentially zero when the critical
effects on rotarod performance were measured (24 hours after cessation of exposure). This
supports the idea that the observed effects are not dependent on the concurrent presence of
97
-------
xylenes in the blood and that they may be a persistent neurological effect. Brain concentrations
would be a better dose surrogate to use in the model, but the model has not been developed to
predict brain concentrations of xylenes.
The TWA dose surrogate is likely to provide a more accurate description of the exposure
experienced by the rats in the study than do the MID or MAX dose surrogates, especially since
the effects were measured after m-xylene had completely cleared from the blood. Thus, using
the TWA as the dose surrogate in extrapolating from the rats in the Korsak et al. (1994) study to
humans, the model predicts a HEC of 46.5 mg/m3. This is very similar, although not identical, to
the HEC (39 mg/m3) predicted using the default NOAEL[HEC] dosimetry methodology, and it
provides support for the RfC derived in Section 5.2.3.1.
98
-------
I.U
1 A
\.l\
1 9
1.0
.§0.8-
X
o
0.6
0.4
0.2
n n
--
--
--
---
---
Rat (50ppm):
TWA=0.198mg/L
MAX=1.095mg/L
MID=0.55 mg/L
0
500
1000 1500
Time (hr)
2000
2500
Figure B2.
-------
0.25
0.20
0.15
O)
E
O
o
o
0.10
0.05
0.00
Human (10.5 ppm):
SS=0.198mg/L
0 50
100 150 200 250 300 350 400 450 500
Time (hr)
Figure 11
-------
0.80
0.70
0.60
0.50
O)
x
O
0.40
0.30
0.20
0.10
0.00
Human (27.4 ppm):
SS=0.55 mg/L
50 100
150 200 250 300 350 400 450 500
Time (hr)
Figure B4.
-------
1.4
1.2
o
to
X
<
O 0.6
0.4
0.2
0.0
Human (49.8 ppm):
SS=1.095 mg/L
0 50 100 150
200 250 300
Time (hr)
350 400 450 500
Figure B5.
-------
APPENDIX C. BENCHMARK DOSE ANALYSIS OF WOLFE ET AL. (1988A)
The decrease in body weight in rats exposed orally to up to 800 mg/kg-day of m-xylene
(Wolfe, 1988a) was analyzed using the models for continuous data in the EPA Benchmark Dose
Software (version 1.3.1.). An appropriate fit was generated using a linear model; model outputs
are included below. A 10% change in body weight was used as the benchmark response. The
resulting BMDL value of 440 mg/kg-day is included for comparison purposes in the derivation
of the RfD. Derivation of an RfD from this point of departure, using the same uncertainty
factors (UFs) as noted in Section 5.1.3., but with an additional UF of 3 for extrapolation from
subchronic to chronic duration (total UF of 3000), would result in a value of 0.1 mg/kg-day,
which is similar to the RfD of 0.2 mg/kg-day in Section 5.1.3.
Linear Model with 0.95 Confidence Level
560
540
520
500
w
CD
Ct
480
460
440
Linear
0 100 200
300
400
dose
500
600
700
800
13:3708/142002
103
-------
Polynomial Model. SRevision: 2.1 $ $Date: 2000/10/11 17:51:39 $
Input Data File: C:\EPA TOX REVIEWS\XYLENES\BMDS
STUFF\WOLFE\WOLFE_A_BODY_WEIGHTS.(d)
Gnuplot Plotting File: C:\EPA TOX REVIEWS\XYLENES\BMDS
STUFF\WOLFE\WOLFE_A_BODY_WEIGHTS.plt
WedAugl4 13:37:502002
BMDS MODEL RUN
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2
Dependent variable = MEAN
Independent variable = Dose
rho is set to 0
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 1
rho = 0 Specified
beta_0= 523.115
beta 1 = -0.0966903
104
-------
Parameter Estimates
Variable
alpha
beta_0
beta 1
Estimate
1364.87
523.849
-0.0972985
Std. Err.
230.706
5.81841
0.013886
Asymptotic Correlation Matrix of Parameter Estimates
alpha beta_0 beta_l
alpha 1 1.8e-006 -1.4e-006
beta_0 1.8e-006 1 -0.65
beta 1 -1.4e-006 -0.65 1
Table of Data and Estimated Values of Interest
Dose N Obs Mean Obs Std Dev Est Mean Est Std Dev ChiA2
Res.
0 20
100 17
200 15
800 18
528
518
492
448
46.3
37.5
31.3
30.4
524
514
504
446
36.9
36.9
36.9
36.9
2.14
1.83
-4.99
1.02
Model Descriptions for likelihoods calculated
Model Al: Yij = Mu(I) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(I) + e(ij)
Var{e(ij)} = Sigma(I)A2
Model R: Yi = Mu + e(I)
Var{e(I)} = SigmaA2
105
-------
44.8298
4.44439
2.17702
6
O
2
<.0001
0.2173
0.3367
Likelihoods of Interest
Model Log(likelihood) DF AIC
Al -286.570013 5 583.140025
A2 -284.347816 8 584.695632
fitted -287.658524 2 579.317047
R -306.762705 2 617.525410
Test 1: Does response and/or variances differ among dose levels
(A2 vs. R)
Test 2: Are Variances Homogeneous (Al vs A2)
Test 3: Does the Model for the Mean Fit (Al vs. fitted)
Tests of Interest
Test -2*log(Likelihood Ratio) Test df p-value
Test 1
Test 2
Test3
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels.
It seems appropriate to model the data
The p-value for Test 2 is greater than .05. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .05. The model chosen appears
to adequately describe the data
Benchmark Dose Computation
Specified effect = 0.1
Risk Type = Relative risk
Confidence level = 0.95
BMD= 538.393
BMDL = 440.224
106
------- |