DRAFT DO NOT CITE OR QUOTE
EPA/635/R-ll/012Aa
www.epa.gov/iris
United States
E n v i ro n m e n f a I Pro t CC t i 0 n
Agency
lexicological Review of Trimethylbenzenes
(CAS No. 25551-13-7, 95-63-6, 526-73-8, and 108-67-8}
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
June 2012
NOTICE
This document is an External Review draft. This information is distributed solely for the
purpose of pre-dissemination peer review under applicable information quality guidelines.
It has not been formally disseminated by EPA. It does not represent and should not be
construed to represent any Agency determination or policy. It is being circulated for
review of its technical accuracy and science policy implications.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
-------
Toxicological Review of Trimethylbenzene
1 DISCLAIMER
2
3 This document is a preliminary draft for review purposes only. This information is
4 distributed solely for the purpose of pre-dissemination peer review under applicable
5 information quality guidelines. It has not been formally disseminated by EPA. It does not
6 represent and should not be construed to represent any Agency determination or policy.
7 Mention of trade names or commercial products does not constitute endorsement of
8 recommendation for use.
This document is a draft for review purposes only and does not constitute Agency policy.
ii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
CONTENTS
4
5 AUTHORS | CONTRIBUTORS | REVIEWERS viii
6 PREFACE x
7 PREAMBLE TO IRIS TOXICOLOGICAL REVIEWS xiii
8 EXECUTIVE SUMMARY xxvi
9 LITERATURE SEARCH STRATEGY | STUDY SELECTION xxxvii
10 1. HAZARD IDENTIFICATION 1-1
11 1.1. Synthesis of Evidence 1-1
12 1.1.1. Neurological Effects 1-1
13 1.1.2. Respiratory Effects 1-21
14 1.1.3. Reproductive and Developmental Effects 1-28
15 1.1.4. Hematological and Clinical Chemistry Effects 1-32
16 1.1.5. Carcinogenicity 1-39
17 1.1.6. Similarities Among TMB Isomers Regarding Observed Inhalation and Oral Toxicity 1-40
18 1.2. Summary and Evaluation 1-42
19 1.2.1. Weight of Evidence for Effects Other than Cancer 1-42
20 1.2.2. Weight of Evidence for Carcinogenicity 1-44
21 1.2.3. Susceptible Populations and Lifestages 1-45
22 2. DOSE-RESPONSE ANALYSIS 2-1
23 2.1. Inhalation Reference Concentration for Effects Other Than Cancer for 1,2,4-TMB 2-1
24 2.1.1. Identification of Candidate Principal Studies and Critical Effects for 1,2,4-TMB 2-1
25 2.1.2. Methods of Analysis for 1,2,4-TMB 2-6
26 2.1.3. Derivation of the Reference Concentration for 1,2,4-TMB 2-9
27 2.1.4. Uncertainties in the Derivation of the Reference Concentration for 1,2,4-TMB 2-12
28 2.1.5. Confidence Statement for 1,2,4-TMB 2-13
29 2.1.6. Comparison of Candidate Reference Concentrations for 1,2,4-TMB 2-14
30 2.2. Inhalation Reference Concentration for Effects Other Than Cancer for 1,2,3-TMB 2-14
31 2.2.1. Identification of Candidate Principal Studies and Critical Effects for 1,2,3-TMB 2-14
32 2.2.2. Methods of Analysis for 1,2,3-TMB 2-18
33 2.2.3. Derivation of the Reference Concentration for 1,2,3-TMB 2-20
34 2.2.4. Uncertainties in the Derivation of the Reference Concentration for 1,2,3-TMB 2-25
35 2.2.5. Confidence Statement for 1,2,3-TMB 2-26
36 2.2.6. Comparison of Candidate Reference Concentrations for 1,2,3-TMB 2-26
37 2.3. Inhalation Reference Concentration for Effects Other Than Cancer for 1,3,5-TMB 2-27
38 2.3.1. Identification of Candidate Principal Studies and Critical Effects for 1,3,5-TMB 2-27
39 2.3.2. Methods of Analysis for 1,3,5-TMB 2-31
This document is a draft for review purposes only and does not constitute Agency policy.
iii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 2.3.3. Derivation of the Reference Concentration for 1,3,5-TMB 2-34
2 2.3.4. Uncertainties in the Derivation of the Reference Concentration for 1,3,5-TMB 2-41
3 2.3.5. Confidence Statement for 1,3,5-TMB 2-41
4 2.4. Oral Reference Dose for Effects Other Than Cancer for 1,2,4-TMB 2-42
5 2.4.1. Identification of Candidate Principal Studies and Critical Effects for 1,2,4-TMB 2-42
6 2.4.2. Methods of Analysis for 1,2,4-TMB 2-42
7 2.4.3. Derivation of the Reference Dose for 1,2,4-TMB 2-43
8 2.4.4. Uncertainties in the Derivation of the Reference Dose for 1,2,4-TMB 2-43
9 2.4.5. Confidence Statementfor 1,2,4-TMB 2-44
10 2.5. Oral Reference Dose for Effects Other Than Cancer for 1,2,3-TMB 2-44
11 2.5.1. Identification of Candidate Principal Studies and Critical Effects for 1,2,3-TMB 2-44
12 2.5.2. Methods of Analysis and Derivation of the Reference Dose for 1,2,3-TMB 2-44
13 2.5.3. Uncertainties in the Derivation of the Reference Dose for 1,2,3-TMB 2-45
14 2.5.4. Confidence Statement for 1,2,3-TMB 2-45
15 2.6. Oral Reference Dose for Effects Other Than Cancer for 1,3,5-TMB 2-45
16 2.6.1. Identification of Candidate Principal Studies and Critical Effects for 1,3,5-TMB 2-45
17 2.6.2. Methods of Analysis and Derivation of the Reference Dose for 1,3,5-TMB 2-46
18 2.6.3. Uncertainties in the Derivation of the Reference Dose for 1,3,5-TMB 2-46
19 2.6.4. Confidence Statement for 1,3,5-TMB 2-46
20 2.7. Cancer Assessment for 1,2,3-TMB, 1,2,4-TMB, and 1,3,5-TMB 2-47
21 REFERENCES R-l
22
23
This document is a draft for review purposes only and does not constitute Agency policy.
iv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
i TABLES AND FIGURES
2 Table ES-1. Summary of inhalation reference concentrations (RfCs) xxvii
3 Table ES-2. Summary of reference concentration (RfC) derivation for 1,2,4-TMB xxviii
4 Table ES-3. Summary of reference concentration (RfC) derivation for 1,2,3-TMB xxix
5 Table ES-4. Summary of reference doses (RfDs) for TMB isomers xxxii
6 Table ES-5. Summary of reference dose (RfD) derivation for 1,2,4-TMB xxxiii
7 Table LS-1. Details of the search strategy employed forTMBs xxxvii
8 Figure LS-1. Literature search and study selection strategy for TMBs xxxviii
9 Table 1-1. Evidence pertaining to neurological effects of TMBs in animals — inhalation
10 exposures 1-9
11 Table 1-2. Evidence pertaining to neurological effects of 1,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB in
12 animals — oral exposures 1-13
13 Figure 1-1. Exposure response array of neurological effects following inhalation exposure to
14 1,2,4-TMB 1-15
15 Figure 1-2. Exposure response array of neurological effects following inhalation exposure to
16 1,2,3-TMB 1-16
17 Figure 1-3. Exposure response array of neurological effects following inhalation exposure to
18 1,3,5-TMB 1-17
19 Figure 1-4. Exposure response array of neurological effects following oral exposure to 1,2,3-
20 TMB, 1,2,4-TMB, or 1,3,5-TMB 1-18
21 Table 1-3. Evidence pertaining to respiratory effects of TMBs in animals — inhalation
22 exposures 1-24
23 Figure 1-5. Exposure response array of respiratory effects following inhalation exposure to
24 1,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB 1-26
25 Table 1-4. Evidence pertaining to reproductive and developmental effects of 1,2,4-TMB and
26 1,3,5-TMB in animals — inhalation exposures 1-30
27 Figure 1-6. Exposure response array of reproductive and developmental effects following
28 inhalation exposure to 1,2,4-TMB or 1,3,5-TMB 1-31
29 Table 1-5. Evidence pertaining to hematological and clinical chemistry effects of 1,2,3-TMB,
30 1,2,4-TMB, or 1,3,5-TMB in animals — inhalation exposures 1-34
31 Figure 1-7. Exposure response array of hematological and clinical chemistry effects following
32 inhalation exposure to 1,2,4-TMB or 1,3,5-TMB 1-37
33 Figure 1-8. Exposure response array of hematological and clinical chemistry effects following
34 inhalation exposure to 1,2,3-TMB 1-38
35 Table 1-6. Similarities between 1,2,3-TMB, 1,2,4-TMB, and 1,3,5-TMB regarding observed
36 inhalation and oral toxicity 1-42
37 Table 2-1. Internal blood dose metrics calculated using the available rat PBPK model (Hissink et
38 al., 2007) 2-3
39 Table 2-2. Endpoints resulting from subchronic inhalation exposure to 1,2,4-TMB considered
40 forthe derivation of the RfC 2-4
This document is a draft for review purposes only and does not constitute Agency policy.
v DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Figure 2-1. Exposure response array of endpoints resulting from inhalation exposure to 1,2,4-
2 TMB considered for the derivation of the RfC 2-6
3 Table 2-3. Summary of dose-response analysis and point of departure estimation for endpoints
4 resulting from subchronic inhalation exposure to 1,2,4-TMB 2-8
5 Table 2-4. PODADJ values, human equivalent concentrations (HECs), uncertainty factors, and
6 candidate RfCs for 1,2,4-TMB 2-9
7 Figure 2-2. Array of candidate PODHEc values with applied UFs and candidate RfCs for
8 neurological, respiratory, and hematological effects resulting from inhalation
9 exposure to 1,2,4-TMB 2-14
10 Table 2-5. Target and actual exposure concentrations used in BMD modeling of 1,2,3-TMB
11 endpoints considered for the derivation of the RfC 2-16
12 Table 2-6. Endpoints resulting from subchronic inhalation exposure to 1,2,3-TMB considered
13 for the derivation of the RfC 2-17
14 Figure 2-3. Exposure response array for endpoints resulting from inhalation exposure to 1,2,3-
15 TMB considered for the derivation of the RfC 2-18
16 Table 2-7. Summary of dose-response analysis and point of departure estimation for endpoints
17 resulting from subchronic inhalation exposure to 1,2,3-TMB 2-20
18 Table 2-8. PODADJ values, human equivalent concentrations (HECs), uncertainty factors, and
19 candidate RfCs for 1,2,3-TMB 2-22
20 Figure 2-4. Array of candidate PODHEc values with applied UFs and candidate RfCs for
21 neurological respiratory, and hematological effects resulting from inhalation
22 exposure to 1,2,3-TMB 2-27
23 Table 2-9. Target and actual exposure concentrations used in BMD modeling of 1,3,5-TMB
24 endpoints considered for the derivation of the RfC 2-29
25 Table 2-10. Endpoints resulting from inhalation exposure to 1,3,5-TMB considered for the
26 derivation of the RfC 2-30
27 Figure 2-5. Exposure response array for endpoints resulting from inhalation exposure to
28 1,3,5-TMB considered for the derivation of the RfC 2-31
29 Table 2-11. Summary of dose-response analysis and point of departure estimation for endpoints
30 resulting from short-term and gestational inhalation exposures to 1,3,5-TMB 2-34
31 Table 2-12. PODADJ values, human equivalent concentrations (HECs), uncertainty factors, and
32 candidate RfCs for 1,3,5-TMB 2-36
33
34
This document is a draft for review purposes only and does not constitute Agency policy.
vi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
ABBREVIATIONS AND ACRONYMS
AAQC Ambient air quality criterion OSHA
ACGIH American Conference of
Governmental Industrial Hygienists p
ADME absorption, distribution, metabolism PBPK
and excretion
AEGL Acute Exposure Guideline Levels PEL
AIC Akaike Information Criterion POD
BAL bronchoalveolar lavage PODADj
BMD benchmark dose POI
BMDL lower confidence limit on the ppm
benchmark dose RBC
BMDS benchmark dose software RDso
BMR benchmark response REL
BW body weight RfC
CAS Chemical Abstracts Service RfD
CASRN Chemical Abstracts Service Registry RGDR
Number ROS
CI confidence interval SCE
CNS central nervous system SD
CYP450 cytochrome P450 SOA
DAF dosimetric adjustment factor TLV
DMBA dimethylbenzoic acid TMB
DMHA dimethylhippuric acid TSCA
DNA deoxyribonucleic acid TWA
ECso half maximal effective concentration UF
EEC electroencephalogram UFA
EPA U.S. Environmental Protection UFn
Agency UFs
GD gestational day
Hb/g-A animal blood:gas partition coefficient UFt
Hb/g-H human blood:gas partition coefficient UFD
HEC human equivalent concentration
HER human epidermal keratinocytes UV
HERO Health and Environmental Research VOC
Online WBC
HEV human epithelial keratinocytes WS
HSDB Hazardous Substances Data Bank %2
IL-8 interleukin-8
i.p. intraperitoneal
IRIS Integrated Risk Information System
JP-8 jet propulsion fuel 8
Km Michaelis-Menten constant
LDH lactate dehydrogenase
LOAEL lowest-observed-adverse-effect level
NCEA National Center for Environmental
Assessment
NIOSH National Institute for Occupational
Safely and Health
NLM National Library of Medicine
NOAEL no-observed-adverse-effect level
OMOE Ontario Ministry of the Environment
Occupational Safety and Health
Administration
probability value
physiologically based
pharmacokinetic (model)
permissible exposure limit
point of departure
duration adjusted POD
point of impingement
parts per million
red blood cell
50% respiratory rate decrease
recommended exposure limit
reference concentration
reference dose
regional gas dose ratio
reactive oxygen species
sister chromatid exchange
standard deviation
secondary organic aerosol
threshold limit value
trimethylbenzene
Toxic Substances Control Act
time-weighted average
uncertainty factor
interspecies uncertainty factor
intraspecies uncertainty factor
subchronic-to-chronic uncertainty
factor
LOAEL-to-NOAEL uncertainty factor
database deficiency uncertainty
factor
ultraviolet
volatile organic compound
white blood cell
white spirit
chi-squared
This document is a draft for review purposes only and does not constitute Agency policy.
vii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
AUTHORS | CONTRIBUTORS | REVIEWERS
Assessment Team
J. Allen Davis, M.S.P.H. (Chemical Manager)
Eva McLanahan, Ph.D. (LCDR, USPHS)
Paul Schlosser, Ph.D.
John Cowden, Ph.D.
Gary Foureman, Ph.D. (Currently ICF Int.)
Andrew Kraft, Ph.D.
Ray Antonelli, B.S.
U.S. Environmental Protection Agency
Office of Research and Development
National Center for Environmental Assessment
Research Triangle Park, NC
U.S. Environmental Protection Agency
Office of Research and Development
National Center for Environmental Assessment
Washington, DC
Oak Ridge Institute for Scientific Education
Research Triangle Park, NC
Scientific Support Team
Reeder Sams, Ph.D.
John Stanek, Ph.D.
Rob Dewoskin, Ph.D.
George Woodall, Ph.D.
Geniece Lehmann, Ph.D.
Connie Meacham, M.S.
Martin Gehlhaus, MHS
Nina Wang, Ph.D.
U.S. Environmental Protection Agency
Office of Research and Development
National Center for Environmental Assessment
Research Triangle Park, NC
U.S. Environmental Protection Agency
Office of Research and Development
National Center for Environmental Assessment
Washington, DC
U.S. Environmental Protection Agency
Office of Research and Development
National Center for Environmental Assessment
Cincinnati, OH
Production Team
Ellen Lorang, M.S.
Deborah Wales
Gerald Gurevich
J. Sawyer Lucy
Maureen Johnson
Vicki Soto
U.S. Environmental Protection Agency
Office of Research and Development
National Center for Environmental Assessment
Research Triangle Park, NC
U.S. Environmental Protection Agency
Office of Research and Development
National Center for Environmental Assessment
Washington, DC
This document is a draft for review purposes only and does not constitute Agency policy.
viii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Contractor Support
Battelle Memorial Institute, Pacific Northwest Division, Richmond, WA
Karla D. Thrall, Ph.D.
Battelle Memorial Institute, Columbus, OH
Jessica D. Sanford, Ph.D.
Maureen A. Wooton
Robert A. Lordo, Ph.D.
Anthony Fristachi
Toxicology Excellence for Risk Assessment (TERA)
Under Battelle Memorial Institute Contract EP-C-09-006
Lisa M. Sweeney, Ph.D., DABT
Melissa J. Kohrman-Vincent, B.A.
Executive Direction
Reeder Sams, Ph.D.
Lyle Burgoon, Ph.D.
John Vandenberg, Ph.D.
Debra Walsh, M.S.
Rebecca Clark, M.S.P.H.
Vincent Cogliano, Ph.D.
Samantha Jones, Ph.D.
Lynn Flowers, Ph.D.
Jamie Strong, Ph.D.
U.S. Environmental Protection Agency
Office of Research and Development
National Center for Environmental Assessment
Research Triangle Park, NC
U.S. Environmental Protection Agency
Office of Research and Development
National Center for Environmental Assessment
Washington, DC
2 Reviewers
3
4
5
6
7
This assessment was provided for review to scientists in EPA's Program and Regional
Offices. Comments were submitted by:
Region 8, Denver, CO
Region 2, New York City, NY
Office of Children's Health Protection, Washington, DC
This assessment was provided for review to other federal agencies and the Executive
Offices of the President. Comments were submitted by:
Chemical Material Risk Management Program, Department of Defense
Agency for Toxic Substances and Disease Registry, Centers for Disease Control and Prevention,
Department of Health & Human Services
The National Institute for Occupational Safety and Health, Centers for Disease Control and
Prevention, Department of Health & Human Services
National Toxicology Program, National Institute of Environmental Health Sciences, National
Institutes of Health, Department of Health & Human Services
Council on Environmental Quality, Executive Office of the President
This document is a draft for review purposes only and does not constitute Agency policy.
ix DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1
2 PREFACE
3 This Toxicological Review critically reviews the publicly available studies on the three
4 isomers of trimethylbenzene (i.e., 1,2,3-trimethylbenzene [1,2,3-TMB], 1,2,4-trimethylbenzene
5 [1,2,4-TMB], and 1,3,5-trimethylbenzene [1,3,5-TMB]) in order to identify their adverse health
6 effects and to characterize exposure-response relationships. Because more types of studies are
7 available for the 1,2,4-TMB isomer, it generally appears first when the individual isomers are listed.
8 This assessment was prepared under the auspices of EPA's Integrated Risk Information System
9 (IRIS) program.
10 This assessment was prepared because of the presence of trimethylbenzenes (TMB) at
11 Superfund sites. Of sites on EPA's National Priorities List that report TMB isomer contamination
12 (38 sites), 93% report 1,3,5-TMB contamination, 85% report 1,2,4-TMB contamination, 12% report
13 1,2,3-TMB contamination, and 17% report contamination by unspecified TMB isomers.
14 The Toxicological Review of Trimethylbenzenes is a new assessment; there is no previous
15 entry on the IRIS Database for 1,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB. This assessment reviews
16 information on all health effects by all exposure routes.
17 This assessment was conducted in accordance with EPA guidance, which is cited and
18 summarized in the Preamble to IRIS Toxicological Reviews. The findings of this assessment and
19 related documents produced during its development are available on the IRIS website
20 (http://www.epa.gov/iris). Appendices for chemical and physical properties, toxicokinetic
21 information, summaries of toxicity studies, and other supporting materials are provided as
22 Supplemental Information (See Appendix A to C).
23 On December 23, 2011, The Consolidated Appropriations Act, 2012, was signed into law1.
24 The report language included direction to EPA for the IRIS Program related to recommendations
25 provided by the National Research Council (NRC) in their review of EPA's draft IRIS assessment of
26 formaldehyde. The NRC's recommendations, provided in Chapter 7 of their review report, offered
27 suggestions to EPA for improving the development of IRIS assessments. The report language
28 included the following:
29
30 The Agency shall incorporate, as appropriate, based on chemical-specific datasets
31 and biological effects, the recommendations of Chapter 7 of the National Research
32 Council's Review of the Environmental Protection Agency's Draft IRIS Assessment of
'Pub. L. No. 112-74, Consolidated Appropriations Act, 2012.
This document is a draft for review purposes only and does not constitute Agency policy.
x DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Formaldehyde into the IRIS process...For draft assessments released in fiscal year
2 2012, the Agency shall include documentation describing how the Chapter 7
3 recommendations of the National Academy of Sciences (NAS) have been
4 implemented or addressed, including an explanation for why certain
5 recommendations were not incorporated.
6
7 Consistent with the direction provided by Congress, documentation of how the recommendations
8 from Chapter 7 of the NRC report have been implemented in this assessment is provided in
9 Appendix D. This documentation also includes an explanation for why certain recommendations
10 were not incorporated.
11 For additional information about this assessment or for general questions regarding IRIS,
12 please contact EPA's IRIS Hotline at 202-566-1676 (phone), 202-566-1749 (fax), or
13 hotline.iris@epa.gov.
14 Chemical Properties and Uses
15 TMBs are aromatic hydrocarbons with three methyl groups attached to a benzene ring and
16 the chemical formula CgHi2. The chemical and physical properties of the TMB isomers are similar to
17 one another. TMBs are colorless, flammable liquids with a strong aromatic odor; an odor threshold
18 of 0.4 parts per million (ppm) of air has been reported (U.S. EPA, 1994a). They are insoluble in
19 water but miscible with organic solvents such as ethyl alcohol, benzene, and ethyl ether (OSHA.
20 1996). Production and use of TMBs may result in their release to the environment through various
21 waste streams. If released to the atmosphere, 1,2,3-TMB, 1,2,4-TMB, and 1,3,5-TMB will exist solely
22 in the vapor phase in the atmosphere under ambient conditions, based on measured vapor
23 pressures of 1.69, 2.10, and 2.48 mm Hg at 25°C, respectively (HSDB. 2011a. b, c). All three isomers
24 are expected to have limited mobility through soil based on their Log Koc values, but are expected to
25 volatilize from both moist and dry soil surfaces and surface waters based on their respective
26 Henry's law constants and vapor pressures (see Appendix B, Table B-l). Degradation of both
27 isomers in the atmosphere occurs by reaction with hydroxyl radicals, the half-life of which is 11-12
28 hours (HSDB, 2011a, b, c). Non-volatilized TMBs may be subject to biodegradation under aerobic
29 conditions (HSDB. 2011a. b, c). The estimated bio-concentration factors (133-439) and high
30 volatility of TMBs suggest that bioaccumulation of these chemicals will not be significant (U.S. EPA.
31 1987). Additional information on the chemical identities and physicochemical properties of TMBs
32 are listed in Table B-l in Appendix B.
33 The commercially available substance known as trimethylbenzene, CAS No. 25551-13-7, is a
34 mixture of three isomers in various proportions, namely CAS No. 526-73-8 (1,2,3-TMB or
35 hemimellitene), CAS No. 95-63-6 (1,2,4-TMB or pseudocumene), and CAS No. 108-67-8 (1,3,5-TMB
36 or mesitylene). Production of TMB isomers occurs during petroleum refining, and 1,2,4-TMB
This document is a draft for review purposes only and does not constitute Agency policy.
xi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 individually makes up approximately 40% of the C9 aromatic fraction (i.e., aromatic hydrocarbons
2 with nine carbons) [U.S. EPA. 1994a). The domestic production of the C9 fraction in 1991 was
3 estimated to be approximately 80 billion pounds (40 million tons) (U.S. EPA, 1994a). Vehicle
4 emissions are a major anthropogenic source of TMBs, due to the widespread use of the C9 fraction
5 as a component of gasoline (U.S. EPA, 1994a). Other uses of TMBs include solvents in research and
6 industry dyestuff intermediate, paint thinner, and as a UV oxidation stabilizer for plastics (HSDB.
7 2011b.cj.
8 Occupational levels of exposure for TMBs have been measured between 20-8,540 |ig/m3
9 (HSDB. 2011a. b, cj Tiun-Horngetal.. 2008). whereas residential exposures are generally much
10 lower: 0.29-7.8 [ig/m3 (Martins etal.. 2010: Choi etal.. 2009: Guo etal.. 2009). Total atmospheric
11 releases of 1,2,4-TMB to the environment in 2008 equaled 5.8 million pounds (2,900 tons), 265,000
12 pounds (132.5 tons) were released to surface waters, underground injection sites, or land (TRI,
13 20081 No information is currently available regarding 1,2,3-TMB or 1,3,5-TMB releases.
14 Assessments by Other National and International Health Agencies
15 Toxicity information on 1,2,4-TMB, 1,2,3-TMB, and 1,3,5-TMB has been evaluated by the
16 National Institute for Occupational Safety and Health (NIOSH), the American Conference of
17 Governmental Industrial Hygienists (ACGIH), the National Advisory Committee for Acute Exposure
18 Guideline Levels for Hazardous Substances, and the Ontario Ministry of the Environment (MOE).
19 The results ofthese assessments are summarized in Appendix A (Table A-l). Itis importantto
20 recognize that these assessments may have been prepared for different purposes and may utilize
21 different methods, and that newer studies may be included in the IRIS assessment.
22
This document is a draft for review purposes only and does not constitute Agency policy.
xii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
PREAMBLE TO IRIS TOXICOLOGICAL REVIEWS
i 1. Scope of the IRIS Program
2 Soon after EPA was established in 1970, it
3 was at the forefront of developing risk
4 assessment as a science and applying it in
5 decisions to protect human health and the
6 environment. The Clean Air Act, for example,
7 mandates that EPA provide "an ample margin of
8 safety to protect public health"; the Safe Drinking
9 Water Act, that "no adverse effects on the health
10 of persons may reasonably be anticipated to
11 occur, allowing an adequate margin of safety."
12 Accordingly, EPA uses information on the
13 adverse effects of chemicals and on exposure
14 levels below which these effects are not
15 anticipated to occur.
16 IRIS assessments critically review the
17 publicly available studies to identify adverse
18 health effects from long-term exposure to
19 chemicals and to characterize exposure-response
20 relationships. An assessment may cover a single
21 chemical, a group of structurally or
22 lexicologically related chemicals, or a complex
23 mixture. Exceptions are chemicals currently used
24 exclusively as pesticides, ionizing and non-
25 ionizing radiation, and criteria air pollutants
26 listed under section 108 of the Clean Air Act
27 (carbon monoxide, lead, nitrogen oxides, ozone,
28 particulate matter, and sulfur oxides; EPA's
29 Integrated Science Assessments evaluate the
30 effects from these pollutants in ambient air).
31 Periodically, the IRIS Program asks other
32 EPA programs and regions, other federal
33 agencies, state government agencies, and the
34 general public to nominate chemicals and
35 mixtures for future assessment or reassessment.
36 These agents may be found in air, water, soil, or
37 sediment. Selection is based on program and
38 regional office priorities and on availability of
39 adequate information to evaluate the potential
40 for adverse effects. IRIS may assess other agents
41 as an urgent public health need arises. IRIS also
42 reassesses agents as significant new studies are
43 published.
44 2. Process for developing and peer-
45 reviewing IRIS assessments
46 The process for developing IRIS assessments
47 (revised in May 2009) involves critical analysis of
48 the pertinent studies, opportunities for public
49 input, and multiple levels of scientific review.
50 EPA revises draft assessments after each review,
51 and external drafts and comments become part
52 of the public record (U.S. EPA. 2009).
53 Step 1. Development of a draft Toxicological
54 Review (usually about 11-1/2 months
55 duration). The draft assessment considers all
56 pertinent publicly available studies and
57 applies consistent criteria to evaluate the
58 studies, identify health effects, weigh the
59 evidence of causation for each effect, identify
60 mechanistic events and pathways, and derive
61 toxicity values.
62 Step 2. Internal review by scientists in EPA
63 programs and regions (2 months). The
64 draft assessment is revised to address
65 comments from within EPA.
66 Step 3. Interagency science consultation with
67 other federal agencies and the Executive
68 Offices of the President (1-1/2 months).
69 The draft assessment is revised to address
70 the interagency comments. The science
71 consultation draft, interagency comments,
72 and EPA's response to major comments
73 become part of the public record.
74 Step 4. External peer review, after public
75 review and comment (3-1/2 months or
76 more, depending on the review process).
77 EPA releases the draft assessment for public
78 review and comment, followed by external
79 peer review. The peer review meeting is
80 open to the public and includes time for oral
81 public comments. The peer reviewers also
82 receive the written public comments. The
83 peer reviewers assess whether the evidence
84 has been assembled and evaluated according
85 to guidelines and whether the conclusions
This document is a draft for review purposes only and does not constitute Agency policy.
xiii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 are justified by the evidence. The peer
2 review draft, peer review report, and written
3 public comments become part of the public
4 record.
5 Step 5. Revision of draft Toxicological Review
6 and development of draft IRIS summary
7 (2 months). The draft assessment is revised
8 to reflect the peer review comments, public
9 comments, and newly published studies that
10 are critical to the conclusions of the
11 assessment. The disposition of peer review
12 comments and public comments becomes
13 part of the public record.
14 Step 6. Final EPA review and interagency
15 science discussion with other federal
16 agencies and the Executive Offices of the
17 President [1-1/2 months). The draft
18 assessment and summary are revised to
19 address EPA and interagency comments. The
20 science discussion draft, written interagency
21 comments, and EPA's response to major
22 comments become part of the public record.
23 Step 7. Completion and posting (1 month). The
24 Toxicological Review and IRIS summary are
25 posted on the IRIS website (http://
26 www.epa.gov/iris/).
27
28 The remainder of this Preamble addresses
29 step 1, the development of a draft Toxicological
30 Review. IRIS assessments follow standard
31 practices of evidence evaluation and peer review,
32 many of which are discussed in EPA guidelines
33 [U.S. EPA, 2005a, b, 2000b, 1998, 1996, 1991,
34 1986a. b) and other methods fU.S. EPA. 2011b.
35 2006a. b, 2002. 2000a. 1994). Transparent
36 application of scientific judgment is of
37 paramount importance. To provide a harmonized
38 approach across IRIS assessments, this Preamble
39 summarizes concepts from these guidelines and
40 emphasizes principles of general applicability.
41 3. Identifying and selecting
42 pertinent studies
43 3.1. Identifying studies
44 Before beginning an assessment, EPA
45 conducts a comprehensive search of the primary
46 scientific literature. The literature search follows
47 standard practices and includes the PubMed and
48 ToxNet databases of the National Library of
49 Medicine and other databases listed in EPA's
50 HERO system (Health and Environmental
51 Research Online, http://hero.epa.gov/). Each
52 assessment specifies the search strategies,
53 keywords, and cut-off dates of its literature
54 searches. EPA posts the results of the literature
55 search on the IRIS website and requests
56 information from the public on additional studies
57 and ongoing research.
58 EPA also considers studies received through
59 the IRIS Submission Desk and studies (typically
60 unpublished) submitted under the Toxic
61 Substances Control Act. Material submitted as
62 Confidential Business Information is considered
63 only if it includes health and safety data that can
64 be publicly released. If a study that may be
65 critical to the conclusions of the assessment has
66 not been peer-reviewed, EPA will have it peer-
67 reviewed.
68 EPA also examines the toxicokinetics of the
69 agent to identify other chemicals (for example,
70 major metabolites of the agent) to include in the
71 assessment if adequate information is available,
72 in order to more fully explain the toxicity of the
73 agent and to suggest dose metrics for subsequent
74 modeling.
75 In assessments of chemical mixtures,
76 mixture studies are preferred for their ability to
77 reflect interactions among components. The
78 literature search seeks, in decreasing order of
79 preference (U.S. EPA. 2000b. 1986b):
80 - Studies of the mixture being assessed.
81 - Studies of a sufficiently similar mixture. In
82 evaluating similarity, the assessment
83 considers the alteration of mixtures in the
84 environment through partitioning and
85 transformation.
86 - Studies of individual chemical components of
87 the mixture, if there are not adequate studies
88 of sufficiently similar mixtures.
89 3.2. Selecting pertinent epidemiologic
90 studies
91 Study design is the key consideration for
92 selecting pertinent epidemiologic studies from
93 the results of the literature search.
94 - Cohort studies and case-control studies
95 provide the strongest epidemiologic
This document is a draft for review purposes only and does not constitute Agency policy.
xiv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 evidence, as they collect information about
2 individual exposures and effects.
3 - Ecologic studies (geographic correlation
4 studies) relate exposures and effects by
5 geographic area. They can provide strong
6 evidence if there are large exposure
7 contrasts between geographic areas,
8 relatively little exposure variation within
9 study areas, and population migration is
10 limited.
11 - Case reports of high or accidental exposure
12 lack definition of the population at risk and
13 the expected number of cases. They can
14 provide information about a rare effect or
15 about the relevance of analogous results in
16 animals.
17 The assessment briefly reviews ecologic
18 studies and case reports but reports details only
19 if they suggest effects not identified by other
20 epidemiologic studies.
21 3.2. Selecting pertinent experimental
22 studies
23 Exposure route is a key design consideration
24 for selecting pertinent experimental studies from
25 the results of the literature search.
26 - Studies of oral, inhalation, or dermal
27 exposure involve passage through an
28 absorption barrier and are considered most
29 pertinent to human environmental exposure.
30 - Injection or implantation studies are often
31 considered less pertinent but may provide
32 valuable toxicokinetic or mechanistic
33 information. They also may be useful for
34 identifying effects in animals if deposition or
35 absorption is problematic (for example, for
36 particles and fibers).
37 Exposure duration is also a key design
38 consideration for selecting pertinent
39 experimental studies.
40 - Studies of effects from chronic exposure are
41 most pertinent to lifetime human exposure.
42 - Studies of effects from less-than-chronic
43 exposure are pertinent but less preferred
44 than studies of chronic exposure.
45 Short-duration studies involving animals or
46 humans may provide toxicokinetic or
47 mechanistic information. Research involving
48 human subjects is considered only if conducted
49 according to ethical principles.
50 For developmental toxicity and reproductive
51 toxicity, irreversible effects may result from a
52 brief exposure during a critical period of
53 development. Accordingly, specialized study
54 designs are used for these effects (U.S. EPA.
55 2006b. 1998.1996.1991).
56 4. Evaluating the quality of
57 individual studies
58 4.1. Evaluating the quality of
59 epidemiologic studies
60 The assessment evaluates design and
61 methodologic aspects that can increase or
62 decrease the weight given to each epidemiologic
63 study in the overall evaluation (U.S. EPA. 2005a.
64 1998,1996,1994,1991):
65 - Documentation of study design, methods,
66 population characteristics, and results.
67 -
68
69 -
70
71 -
72
73 -
74
75
76 -
77
78 -
79
80 -
81
82 -
83
84
85
86
87
Definition and selection of the study and
comparison populations.
Ascertainment of exposure and the potential
for misclassification.
Ascertainment of disease or effect and the
potential for misclassification.
Duration of exposure and follow-up and
adequacy for assessing the occurrence of
effects, including latent effects.
Characterization of exposure during critical
periods.
Sample size and statistical power to detect
anticipated effects.
Participation rates and
potential for selection bias.
the resulting
Potential confounding and other sources of
bias are identified and addressed in the
study design or in the analysis of results. The
basis for consideration of confounding is a
reasonable expectation that the confounder
is prevalent in the population and is related
to both exposure and outcome.
89 For developmental toxicity, reproductive
90 toxicity, neurotoxicity, and cancer there is further
91 guidance on the nuances of evaluating
This document is a draft for review purposes only and does not constitute Agency policy.
xv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 epidemiologic studies of these effects [U.S. EPA.
2 2005a. 1998.1996.1991).
3 4.2. Evaluating the quality of
4 experimental studies
5 The assessment evaluates design and
6 methodologic aspects that can increase or
7 decrease the weight given to each experimental
8 study in the overall evaluation [U.S. EPA. 2005a.
9 1998.1996.1994.1991):
10 - Documentation of study design, animals or
11 study population, methods, basic data, and
12 results.
13 - Relevance to humans of the animal model
14 and experimental methods.
15 - Characterization of the nature and extent of
16 impurities and contaminants of the
17 administered chemical or mixture.
18 - Characterization of dose and dosing regimen
19 (including age at exposure) and their
20 adequacy to elicit adverse effects, including
21 latent effects.
22 - Sample sizes and statistical power to detect
23 dose-related differences or trends.
24 - Ascertainment of survival, vital signs, disease
25 or effects, and cause of death.
26 - Control of other variables that
27 influence the occurrence of effects.
could
28 The assessment uses statistical tests to
29 evaluate whether the observations may be due to
30 chance. The standard for determining statistical
31 significance of a response is a trend test or
32 comparison of outcomes in the exposed groups
33 against those of concurrent controls. In some
34 situations, examination of historical control data
35 from the same laboratory within a few years of
36 the study may improve the analysis. For an
37 uncommon effect that is not statistically
38 significant compared with concurrent controls,
39 historical controls may show that the effect is
40 unlikely to be due to chance. For a response that
41 appears significant against a concurrent control
42 response that is unusual, historical controls may
43 offer a different interpretation [U.S. EPA. 2005a).
44 For developmental toxicity, reproductive
45 toxicity, neurotoxicity, and cancer there is further
46 guidance on the nuances of evaluating
47 experimental studies of these effects [U.S. EPA.
48 2005a. 1998. 1996. 1991). In multi-generation
49 studies, agents that produce developmental
50 effects at doses that are not toxic to the maternal
51 animal are of special concern. Effects that occur
52 at doses associated with mild maternal toxicity
53 are not assumed to result only from maternal
54 toxicity. Moreover, maternal effects may be
55 reversible, while effects on the offspring may be
56 permanent [U.S. EPA. 1998.1991).
57 4.3. Reporting study results
58 The assessment uses evidence tables to
59 summarize details of the design and key results
60 of pertinent studies. There may be separate
61 tables for each site of toxicity or type of study.
62 If a large number of studies observe the same
63 effect, the assessment considers the study
64 characteristics in this section to identify the
65 strongest studies or types of study. The tables
66 report details from these studies, and the
67 assessment explains the reasons for not
68 reporting details of other studies or groups of
69 studies that do not add new information.
70 Supplemental material provides references to all
71 studies considered, including those not
72 summarized in the tables.
73 The assessment discusses strengths and
74 limitations that affect the interpretation of each
75 study. If the interpretation of a study in the
76 assessment differs from that of the study authors,
77 the assessment discusses the basis for the
78 difference.
79 As a check on the selection and evaluation of
80 pertinent studies, EPA asks peer reviewers to
81 identify studies that were not adequately
82 considered.
83 5. Weighing the overall evidence of
84 each effect
85 5.1. Weighing epidemiologic evidence
86 For each effect, the assessment evaluates the
87 evidence from the epidemiologic studies as a
88 whole to determine the extent to which any
89 observed associations may be causal. Positive,
90 negative, and null results are given weight
91 according to study quality. This evaluation
92 considers aspects of an association that suggest
93 causality, discussed by Hill [1965) and
94 elaborated by Rothman and Greenland [1998)
This document is a draft for review purposes only and does not constitute Agency policy.
xvi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 fU.S. EPA. 2005a: CDC. 2004: U.S. EPA. 2002.
2 19941
3 Strength of association: The finding of a large
4 relative risk with narrow confidence
5 intervals strongly suggests that an
6 association is not due to chance, bias, or
7 other factors. Modest relative risks, however,
8 may reflect a small range of exposures, an
9 agent of low potency, an increase in an effect
10 that is common, exposure misclassification,
11 or other sources of bias.
12 Consistency of association: An inference of
13 causality is strengthened if elevated risks are
14 observed in independent studies of different
15 populations and exposure scenarios.
16 Reproducibility of findings constitutes one of
17 the strongest arguments for causality.
18 Discordant results sometimes reflect
19 differences in study design, exposure, or
20 confounding factors.
21 Specificity of association: As originally
22 intended, this refers to one cause associated
23 with one effect. Current understanding that
24 many agents cause multiple effects and many
25 effects have multiple causes make this a less
26 informative aspect of causality, unless the
27 effect is rare or unlikely to have multiple
28 causes.
29 Temporal relationship: A causal interpretation
30 requires that exposure precede development
31 of the effect.
32 Biologic gradient (exposure-response
33 relation-ship): Exposure-response
34 relationships strongly suggest causality. A
35 monotonic increase is not the only pattern
36 consistent with causality. The presence of an
37 exposure-response gradient also weighs
38 against bias and confounding as the source of
39 an association.
40 Biologic plausibility: An inference of causality is
41 strengthened by data demonstrating
42 plausible biologic mechanisms, if available.
43 Coherence: An inference of causality is
44 strengthened by supportive results from
45 animal experiments, toxicokinetic studies,
46 and short-term tests. Coherence may also be
47 found in other lines of evidence, such as
48 changing disease patterns in the population.
49 "Natural experiments": A change in exposure
50 that brings about a change in disease
51 frequency provides strong evidence of
52 causality, for example, an intervention to
53 reduce exposure in the workplace or
54 environment that is followed by a reduction
55 of an adverse effect.
56 Analogy: Information on structural analogues or
57 on chemicals that induce similar mechanistic
58 events can provide insight into causality.
59 These considerations are consistent with
60 guidelines for systematic reviews that evaluate
61 the quality and weight of evidence. Confidence is
62 increased if the magnitude of effect is large, if
63 there is evidence of an exposure-response
64 relationship, or if an association was observed
65 and the plausible biases would tend to decrease
66 the magnitude of the reported effect. Confidence
67 is decreased for study limitations, inconsistency
68 of results, indirectness of evidence, imprecision,
69 or reporting bias [Guyatt et al., 2008a: Guyatt et
70 al..2008b).
71 To make clear how much the epidemiologic
72 evidence contributes to the overall weight of the
73 evidence, the assessment may choose a
74 descriptor such as sufficient evidence, suggestive
75 evidence, inadequate evidence, or evidence
76 suggestive of no causal relationship to
77 characterize the epidemiologic evidence of each
78 effect [CDC. 2004].
79 5.2. Weighing experimental animal
80 evidence
81 For each effect, the assessment evaluates the
82 evidence from the animal experiments as a whole
83 to determine the extent to which they indicate a
84 potential for effects in humans. Consistent results
85 across various species and strains increase
86 confidence that similar results would occur in
87 humans. Several concepts discussed by Hill
88 [1965] are pertinent to the weight of
89 experimental results: consistency of response,
90 dose-response relationships, strength of
91 response, biologic plausibility, and coherence
92 [U.S. EPA. 2005a. 2002.1994].
93 In weighing evidence from multiple
94 experiments, [U.S. EPA. 2005a] distinguishes
95 Conflicting evidence [that is, mixed positive and
96 negative results in the same sex and strain
97 using a similar study protocol] from
This document is a draft for review purposes only and does not constitute Agency policy.
xvii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Differing results (that is, positive results and
2 negative results are in different sexes or
3 strains or use different study protocols).
4 Negative or null results do not invalidate positive
5 results in a different experimental system. EPA
6 regards all as valid observations and looks to
7 methodological differences or, if available,
8 mechanistic information to reconcile differing
9 results.
10 It is well established that there are critical
11 periods for some developmental and
12 reproductive effects. Accordingly, the assessment
13 determines whether critical periods have been
14 adequately investigated [U.S. EPA. 2006b. 2005a.
15 b, 1998. 1996. 1991]. Similarly, the assessment
16 determines whether the database is adequate to
17 evaluate other critical sites and effects.
18 In evaluating evidence of genotoxicity:
19 - Demonstration of gene mutations,
20 chromosome aberrations, or aneuploidy in
21 humans or experimental mammals (in vivo)
22 provides the strongest evidence.
23 - This is followed by positive results in lower
24 organisms or in cultured cells (in vitro) or for
25 other genetic events.
26 - Negative results carry less weight, partly
27 because they cannot exclude the possibility
28 of effects in other tissues (IARC. 2 006).
29 For germ-cell mutagenicity, EPA has defined
30 categories of evidence, ranging from positive
31 results of human germ-cell mutagenicity to
32 negative results for all effects of concern [U.S.
33 EPA. 1986a).
34 5.3. Characterizing modes of action
35 For each effect, the assessment discusses the
36 available information on its modes of action and
37 associated key events (key events being
38 empirically observable, necessary precursor
39 steps or biologic markers of such steps; mode of
40 action being a series of key events involving
41 interaction with cells, operational and anatomic
42 changes, and resulting in disease). Pertinent
43 information may also come from studies of
44 metabolites or of compounds that are
45 structurally similar or that act through similar
46 mechanisms. Information on mode of action is
47 not required for a conclusion that an effect is
48 causally related to an agent [U.S. EPA. 2005a).
49 The assessment addresses several questions
50 about each hypothesized mode of action [U.S.
51 EPA.2005a).
52 (1) Is the hypothesized mode of action
53 sufficiently supported in test animals?
54 Strong support for a key event being
55 necessary to a mode of action can come from
56 experimental challenge to the hypothesized
57 mode of action, in which studies that
58 suppress a key event observe suppression of
59 the effect. Support for a mode of action is
60 meaningfully strengthened by consistent
61 results in different experimental models,
62 much more so than by replicate experiments
63 in the same model. The assessment may
64 consider various aspects of causality in
65 addressing this question.
66 (2) Is the hypothesized mode of action
67 relevant to humans? The assessment
68 reviews the key events to identify critical
69 similarities and differences between the test
70 animals and humans. Site concordance is not
71 assumed between animals and humans,
72 though it may hold for certain effects or
73 modes of action. Information suggesting
74 quantitative differences in doses where
75 effects would occur in animals or humans is
76 considered in the dose-response analysis but
77 is not used to determine relevance. Similarly,
78 anticipated levels of human exposure are not
79 used to determine relevance.
80 (3) Which populations or lifestages can be
81 particularly susceptible to the
82 hypothesized mode of action? The
83 assessment reviews the key events to
84 identify populations and lifestages that might
85 be susceptible to their occurrence.
86 Quantitative differences may result in
87 separate toxicity values for susceptible
88 populations or lifestages.
89 The assessment discusses the likelihood that
90 an agent operates through multiple modes of
91 action. An uneven level of support for different
92 modes of action can reflect disproportionate
93 resources spent investigating them [U.S. EPA.
94 2005a). It should be noted that in clinical
95 reviews, the credibility of a series of studies is
96 reduced if evidence is limited to studies funded
97 by one interested sector [Guyatt et al.. 2008a).
98 For cancer, the assessment evaluates
99 evidence of a mutagenic mode of action to guide
This document is a draft for review purposes only and does not constitute Agency policy.
xviii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 extrapolation to lower doses and consideration
2 of susceptible lifestages. Key data include the
3 ability of the agent or a metabolite to react with
4 or bind to DNA, positive results in multiple test
5 systems, or similar properties and structure-
6 activity relationships to mutagenic carcinogens
7 [U.S. EPA. 2005a).
9 5.4. Characterizing the overall weight
10 of the evidence
11 After weighing the epidemiologic and
12 experimental studies pertinent to each effect, the
13 assessment may select a standard descriptor to
14 characterize the overall weight of the evidence.
15 For example, the following standard descriptors
16 combine epidemiologic, experimental, and
17 mechanistic evidence of carcinogenicity [U.S.
18 EPA. 20053).
19 Carcinogenic to humans: There is convincing
20 epidemiologic evidence of a causal
21 association [that is, there is reasonable
22 confidence that the association cannot be
23 fully explained by chance, bias, or
24 confounding); or there is strong human
25 evidence of cancer or its precursors,
26 extensive animal evidence, identification of
27 key precursor events in animals, and strong
28 evidence that they are anticipated to occur in
29 humans.
30 Likely to be carcinogenic to humans: The
31 evidence demonstrates a potential hazard to
32 humans but does not meet the criteria for
33 carcinogenic. There may be a plausible
34 association in humans, multiple positive
35 results in animals, or a combination of
36 human, animal, or other experimental
37 evidence.
38 Suggestive evidence of carcinogenic potential:
39 The evidence raises concern for effects in
40 humans but is not sufficient for a stronger
41 conclusion. This descriptor covers a range of
42 evidence, from a positive result in the only
43 available study to a single positive result in
44 an extensive database that includes negative
45 results in other species.
46 Inadequate information to assess carcinogenic
47 potential: No other descriptors apply.
48 Conflicting evidence can be classified as
49 inadequate information if all positive results
50 are opposed by negative studies of equal
51 quality in the same sex and strain. Differing
52 results, however, can be classified as
53 suggestive evidence or as likely to be
54 carcinogenic.
55 Not likely to be carcinogenic to humans: There
56 is robust evidence for concluding that there
57 is no basis for concern. There may be no
58 effects in both sexes of at least two
59 appropriate animal species; positive animal
60 results and strong, consistent evidence that
61 each mode of action in animals does not
62 operate in humans; or convincing evidence
63 that effects are not likely by a particular
64 exposure route or below a defined dose.
65 Multiple descriptors may be used if there is
66 evidence that carcinogenic effects differ by dose
67 range or exposure route [U.S. EPA. 2005a).
68 EPA is investigating and may on a trial basis
69 propose standard descriptors to characterize the
70 overall weight of the evidence for effects other
71 than cancer.
72 6. Selecting studies for derivation
73 of toxicity values
74 For each effect where there is credible
75 evidence of an association with the agent, the
76 assessment derives toxicity values if there are
77 suitable epidemiologic or experimental data. The
78 decision to derive toxicity values may be linked
79 to the weight-of-evidence descriptor. For
80 example, EPA typically derives toxicity values for
81 agents classified as carcinogenic to humans or as
82 likely to be carcinogenic [U.S. EPA. 2005a).
83 Dose-response analysis requires quantitative
84 measures of dose and response. Then, other
85 factors being equal [U.S. EPA. 2005a. 1994):
86 - Epidemiologic studies are preferred over
87 animal studies, if quantitative measures of
88 exposure are available and effects can be
89 attributed to the agent.
90 - Among experimental animal models, those
91 that respond most like humans are
92 preferred, if the comparability of response
93 can be determined.
94 - Studies by a route of human environmental
95 exposure are preferred, although a validated
96 toxicokinetic model can be used to
97 extrapolate across exposure routes.
This document is a draft for review purposes only and does not constitute Agency policy.
xix DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 - Studies of longer exposure duration and
2 follow-up are preferred, to minimize
3 uncertainty about whether effects are
4 representative of lifetime exposure.
5 - Studies with multiple exposure levels are
6 preferred for their ability to provide
7 information about the shape of the exposure-
8 response curve.
9 - Studies with adequate power to detect
10 effects at lower exposure levels are
11 preferred, to minimize the extent of
12 extrapolation to levels found in the
13 environment.
14 Studies with non-monotonic exposure-
15 response relationships are not necessarily
16 excluded from the analysis. A diminished effect at
17 higher exposure levels may be satisfactorily
18 explained by factors such as competing toxicity,
19 saturation of absorption or metabolism, exposure
20 misclassification, or selection bias.
21 If a large number of studies are suitable for
22 dose-response analysis, the assessment
23 considers the study characteristics in this section
24 to focus on the most informative data. The
25 assessment explains the reasons for not
26 analyzing other groups of studies. As a check on
27 the selection of studies for dose-response
28 analysis, EPA asks peer reviewers to identify
29 studies that were not adequately considered.
30 7. Deriving toxicity values
31 7.1. General framework for dose-
32 response analysis
33 EPA uses a two-step approach that
34 distinguishes analysis of the observed dose-
35 response data from inferences about lower doses
36 fU.S. EPA.2005a1.
37 Within the observed range, the preferred
38 approach is to use modeling to incorporate a
39 wide range of data into the analysis. The
40 modeling yields a point of departure (an exposure
41 level near the lower end of the observed range,
42 without significant extrapolation to lower doses)
43 (sections 7.2-7.3).
44 Extrapolation to lower doses considers what
45 is known about the modes of action for each
46 effect (sections 7.4-7.5). When response
47 estimates at lower doses are not required, an
48 alternative is to derive reference values, which
49 are calculated by applying factors that account
50 for sources of uncertainty and variability to the
51 point of departure (section 7.6).
52 For a group of agents that induce an effect
53 through a common mode of action, the dose-
54 response analysis may derive a relative potency
55 factor for each agent. A full dose-response
56 analysis is conducted for one well-studied index
57 chemical in the group, then the potencies of other
58 members are expressed in relative terms based
59 on relative toxic effects, relative absorption or
60 metabolic rates, quantitative structure-activity
61 relationships, or receptor binding characteristics
62 (U.S. EPA.2005a. 2000b).
63 Increasingly, EPA is basing toxicity values on
64 combined analyses of multiple data sets or
65 multiple responses. EPA also considers multiple
66 dose-response approaches when they can be
67 supported by robust data.
68 7.2. Modeling dose
69 The preferred approach for analysis of dose
70 is toxicokinetic modeling because of its ability to
71 incorporate a wide range of data. The preferred
72 dose metric would refer to the active agent at the
73 site of its biologic effect or to a close, reliable
74 surrogate measure. The active agent may be the
75 administered chemical or a metabolite.
76 Confidence in the use of a toxicokinetic model
77 depends on the robustness of its validation
78 process and on the results of sensitivity analyses
79 fU.S. EPA. 2006a. 2005a. 19941
80 Because toxicokinetic modeling can require
81 many parameters and more data than are
82 typically available, EPA has developed standard
83 approaches that can be applied to typical data
84 sets. These standard approaches also facilitate
85 comparison across exposure patterns and
86 species.
87 - Intermittent study exposures are
88 standardized to a daily average over the
89 duration of exposure. For chronic effects,
90 daily exposures are averaged over the
91 lifespan. Exposures during a critical period,
92 however, are not averaged over a longer
93 duration (U.S. EPA. 2005a. 1998. 1996.
94 19911.
95 - Doses are standardized to equivalent human
96 terms to facilitate comparison of results from
97 different species.
This document is a draft for review purposes only and does not constitute Agency policy.
xx DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 - Oral doses are scaled allometrically
2 using mg/kg3/4-d as the equivalent dose
3 metric across species. Allometric scaling
4 pertains to equivalence across species,
5 not across lifestages, and is not used to
6 scale doses from adult humans or
7 mature animals to infants or children
8 fU.S. EPA.2011b.2005al
9 - Inhalation exposures are scaled using
10 dosimetry models that apply species-
11 specific physiologic and anatomic factors
12 and consider whether the effect occurs
13 at the site of first contact or after
14 systemic circulation [U.S. EPA. 1994].
15 It can be informative to convert doses across
16 exposure routes. If this is done, the assessment
17 describes the underlying data, algorithms, and
18 assumptions (U.S. EPA. 2005a).
19 In the absence of study-specific data on, for
20 example, intake rates or body weight, EPA has
21 developed recommended values for use in dose-
22 response analysis [U.S. EPA. 19881.
23 7.3. Modeling response in the range
24 of observation
25 Toxicodynamic ("biologically based")
26 modeling can incorporate data on biologic
27 processes leading to an effect. Such models
28 require sufficient data to ascertain a mode of
29 action and to quantitatively support model
30 parameters associated with its key events.
31 Because different models may provide equivalent
32 fits to the observed data but diverge substantially
33 at lower doses, critical biologic parameters
34 should be measured from laboratory studies, not
35 by model fitting. Confidence in the use of a
36 toxicodynamic model depends on the robustness
37 of its validation process and on the results of
38 sensitivity analyses. Peer review of the scientific
39 basis and performance of a model is essential
40 fU.S. EPA. 20053).
41 Because toxicodynamic modeling can require
42 many parameters and more knowledge and data
43 than are typically available, EPA has developed a
44 standard set of empirical ("curve-fitting") models
45 (http://www.epa.gov/ncea/bmds/) that can be
46 applied to typical data sets, including those that
47 are nonlinear. EPA has also developed guidance
48 on modeling dose-response data, assessing
49 model fit, selecting suitable models, and
50 reporting modeling results (U.S. EPA. 2000a).
51 Additional judgment or alternative analyses are
52 used when the procedure fails to yield reliable
53 results, for example, if the fit is poor, modeling
54 may be restricted to the lower doses, especially if
55 there is competing toxicity at higher doses (U.S.
56 EPA.2005a).
57 Modeling is used to derive a point of
58 departure (U.S. EPA. 2005a. 2000a). (See section
59 7.6 for alternatives if a point of departure cannot
60 be derived by modeling.)
61 - For dichotomous responses, the point of
62 departure is often the 95% lower bound on
63 the dose associated with a 10% response,
64 but a lower response that falls within the
65 observed range may be used instead. For
66 example, reproductive or developmental
67 studies often have power to detect a 5%
68 response; epidemiologic studies, 1% or
69 lower.
70 - For continuous responses, the point of
71 departure is ideally the dose where the effect
72 becomes biologically significant. In the
73 absence of such definition, both statistical
74 and biologic factors are considered.
75 7.4. Extrapolating to lower doses
76 The purpose of extrapolating to lower doses
77 is to estimate responses at exposures below the
78 observed data. Low-dose extrapolation is
79 typically used for known and likely carcinogens.
80 Low-dose extrapolation considers what is known
81 about modes of action (U.S. EPA. 2005a).
82 (1) If a biologically based model has been
83 developed and validated for the agent,
84 extrapolation may use the fitted model below
85 the observed range if significant model
86 uncertainty can be ruled out with reasonable
87 confidence.
88 (2) Linear extrapolation is used if the dose-
89 response curve is expected to have a linear
90 component below the point of departure.
91 This includes:
92
93
94
95
96
97
98
Agents or their metabolites that are
DNA-reactive and have direct mutagenic
activity.
Agents or their metabolites for which
human exposures or body burdens are
near doses associated with key events
leading to an effect.
This document is a draft for review purposes only and does not constitute Agency policy.
xxi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Linear extrapolation is also used if the
2 evidence is insufficient to establish a mode of
3 action.
4 The result of linear extrapolation is
5 described by an oral slope factor or an
6 inhalation unit risk, which is the slope of the
7 dose-response curve at lower doses or
8 concentrations, respectively.
9 (3) Nonlinear extrapolation is used if there are
sufficient data to ascertain the mode of
action and to conclude that it is not linear at
lower doses, and the agent does not
demonstrate mutagenic or other activity
consistent with linearity at lower doses. If
nonlinear extrapolation is appropriate but no
model is developed, an alternative is to
calculate reference values.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
If linear extrapolation is used, the
assessment develops a candidate slope factor or
unit risk for each suitable data set. These results
are arrayed, using common dose metrics, to show
the distribution of relative potency across
various effects and experimental systems. The
assessment then derives an overall slope factor
and an overall unit risk for the agent, considering
the various dose-response analyses, the study
preferences discussed in section 6, and the
possibility of basing a more robust result on
multiple data sets.
30 7.5. Considering susceptible
31 populations and lifestages
32 The assessment analyzes the available
33 information on populations and lifestages that
34 may be particularly susceptible to each effect. A
35 tiered approach is used [U.S. EPA. 2005aj.
36 (1] If an epidemiologic or experimental study
37 reports quantitative results for a susceptible
38 population or lifestage, these data are
39 analyzed to derive separate toxicity values
40 for susceptible individuals.
41 (2) If data on risk-related parameters allow
42 comparison of the general population and
43 susceptible individuals, these data are used
44 to adjust the general-population toxicity
45 values for application to susceptible
46 individuals.
47 (3) In the absence of chemical-specific data, EPA
48 has developed age-dependent adjustment
49 factors for early-life exposure to suspected
50 carcinogens that have a mutagenic mode of
51 action. There is evidence of early-life
52 susceptibility to various carcinogenic agents,
53 but most epidemiologic studies and cancer
54 bioassays do not include early-life exposure.
55 To address the potential for early-life
56 susceptibility, EPA recommends [U.S. EPA.
57 20Q5b):
58 - 10-fold adjustment for exposures before
59 age 2 years.
60 - 3-fold adjustment for exposures
61 between ages 2 and 16 years.
62 7.6. Reference values and uncertainty
63 factors
64 An oral reference dose or an inhalation
65 reference concentration is an estimate of an
66 exposure (including in susceptible subgroups)
67 that is likely to be without an appreciable risk of
68 adverse health effects over a lifetime [U.S. EPA.
69 2002]. Reference values are typically calculated
70 for effects other than cancer and for suspected
71 carcinogens if a well characterized mode of
72 action indicates that a necessary key event does
73 not occur below a specific dose. Reference values
74 provide no information about risks at higher
75 exposure levels.
76 The assessment characterizes effects that
77 form the basis for reference values as adverse,
78 considered to be adverse, or a precursor to an
79 adverse effect. For developmental toxicity,
80 reproductive toxicity, and neurotoxicity there is
81 guidance on adverse effects and their biologic
82 markers [U.S. EPA. 1998.1996.1991].
83 To account for uncertainty and variability in
84 the derivation of a lifetime human exposure
85 where effects are not anticipated to occur,
86 reference values are calculated by applying a
87 series of uncertainty factors to the point of
88 departure. If a point of departure cannot be
89 derived by modeling, a no-observed-adverse-
90 effect level or a lowest-observed-adverse-effect
91 level is used instead. The assessment discusses
92 scientific considerations involving several areas
93 of variability or uncertainty.
94 Human variation. A factor of 10 is applied to
95 account for variation in susceptibility across
96 the human population and the possibility
97 that the available data may not be
98 representative of individuals who are most
This document is a draft for review purposes only and does not constitute Agency policy.
xxii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 susceptible to the effect. This factor is
2 reduced only if the point of departure is
3 derived specifically for susceptible
4 individuals (not for a general population that
5 includes both susceptible and non-
6 susceptible individuals) [U.S. EPA. 2002.
7 1998.1996.1994.19911
8 Animal-to-human extrapolation. A factor of 10
9 is applied if animal results are used to make
10 inferences about humans. This factor is often
11 regarded as comprising toxicokinetics and
12 toxicodynamics in equal parts. Accordingly, if
13 the point of departure is based on
14 toxicokinetic modeling, dosimetry modeling,
15 or allometric scaling across species, a factor
16 of 10V2 (rounded to 3) is applied to account
17 for the remaining uncertainty involving
18 toxicodynamic differences. An animal-to-
19 human factor is not applied if a biologically
20 based model adjusts fully for toxicokinetic
21 and toxicodynamic differences across species
22 (U.S. EPA, 2011b, 2002, 1998, 1996, 1994,
23 19911
24 Adverse-effect level to no-observed-adverse-
25 effect level. If a point of departure is based
26 on a lowest-observed-adverse-effect level,
27 the assessment must infer a dose where such
28 effects are not expected. This can be a matter
29 of great uncertainty, especially if there is no
30 evidence available at lower doses. A factor of
31 10 is applied to account for the uncertainty
32 in making this inference. A factor other than
33 10 may be used, depending on the magnitude
34 and nature of the response and the shape of
35 the dose-response curve (U.S. EPA. 2002.
36 1998.1996.1994.19911
37 Subchronic-to-chronic exposure. If a point of
38 departure is based on subchronic studies, the
39 assessment considers whether lifetime
40 exposure could have effects at lower levels of
41 exposure. A factor of 10 is applied to account
42 for the uncertainty in using subchronic
43 studies to make inferences about lifetime
44 exposure. This factor may also be applied for
45 developmental or reproductive effects if
46 exposure covered less than the full critical
47 period. A factor other than 10 may be used,
48 depending on the duration of the studies and
49 the nature of the response (U.S. EPA. 2002.
50 1998.1994].
51 Incomplete database. If an incomplete database
52 raises concern that further studies might
53 identify a more sensitive effect, organ
54 system, or lifestage, the assessment may
55 apply a database uncertainty factor (U.S.
56 EPA. 2002.1998.1996.1994.1991]. The size
57 of the factor depends on the nature of the
58 database deficiency. For example, EPA
59 typically follows the suggestion that a factor
60 of 10 be applied if both a prenatal toxicity
61 study and a two-generation reproduction
62 study are missing and a factor of 10V2 if
63 either is missing (U.S. EPA. 2002].
64 In this way, the assessment derives
65 candidate reference values for each suitable data
66 set and effect that is credibly associated with the
67 agent. These results are arrayed, using common
68 dose metrics, to show where effects occur across
69 a range of exposures (U.S. EPA. 1994]. The
70 assessment then selects an overall reference
71 dose and an overall reference concentration for
72 the agent to represent lifetime human exposure
73 levels where effects are not anticipated to occur.
74 The assessment may also report reference
75 values for each effect. This would facilitate
76 subsequent cumulative risk assessments that
77 consider the combined effect of multiple agents
78 acting at a common site or through common
79 mechanisms (U.S. EPA. 2002].
80 7.7. Confidence and uncertainty in the
81 reference values
82 The assessment selects a standard descriptor
83 to characterize the level of confidence in each
84 reference value, based on the likelihood that the
85 value would change with further testing.
86 Confidence in reference values is based on
87 quality of the studies used and completeness of
88 the database, with more weight given to the
89 latter. The level of confidence is increased for
90 reference values based on human data supported
91 by animal data (U.S. EPA. 1994].
92 High confidence: The reference value is not
93 likely to change with further testing, except
94 for mechanistic studies that might affect the
95 interpretation of prior test results.
96 Medium confidence: This is a matter of
97 judgment, between high and low confidence.
This document is a draft for review purposes only and does not constitute Agency policy.
xxiii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Low confidence: The reference value is
2 especially vulnerable to change with further
3 testing.
4 These criteria are consistent with guidelines
5 for systematic reviews that evaluate the quality
6 of evidence. These also focus on whether further
7 research would be likely to change confidence in
8 the estimate of effect (Guyatt etal., 2008a).
9 All assessments discuss the significant
10 uncertainties encountered in the analysis. EPA
11 provides guidance on characterization of
12 uncertainty [U.S. EPA. 2005a). For example, the
13 discussion distinguishes model uncertainty (lack
14 of knowledge about the most appropriate
15 experimental or analytic model) and parameter
16 uncertainty (lack of knowledge about the
17 parameters of a model). Assessments also
18 discuss human variation (interpersonal
19 differences in biologic susceptibility or in
20 exposures that modify the effects of the agent).
21
22 References
23 CDC (Centers for Disease Control and
24 Prevention). (2004). The health
25 consequences of smoking: A report of the
26 Surgeon General. Washington, DC: U.S.
27 Department of Health and Human Services.
28 http://www.surgeongeneral.gov/library/sm
29 okingconsequences/.
30 Guyatt. GH: Oxman. AD: Vist. GE: Kunz. R: Falck-
31 Ytter. Y: Alonso-Coello. P: Schiinemann. HI.
32 (2008a). GRADE: An emerging consensus on
33 rating quality of evidence and strength of
34 recommendations. BMJ 336: 924-926.
35 http://dx.doi.org/10.1136/bmi.39489.4703
36 47.AD.
37 Guyatt. GH: Oxman. AD: Kunz. R: Vist. GE: Falck-
38 Ytter. Y: Schiinemann. HI. (2008b). GRADE:
39 What is "quality of evidence" and why is it
40 important to clinicians? BMJ 336: 995-998.
41 http://dx.doi.org/10.1136/bmj.39490.5510
42 19.BE.
43 Hill. AB. (1965). The environment and disease:
44 Association or causation? Proc R Soc Med 58:
45 295-300.
46 IARC (International Agency for Research on
47 Cancer). (2006). Preamble to the IARC
48 monographs. Lyon, France.
49 http://monographs.iarc.fr/ENG/Preamble/.
50 Rothman. Kl: Greenland. S. (1998). Modern
51 epidemiology (2nd ed.). Philadelphia, PA:
52 Lippincott, Williams, & Wilkins.
53 U.S. EPA (U.S. Environmental Protection Agency).
54 (1986a). Guidelines for mutagenicity risk
55 assessment [EPA Report]. (EPA/630/R-
56 98/003). Washington, DC.
57 http://www.epa.gov/iris/backgrd.html.
58 U.S. EPA (U.S. Environmental Protection Agency).
59 (1986b). Guidelines for the health risk
60 assessment of chemical mixtures [EPA
61 Report] (pp. 34014-34025). (EPA/630/R-
62 98/002). Washington, DC.
63 http://cfpub.epa.gov/ncea/cfm/recordisplay
64 .cfm?deid=22567.
65 U.S. EPA (U.S. Environmental Protection Agency).
66 (1988). Recommendations for and
67 documentation of biological values for use in
68 risk assessment. (EPA/600/6-87/008).
69 Cincinnati, OH: U.S. Environmental
70 Protection Agency, Environmental Criteria
71 and Assessment Office.
72 http://cfpub.epa.gov/ncea/cfm/recordisplay
73 .cfm?deid=34855.
74 U.S. EPA (U.S. Environmental Protection Agency).
75 (1991). Guidelines for developmental
76 toxicity risk assessment [EPA Report].
77 (EPA/600/FR-91/001). Washington, DC: U.S.
78 Environmental Protection Agency, Risk
79 Assessment Forum.
80 http://www.epa.gov/iris/backgrd.html.
81 U.S. EPA (U.S. Environmental Protection Agency).
82 (1994). Methods for derivation of inhalation
83 reference concentrations and application of
84 inhalation dosimetry. (EPA/600/8-
85 90/066F). Research Triangle Park, NC: U.S.
86 Environmental Protection Agency, Office of
87 Research and Development, Office of Health
88 and Environmental Assessment,
89 Environmental Criteria and Assessment
90 Office.
91 http://cfpub.epa.gov/ncea/cfm/recordisplay
92 .cfm?deid=71993.
93 U.S. EPA (U.S. Environmental Protection Agency).
94 (1996). Guidelines for reproductive toxicity
95 risk assessment [EPA Report]. (EPA/630/R-
96 96/009). Washington, DC: U.S.
97 Environmental Protection Agency, Risk
98 Assessment Forum.
99 http://www.epa.gov/raf/publications/pdfs/
100 REPR051.PDF.
This document is a draft for review purposes only and does not constitute Agency policy.
xxiv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 U.S. EPA (U.S. Environmental Protection Agency).
2 (1998). Guidelines for neurotoxicity risk
3 assessment. (EPA/630/R-95/001F).
4 Washington, DC: U.S. Environmental
5 Protection Agency, Risk Assessment Forum.
6 http://www.epa.gov/raf/publications/pdfs/
7 NEUROTOX.PDF.
8 U.S. EPA (U.S. Environmental Protection Agency).
9 (2000a). Benchmark dose technical guidance
10 document [external review draft].
11 (EPA/630/R-00/001). Washington, DC: U.S.
12 Environmental Protection Agency, Risk
13 Assessment Forum.
14 http://www.epa.gov/raf/publications/benc
15 hmark-dose-doc-draft.htm.
16 U.S. EPA (U.S. Environmental Protection Agency).
17 (2000b). Supplementary guidance for
18 conducting health risk assessment of
19 chemical mixtures [EPA Report].
20 (EPA/630/R-00/002).
21 http://cfpub.epa.gov/ncea/cfm/recordisplay
22 .cfm?deid=20533.
23 U.S. EPA (U.S. Environmental Protection Agency).
24 (2002). A review of the reference dose and
25 reference concentration processes.
26 (EPA/630/P-02/002F). Washington, DC.
27 http://cfpub.epa.gov/ncea/cfm/recordisplay
28 .cfm?deid=51717.
29 U.S. EPA (U.S. Environmental Protection Agency).
30 (2005a). Guidelines for carcinogen risk
31 assessment. (EPA/630/P-03/001F).
32 Washington, DC.
33 http://www.epa.gov/cancerguidelines/.
34 U.S. EPA (U.S. Environmental Protection Agency).
35 (2005b). Supplemental guidance for
36 assessing susceptibility from early-life
37 exposure to carcinogens. (EPA/630/R-
38 03/003F). Washington, DC: U.S.
39 Environmental Protection Agency, Risk
40 Assessment Forum.
41 http://www.epa.gov/cancerguidelines/guid
42 elines-carcinogen-supplementhtm.
43 U.S. EPA (U.S. Environmental Protection Agency).
44 (2006a). Approaches for the application of
45 physiologically based pharmacokinetic
46 (PBPK) models and supporting data in risk
47 assessment (Final Report). (EPA/600/R-
48 05/043F). Washington, DC: U.S.
49 Environmental Protection Agency, Office of
50 Research and Development.
51 U.S. EPA (U.S. Environmental Protection Agency).
52 (2006b). A framework for assessing health
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
U
risk of environmental exposures to children.
(EPA/600/R-05/093F). Washington, DC.
http://cfpub.epa.gov/ncea/cfm/recordisplay
.cfm?deid=158363.
L (U.S. Environmental Protection Agency).
(2009). EPAs Integrated Risk Information
System: Assessment development process.
Washington, DC.
http://epa.gov/iris/process.htm.
,S. EPA (U.S. Environmental Protection Agency).
(2011a). Recommended use of body weight
3/4 as the default method in derivation of
the oral reference dose.
(EPA/100/R11/0001). Washington, DC.
http://www.epa.gov/raf/publications/inters
pecies-extrapolation.htm.
May 2012 version
This document is a draft for review purposes only and does not constitute Agency policy.
xxv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
EXECUTIVE SUMMARY
3 Occurrence and Health Effects
4
5 Trimethylbenzenes are a commercially available mixture of three individual
6 isomers: 1,2,3-, 1,2,4-, and 1,3,5-trimethylbenzene (TMBs). 1MB isomers are
7 produced during petroleum refining and production of aromatic hydrocarbons with
8 nine carbons (i.e., C9 aromatic fraction). As the vast majority of the C9 fraction is
9 used as a component of gasoline, vehicle emissions are expected to be the major
10 anthropogenic source of TMBs. TMBs are volatile hydrocarbons, and thus humans
11 are exposed to these isomers primarily through breathing air containing TMB
12 vapors, although ingestion through food or drinking water is also possible.
13 Effects on the nervous system, respiratory system, and hematological
14 system (i.e., blood) have been reported in occupationally- and residentially-exposed
15 humans, but these effects were observed following exposure to complex mixtures
16 containing TMB isomers, thus making it difficult to determine the contribution of
17 each TMB isomer to the observed health effects. Health effects that are roughly
18 analogous to those seen in humans have been observed in animals exposed to the
19 individual isomers. Effects on the nervous system, including cognitive effects and
20 decreased pain sensitivity, are the most widely observed effects in animals. Effects
21 on other organ systems, including the respiratory and hematological systems, have
22 also been observed in animals. Both 1,2,4-TMB and 1,3,5-TMB have been observed
23 to elicit effects on pregnant animals and developing fetuses, but at exposure levels
24 greater than those that cause effects on the nervous system. There is inadequate
25 information to evaluate the carcinogenicity of TMBs.
26 Effects Other Than Cancer Following Inhalation Exposure
27 The relationship between exposure to 1,2,3-TMB, 1,2,4-TMB, 1,3,5-TMB and health effects
28 has been evaluated in studies of (1) exposed human adults, (2) animals exposed via inhalation for
29 acute, short-term, and subchronic durations, and (3) animals exposed gestationally via inhalation.
30 Human studies included occupational exposure to various solvent mixtures containing
31 TMBs. Health effects noted in these studies were eye irritation, neurological (hand tremble,
32 abnormal fatigue, lack of coordination), and hematological effects (Chen etal.. 1999: Norseth etal..
33 1991: Baettigetal.. 1958: Battigetal.. 1956). Also, residential exposure to mixtures containing
34 1,2,4-TMB were observed to result in asthma (Billionnetetal., 2011). However, as these studies
35 involved exposures to mixtures containing multiple TMB isomers and other volatile organic
This document is a draft for review purposes only and does not constitute Agency policy.
xxvi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 compounds (VOCs), it is difficult to ascertain the specific contribution of each 1MB isomer to the
2 specific health effects reported. Controlled human exposures to individual isomers also exist,
3 although these studies generally report little or no effect on respiratory or sensory irritation [Tones
4 etal.. 2006: Jarnbergetal.. 1997a: Tarnbergetal.. 1997b: Kostrzewskietal.. 1997: Tarnbergetal..
5 1996: Kostrewski and Wiaderna-Brycht, 1995]. One controlled human exposure study reported
6 some deficits in attention following exposure to white spirit (WS), a complex mixture containing
7 1,2,4-TMB [Lammers etal.. 2007).
8 Animal inhalation studies [Wiaderna et al., 1998][Wiadernaetal., 2002: Gralewicz and
9 Wiaderna. 2001: Gralewicz etal.. 1997a: Gralewicz etal.. 1997b: Korsak et al.. 1995] included acute
10 and short-term studies of TMBs that reported respiratory irritation (decreased respiration rates]
11 and neurological (decreased pain sensitivity, altered cognitive function, and decreased anxiety
12 and/or increased motor function] effects that support effects seen in human studies. Four
13 subchronic inhalation studies for 1,2,3-TMB and 1,2,4-TMB observed exposure-response effects in
14 multiple organ systems, including the nervous, hematological, and respiratory systems (Korsak et
15 al., 2000a, b: Korsak et al., 1997: Korsak and Rydzynski, 1996]. In these studies, disturbances in
16 central nervous system (CNS] function, including decreased pain sensitivity and decreased
17 neuromuscular function and coordination, appear to be the most sensitive endpoints following
18 exposure to 1,2,3-TMB or 1,2,4-TMB. No subchronic studies were found that investigated exposure
19 to 1,3,5-TMB. One developmental toxicity study (Saillenfaitetal., 2005] observed similar levels of
20 maternal and fetal toxicity (i.e., decreased maternal weight gain and fetal weight] following
21 exposure to either 1,2,4-TMB or 1,3,5-TMB; other indices of fetal toxicity (i.e., fetal death and
22 malformations] were not affected by exposure.
23 Table ES-1 summarizes the RfCs derived for all three TMB isomers, and the sections that
24 follow provide details on the RfC derivation for each isomer.
25 Table ES-1. Summary of inhalation reference concentrations (RfCs)
Isomer
Source
Reference value
Confidence
Inhalation reference concentration (mg/m3)
1,2,4-TMB
1,2,3-TMB
1,3,5-TMB
Decreased pain sensitivity
Decreased pain sensitivity
Adopted from 1,2,4-TMB based on
sufficient similarity of these isomers
2 x 10"2
2 x 10"2
2 x 10"2
Low-to-medium
Low-to-medium
Low
This document is a draft for review purposes only and does not constitute Agency policy.
xxvii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Inhalation Reference Concentration (RfC) for 1,2,4-TMB for Effects Other Than
Cancer
Table ES-2. Summary of reference concentration (RfC) derivation for
1,2,4-TMB
Critical effect
Decreased pain sensitivity
90 d male rat study
Korsak and Rydzyriski (1996)
Point of departure
PODHEC(mg/m3) = 15.8
Uncertainty
factor
1,000
Chronic RfC
(mg/m3)
2 x 10"2
Decreased pain sensitivity was observed in multiple studies of acute, short-term, and
subchronic durations [Gralewicz and Wiaderna, 2001: Gralewiczetal., 1997a: Korsak and
Rydzynski. 1996: Korsak etal., 1995]. Given the consistency of this effect and the determination
that decreased pain sensitivity is an adverse effect, in accordance with the U.S. EPA's Guidelines for
Neurotoxicity Risk Assessment [1998], decreased pain sensitivity was selected as the critical
effect and Korsak and Rydzynski (1996) was selected as the principal study for derivation of
the RfC for 1,2,4-TMB.
The RfC calculation is summarized in Table ES-2. The available rat PBPK model [Hissink et
al.. 2007] was used to convert the external concentrations (in mg/m3] from the animal study to the
internal blood metric of weekly average venous 1,2,4-TMB concentration (in mg/L]. These internal
blood metrics were then used as the dose inputs for benchmark dose (BMD] modeling. A
benchmark response (BMR] equal to a change in the mean equal to 1 standard deviation of the
model estimated control mean for decreased pain sensitivity was used. A BMDLiso of 0.086 mg/L
was estimated for decreased pain sensitivity in male rats exposed to 1,2,4-TMB via inhalation for 90
days (6 hours/day, 5 days/week] (Korsak and Rydzynski. 1996].
The available human PBPK model (Hissink et al., 2007] was then used to estimate a human
equivalent concentration (HEC] of 15.8 mg/m3 from the BMDLiso of 0.086 mg/L. This HEC was
used as the PODnEc with which to derive the RfC. A composite uncertainty factor (UF] of 1,000 was
applied: 3 to account for uncertainty in extrapolating from laboratory animals to humans
(interspecies variability], 10 to account for variation in susceptibility among members of the
human population (interindividual variability], 10 to account for subchronic-to-chronic
extrapolation due to the use of a subchronic study, and 3 to account for deficiencies in the database
(no two-generation reproductive/developmental toxicity or developmental neurotoxicity studies
were available]. Dividing the PODHEc by the composite UF of 1,000 yielded a chronic RfC of 2 x 10-
2 mg/m3 for 1,2,4-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
xxviii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Confidence in the Chronic Inhalation RfC for 1,2,4-TMB
A confidence level of high, medium, or low is assigned to the study used to derive the RfC,
the overall database, and the RfC itself, as described in Section 4.3.9.2 of EPA's Methods for
Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry [U.S. EPA,
1994bl
Confidence in the study from which the critical effect was identified, Korsak and Rydzyhski
[1996]. is medium. The study is a well-conducted, peer-reviewed study that utilized three dose
groups plus untreated controls, an appropriate number of animals per dose group, and performed
appropriate statistical analyses.
One area of uncertainty regarding this study is the lack of reported actual concentrations.
However, as the methods by which the test atmosphere was generated and analyzed were reported
in sufficient detail, and given the fact that this laboratory has used this methodology in subsequent
studies [Korsak et al., 2000a, b) and achieved appropriate actual concentrations (i.e., within 10% of
target concentrations), the concern regarding the lack of reported actual concentrations is minimal.
The critical effect on which the RfC is based is well-supported as the weight of evidence for
1,2,4-TMB-induced neurotoxicity is coherent across multiple animals species (i.e., human, mouse,
and rat) and consistent across multiple exposure durations (i.e., acute, short-term, and subchronic)
(Gralewicz and Wiaderna. 2001: Chen etal.. 1999: Wiaderna etal.. 1998: Gralewicz etal.. 1997a:
Gralewicz et al.. 1997b: Korsak and Rydzynski. 1996: Norseth et al.. 1991).
The database for 1,2,4-TMB includes acute, short-term, subchronic, and developmental
toxicity studies in rats and mice. However, confidence in the database is low to medium because it
lacks chronic, multi-generation reproductive/developmental, and developmental neurotoxicity
studies, and the studies supporting the critical effect predominantly come from the same research
institute. Consequently, the overall confidence in the RfC for 1,2,4-TMB is low to medium.
Inhalation Reference Concentration (RfC) for 1,2,3-TMB for Effects Other Than
Cancer
Table ES-3. Summary of reference concentration (RfC) derivation for
1,2,3-TMB
Critical effect
Decreased pain sensitivity
90 d male rat study
Korsak and Rydzynski
(1996)
Point of departure
PODHEC(mg/m3) = 16.3
Uncertainty
factor
1,000
Chronic RfC
(mg/m3)
2 x 10"2
29
30
Decreased pain sensitivity was observed in multiple studies of acute, short-term, and
subchronic durations (Lutz etal., 2010: Wiaderna etal., 1998: Korsak and Rydzynski, 1996). Given
This document is a draft for review purposes only and does not constitute Agency policy.
xxix DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 the consistency of this effect and the determination that decreased pain sensitivity is an adverse
2 effect, in accordance with the U.S. EPA's Guidelines for Neurotoxicity Risk Assessment [1998].
3 decreased pain sensitivity was selected as the critical effect and Korsak and Rydzynski
4 (1996) was selected as the principal study for derivation of the RfC for 1,2,3-TMB.
5 The RfC calculation is summarized in Table ES-3. BMD modeling was used in order to
6 identify the POD for decreased pain sensitivity. A BMR equal to a change in the control mean equal
7 to 1 standard deviation of the model estimated control mean was used. A BMDLiso of 17.36 mg/m3
8 was estimated for decreased pain sensitivity in male rats exposed to 1,2,3-TMB via inhalation for 90
9 days (6 hours/day, 5 days/week) [Korsak and Rydzynski. 1996).
10 As no PBPK model was available for 1,2,3-TMB, default dosimetry methodologies were
11 used to estimate the HEC of 16.3 mg/m3, based on the ratio of the human and animal blood:air
12 partition coefficients [U.S. EPA. 1994b]. This PODHEc was used to derive the RfC. A composite
13 uncertainty factor (UF) of 1,000 was applied: 3 to account for uncertainty in extrapolating from
14 laboratory animals to humans (interspecies variability), 10 to account for variation in susceptibility
15 among members of the human population (interindividual variability), 10 to account for
16 subchronic-to-chronic extrapolation due to the use of a subchronic study, and 3 to account for
17 deficiencies in the database (no two-generation reproductive/developmental toxicity,
18 developmental toxicity, or developmental neurotoxicity studies were available). Dividing the
19 PODHEc by the composite UF of 1,000 yielded a chronic RfC of 2 x 1Q-2 mg/m3 for 1,2,3-TMB.
20 Confidence in the Chronic Inhalation RfC for 1,2,3-TMB
21 Confidence in the study from which the critical effect was identified, Korsak and Rydzynski
22 [1996] is medium. The study is a well-conducted, peer-reviewed study that utilized three dose
23 groups plus untreated controls, an appropriate number of animals per dose group, and
24 appropriately performed statistical analyses.
25 One area of uncertainty regarding this study is the lack of reported actual concentrations.
26 However, as the methods by which the test atmosphere was generated and analyzed were reported
27 in sufficient detail, and given the fact that this laboratory has used this methodology in subsequent
28 studies [Korsak et al., 2000a, b) and achieved appropriate actual concentrations (i.e., within 10% of
29 target concentrations), the concern regarding the lack of reported actual concentrations is minimal.
30 The critical effect on which the RfC is based is well-supported as the weight of evidence for 1,2,3-
31 TMB-induced neurotoxicity is coherent across multiple animals species (i.e., mouse, and rat) and
32 consistent across multiple exposure durations (i.e., acute, short-term, and subchronic) (Lutz etal.,
33 2010: Wiadernaetal.. 1998: Korsak and Rydzynski. 1996).
34 The database for 1,2,3-TMB includes acute, short-term, and subchronic toxicity studies in
35 rats and mice. However, confidence in the database is low to medium because it lacks chronic,
36 multi-generation reproductive/developmental, developmental toxicity, or developmental
37 neurotoxicity studies, and the studies supporting the critical effect predominantly come from the
This document is a draft for review purposes only and does not constitute Agency policy.
xxx DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 same research institute. Consequently, the overall confidence in the RfC for 1,2,3-1MB is low to
2 medium.
3 Inhalation Reference Concentration (RfC) for 1,3,5-TMB for Effects Other Than
4 Cancer
5 No chronic or subchronic studies exist that would support the derivation of an RfC for
6 1,3,5-TMB, however two short-term neurotoxicity studies [Wiaderna et al., 2002: Gralewicz and
7 Wiaderna. 2001] and one developmental toxicity study [Saillenfaitetal., 2005] were identified as
8 potential studies from which to identify a critical effect for RfC derivation. Ultimately, the two
9 short-term neurotoxicity studies were inappropriate for the derivation of an RfC due to the
10 magnitude of the composite uncertainty factor associated with those data sets (i.e., a composite UF
11 > 10,000).
12 A developmental study by Saillenfait et al. [2005] showing decreased maternal weight gain
13 would result in an RfC 15-fold greater than that derived for 1,2,4-TMB (3 x lQ-i vs. 2 x 10-2 mg/m3).
14 This large difference is not consistent with the rest of the toxicological database for 1,2,4-TMB and
15 1,3,5-TMB, which demonstrates that the two isomers are similar to one another with regard to
16 respiratory and developmental toxicity in acute and developmental studies [Saillenfaitetal., 2005:
17 Korsak and Rydzynski, 1996: Korsaketal., 1995]. The 1,3,5-TMB isomer was observed to induce
18 some measures of neurotoxicity at lower doses than 1,2,4-TMB, and induces effects at a slightly
19 earlier time point compared to 1,2,4-TMB at the same concentration in short-term studies
20 [Wiaderna etal.. 2002: Gralewicz and Wiaderna, 2001]. Additionally, available toxicokinetic data
21 regarding blood:air partition coefficients, respiratory uptake, and absorption into the bloodstream
22 in humans and rats do not suggest any appreciable differences can be expected between the two
23 isomers [Meulenbergand Vijverberg. 2000: Tarnberg etal.. 1996: Dahl etal.. 1988].
24 Therefore, the chronic RfC of 2 x 10 2 mg/m3 derived for 1,2,4-TMB was adopted as the
25 RfC for 1,3,5-TMB. This is based on the determination of sufficient similarity with regard to
26 chemical properties, kinetics, and toxicity between the two isomers (see Section 2.3.3].
27 Confidence in the Chronic Inhalation RfC for 1,3,5-TMB
28 The chronic RfC for 1,2,4-TMB was adopted as the RfC for 1,3,5-TMB; thus, confidence in the
29 study from which the critical effect was identified, Korsak and Rydzynski (1996]. is medium (see
30 above]. The database for 1,3,5-TMB includes acute, short-term, and developmental toxicity studies
31 in rats and mice. However, confidence in the database is low to medium because it lacks chronic,
32 subchronic, multi-generation reproductive/developmental toxicity, and developmental
33 neurotoxicity studies and most of the studies supporting the critical effect come from the same
34 research institute.
This document is a draft for review purposes only and does not constitute Agency policy.
xxxi DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Toxicological Review of Trimethylbenzene
Reflecting the confidence in the study and the database and the uncertainty surrounding the
adoption of the RfC derived for 1,2,4-TMB as the RfC for 1,3,5-TMB, the overall confidence in the
RfC for 1,3,5-TMB is low.
Effects Other Than Cancer Observed Following Oral Exposure
No chronic, subchronic, or short-term studies were identified that examined the effects of
oral exposure to 1,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB. A series of studies utilizing single exposures
(oral gavage or i.p. injection) were identified that investigated the acute neurotoxic effects of TMBs
[Tomasetal..l999a: Tomas etal.. 1999b: Tomasetal.. 1999c]. In these studies, exposed rats
demonstrated changes in electrocortical arousal, altered EEC activity in the cortical and
hippocampal regions of the brain, and altered locomotor activity in open field tests. As these effects
were only observed in studies investigating acute exposures, they were considered insufficient for
derivation of oral toxicity reference values. Therefore, RfDs were derived for 1,2,4-TMB using
route-to-route extrapolation and for 1,2,3-TMB and 1,3,5-TMB based on sufficient similarity.
Table ES-4 below summarizes the RfDs derived for all three TMB isomers, and the sections
that follow provide details on the derivation of the RfD for each isomer.
Table ES-4. Summary of reference doses (RfDs) for TMB isomers
Isomer
Source
Reference value
Confidence
Oral reference dose (mg/kg-d)
1,2,4-TMB
1,2,3-TMB
1,3,5-TMB
Route-to-route extrapolation from RfC
for 1,2,4-TMB
Adopted from 1,2,4-TMB based on
sufficient similarity of these isomers
Adopted from 1,2,4-TMB based on
sufficient similarity of these isomers
6 x 10"3
6 x 10"3
6 x 10"3
Low
Low
Low
18
This document is a draft for review purposes only and does not constitute Agency policy.
xxxii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Oral Reference Dose (RfD) for 1,2,4-TMB for Effects Other Than Cancer
Table ES-5. Summary of reference dose (RfD) derivation for 1,2,4-TMB
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Critical effect
Decreased pain sensitivity
90 d male rat study
Korsak and Rydzyriski (1996)
Point of departure
Route-to-route extrapolation
using Korsak and Rydzyriski
(1996) subchronic inhalation
study in Wistar rats
PODHED (mg/kg-day) = 6.3
Uncertainty
factor
1,000
Chronic RfD
(mg/kg-day)
6 x 10"3
A human PBPK model (Hissink et al., 2007), modified by EPA to include an oral
compartment, was available for estimating the oral dose that would yield a blood concentration
equal to the blood concentration at the POD used in the derivation of the RfC for 1,2,4-TMB (Section
B.3.3.5, Appendix B). The RfD calculation is summarized in Table ES-5. Under the assumption of
constant oral ingestion and 100% absorption of 1,2,4-TMB via constant infusion rate into the liver,
a PODnED of 6.3 mg/kg-day was derived. Hepatic first-pass metabolism was also evaluated in
humans using the modified PBPK model: following 50 days of low daily doses, inhalation doses
were estimated to result in steady state venous blood concentrations 4-fold higher than blood
concentrations resulting from equivalent oral doses due to hepatic first pass metabolism (see
Figure B-18, Appendix B). The same composite UF of 1,000 used for the RfC derivation was applied:
3 to account for uncertainty in extrapolating from laboratory animals to humans (interspecies
variability), 10 to account for variation in susceptibility among members of the human population
(interindividual variability), 10 to account for subchronic-to-chronic extrapolation due to the use of
a subchronic study, and 3 to account for deficiencies in the database (no multi-generation
reproductive/developmental toxicity or developmental neurotoxicity studies). Dividing the PODHED
by the composite UF of 1,000 yielded a chronic RfD of 6 x 1Q-3 mg/kg-day for 1,2,4-TMB.
Confidence in the Chronic Oral RfD for 1,2,4-TMB
A PBPK model was utilized to perform a route-to-route extrapolation to determine a POD
for the derivation of the RfD from the Korsak and Rydzyhski (1996) inhalation study and
corresponding critical effect The confidence in the study from which the critical effect was
identified, Korsak and Rydzyhski (1996), is medium (see above). The database for 1,2,4-TMB
includes acute, short-term, subchronic, and developmental toxicity studies in rats and mice.
However, confidence in the database for 1,2,4-TMB is low to medium because it lacks chronic,
multi-generation reproductive/developmental and developmental neurotoxicity studies, and the
studies supporting the critical effect predominantly come from the same research institute.
This document is a draft for review purposes only and does not constitute Agency policy.
xxxiii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Reflecting the confidence in the study and the database and the uncertainty surrounding the
2 application of the available PBPK model for the purposes of a route-to-route extrapolation, the
3 overall confidence in the RfD for 1,2,4-TMB is low.
4 Oral Reference Dose (RfD) for 1,2,3-TMB for Effects Other Than Cancer
5 The oral database is inadequate to derive an RfD for 1,2,3-TMB. No chronic, subchronic, or
6 short-term oral exposure studies were found in the literature. However, as discussed in Sections
7 1.1.7 and B.2, the toxicokinetic and toxicity similarities between 1,2,3-TMB and 1,2,4-TMB support
8 adopting the RfD for 1,2,4-TMB as the RfD for 1,2,3-TMB. 1,2,3-TMB is observed to elicit the same
9 neurotoxic effects in rats (decreased pain sensitivity) as 1,2,4-TMB following subchronic inhalation
10 exposures, and the calculated RfCs for these two isomers are equal: 2 x 10~2 mg/m3. In addition to
11 the outlined similarities in toxicokinetics, the qualitative metabolic profiles for the two isomers are
12 similar such that first-pass metabolism through the liver is not expected to differ greatly between
13 1,2,4-TMB and 1,2,3-TMB. Therefore, the chronic RfC of 6 x 10-3 mg/kg-day derived for
14 1,2,4-TMB was adopted as the RfD for 1,2,3-TMB based on the determination of sufficient
15 similarity between the two isomers with regard to chemical properties, toxicokinetics, and toxicity.
16 Confidence in the Chronic Oral RfD for 1,2,3-TMB
17 The chronic oral RfD for 1,2,4-TMB was adopted as the chronic oral RfD for 1,2,3-TMB; thus,
18 confidence in the study from which the critical effect was identified, Korsak and Rydzyhski [1996].
19 is medium (see above). The database for 1,2,3-TMB includes acute, short-term, and subchronic
20 studies in rats and mice. However, confidence in the database is low to medium because it lacks
21 chronic, multi-generation reproductive/developmental, developmental toxicity, or developmental
22 neurotoxicity studies, and the studies supporting the critical effect predominantly come from the
23 same research institute. Reflecting the confidence in the study and the database and the
24 uncertainty surrounding the adoption of the RfD derived for 1,2,4-TMB as the RfD for 1,2,3-TMB,
25 the overall confidence in the RfD for 1,2,3-TMB is low.
26 Oral Reference Dose (RfD) for 1,3,5-TMB for Effects Other Than Cancer
27 The oral database is inadequate to derive an RfD for 1,3,5-TMB. No chronic, subchronic, or
28 short-term oral exposure studies were found in the literature. However, as determined for the RfC
29 derivation for 1,3,5-TMB, the toxicokinetic and toxicological similarities between 1,3,5-TMB and
30 1,2,4-TMB demonstrate sufficient similarity between the two isomers to support adopting the RfD
31 for 1,2,4-TMB for the RfD for 1,3,5-TMB. In addition to the previously discussed similarities in
32 toxicokinetics, the qualitative metabolic profiles for the two isomers are similar to such a degree
33 that first-pass metabolism through the liver is not expected to differ greatly between 1,2,4-TMB and
34 1,3,5-TMB. Therefore, the chronic RfD of 6 x 10-3 mg/kg-day derived for 1,2,4-TMB was
This document is a draft for review purposes only and does not constitute Agency policy.
xxxiv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 adopted as the RfD for 1,3,5-TMB based on the determination of sufficient similarity between the
2 two isomers with regard to chemical properties, toxicokinetics, and toxicity.
3 Confidence in the Chronic Oral RfD for 1,3,5-TMB
4 The chronic oral RfD for 1,2,4-TMB was adopted as the chronic oral RfD for 1,3,5-TMB; thus
5 confidence in the study from which the critical effect was identified, Korsak and Rydzyhski [1996].
6 is medium (see above). The database for 1,3,5-TMB includes acute, short-term, and developmental
7 toxicity studies in rats and mice. However, confidence in the database is low to medium because it
8 lacks chronic, multi-generation reproductive/developmental and developmental neurotoxicity
9 studies, and the studies supporting the critical effect predominately come from the same research
10 institute. Reflecting the confidence in the study and the database and the uncertainty surrounding
11 the adoption of the RfD derived for 1,2,4-TMB as the RfD for 1,3,5-TMB, the overall confidence in
12 the RfD for 1,3,5-TMB is low.
13 Evidence of Carcinogenicity
14 Under EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005). there is
15 "inadequate information to assess carcinogenic potential" of TMBs. No chronic inhalation
16 studies that investigated cancer outcomes were identified in the literature for 1,2,3-TMB, 1,2,4-
17 TMB, or 1,3,5-TMB. One cancer study in which rats were exposed to 1,2,4-TMB via oral gavage at
18 one experimental dose of 800 mg/kg-day observed marginal increases in total malignant tumors
19 and head tumors (e.g., neuroesthesioepitheliomas), but provided no statistical analyses of the
20 results (Maltoni etal.. 1997]. A number of methodological issues limit the utility of this study (e.g.,
21 only one dose group and no discussion of histopathological analyses]. When Fisher's exact test was
22 performed by EPA on the incidences calculated from the reported percentages of animals bearing
23 tumors in the control and 800 mg/kg dose groups, no statistically significant elevations were
24 observed. Therefore, a quantitative cancer assessment for TMBs was not conducted.
25 Susceptible Populations and Lifestages
26 No chemical-specific data that would allow for the identification of populations or lifestages
27 with increased susceptibility to TMB exposure exist However, some inferences can be made based
28 on the toxicokinetics of TMB isomers. TMB isomers are metabolized via side-chain oxidation to
29 form alcohols and aromatic carboxylic/mercapturic acids or by hydroxylation to form phenols,
30 which are then conjugated with glucuronic acid, glycine, or sulfates for urinary excretion. The
31 activities of multiple cytochrome P450 (GYP P450] mono-oxygenase isozymes and rates of
32 glucuronidation and sulfation conjugation are reduced in children up to 1 year in age, and renal
33 clearance is reduced in infants up to 2 months of age (Ginsberg et al., 2004]. Therefore, as GYP P450
34 mono-oxygenase activities, the rate of glucuronidation and sulfation, and renal clearance appear to
35 be decreased in early life, newborns and young infants may experience higher and more persistent
This document is a draft for review purposes only and does not constitute Agency policy.
xxxv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 blood concentrations of 1,2,3-TMB, 1,2,4-TMB, 1,3,5-TMB, and/or their respective metabolites
2 compared with adults at similar exposure levels. Additionally, those with pre-existing respiratory
3 diseases (e.g., asthma) may be more sensitive to the respiratory irritative and inflammatory effects
4 resulting from exposure to 1MB isomers.
5 Key Issues Addressed in the Assessment: Adoption of 1,2,4-TMB Toxicity Values for
6 the 1,3,5-and 1,2,3-TMB Isomers
7 The toxicity database for 1,3,5-TMB was inadequate for derivation of either a reference
8 concentration or a reference dose. The chemical, toxicokinetic, and toxicological properties of the
9 individual isomers are sufficiently similar to one another to support adoption of 1,2,4-TMB's
10 reference values for 1,3,5-TMB (see Section 2.3.3). Both isomers are similar in their (1) chemical
11 properties (e.g., blood:tissue partition coefficients), (2) toxicokinetic properties (i.e., absorption,
12 metabolism, and excretion profiles), and (3) toxicity profiles across studies utilizing multiple
13 durations of exposure and multiple endpoints (i.e., neurological, respiratory, maternal, and fetal
14 effects). Therefore, given these similarities, the RfC and RfD derived for 1,2,4-TMB were adopted as
15 the RfC and RfD for 1,3,5-TMB.
16 The toxicity database for 1,2,3-TMB was inadequate for derivation of a reference dose. No
17 chemical-specific PBPK model is available for 1,2,3-TMB, and therefore, no route-to-route
18 extrapolation can be performed on which to derive an RfD from the RfC for 1,2,3-TMB. The
19 chemical, toxicokinetic, and toxicological properties of the individual isomers are sufficiently
20 similar to one another to support adoption of 1,2,4-TMB's reference dose for 1,2,3-TMB (see
21 Section 2.5.2). Both isomers are similar in their (1) chemical properties (e.g., blood:air and
22 tissue:air partition coefficients), (2) toxicokinetic properties (i.e., the degree of absorption into the
23 bloodstream between the two isomers indicates the internal blood dose metrics for 1,2,3-TMB
24 would be similar to those calculated for 1,2,4-TMB by that isomer's available PBPK model), and (3)
25 toxicity profiles (i.e., the observation that both isomers affected pain sensitivity to an equal degree
26 and that the two isomer's RfCs for this effect were equal). Therefore, given these similarities, the
27 deficiencies in the 1,2,3-TMB oral database, and the lack of a 1,2,3-TMB PBPK model with which to
28 perform a route-to-route extrapolation, the RfD derived for 1,2,4-TMB was adopted as the RfD for
29 1,2,3-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
xxxvi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
LITERATURE SEARCH STRATEGY | STUDY
SELECTION
4
5
6
7
8
9
10
11
12
13
The literature search strategy used to identify primary, peer-reviewed literature pertaining
to TMBs was conducted using the databases and keywords listed in Table LS-1. References from
health assessments developed by other national and international health agencies were also
examined. Other peer-reviewed information, including review articles, literature necessary for the
interpretation of TMB-induced health effects, and independent analyses of the health effects data
were retrieved and included in the assessment where appropriate. EPA requested public
submissions of additional information in April 2008; no submissions in response to the data call-in
were received. A comprehensive literature search was last conducted in December 2011.
Table LS-1: Details of the search strategy employed for TMBs
Databases
Keywords'
a,b
EBSCO
DISCOVERY
SERVICE:
HERO
SCI
NLM
TOXLINE
WOS
Chemical name, CASRN, and synonym search: 1,2,4-trimethylbenzene OR pseudocumene OR 95-
63-6; 1,2,3-trimethylbenzene OR hemimellitene OR 526-73-8; 1,3,5-trimethylbenzene OR
mesitylene OR 108-67-8
Keyword search: neurotoxicity, genotoxicity, developmental toxicity, inflammation, irritation,
toxicokinetics, pbpk, mode of action, white spirit, C9, C9 fraction, JP-8
Additional search on specific metabolites: 2,3-dimethylbenzoic acid OR 26998-80-1; 2,3-
dimethylhippuric acid OR 187980-99-0; 2,4-dimethylbenzoic acid OR 611-01-8; 2,4-
dimethylhippuric acid OR 41859-41-0; 2,5-dimethylbenzoic acid OR 610-72-0; 2,5-dimethylhippuric
acid OR 41859-40-9; 2,6-dimethylbenzoic acid OR 632-46-2; 2,6-dimethylhippuric acid OR 187980-
98-9; 3,4-dimethylbenzoic acid OR 619-04-5; 3,4-dimethylhippuric acid OR 23082-12-4; 2,4,5-
trimethylphenol OR 496-78-6; 2,3,5-trimethylphenol OR 697-82-5; 2,3,6-trimethylphenol OR 2416-
94-6; 2,4,6-trimethylphenol OR 527-60-6; 3,5-dimethylbenzoic acid OR 499-06-9; 3,5-
dimethylhippuric acid OR 23082-14-6
14
15
16
Potentially relevant publications on TMBs were identified through a literature search conducted with the EBSCO
Discovery Service feature of Health and Environmental Research Online (HERO), a meta-search engine with
access to numerous databases including the Science Citation Index (SCI), Toxicology Literature Online (TOXLINE),
The National Library of Medicine (NLM, PubMed/Medline), and Web of Science (WOS).
bLiterature search was performed using related words (i.e., lemmatization) of included search terms. Search terms
were entered into the EBSCO Discovery Service portal with no qualifiers and the results from individual search
engines were returned and exported to HERO.
Figure LS-1 depicts the literature search and study selection strategy and the number of
references obtained at each stage of the literature screening. Approximately 4300 references were
obtained from the chemical name, keyword, and metabolite searches for 1,2,4-TMB, 1,2,3-TMB, and
This document is a draft for review purposes only and does not constitute Agency policy.
xxxvii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1
2
3
4
5
6
7
8
9
10
11
1,3,5-TMB including references retrieved from specific literature searches necessary for the
interpretation of TMB-induced health effects (e.g., literature on specific modes of action, PBPK
analysis). From this full list of references, there were 218 references that were considered for
inclusion in the Toxicological Review.
References identified based on initial keyword search {see Table LS-1): "4300
Reference excluded based on preliminary manual screen of
titles/abstracts: "4100
Reasons/or excluding references Included the following:
* Published in non-relevant journals (i.e., Journal of Chemical
Catalysis
• Topics not relevant to TMB toxicity
Reference considered for inclusion in the Toxicological Review: 218
Hum an Studies: 40
Animal Studies: 74
Other supporting studies: 138
Including:
• Reviews
* Background and physical/chemical properties
• Studies related to mode-of-action
• Studies related to toxicokinetics or PBPK analyses
References excluded based on manual review of
papers/abstracts: 61
Typesof papers evaluated and not considered further for
Inclusion in the Toxicological Review:
' Studies of exposure assessment
•Studies of chemicals other than TMBs
• Non-relevant human and animal toxicity studies
• Case reports
• Ecotoxicity studies
• In vitro studies
• Not available in English
References cited in the Toxicological Review
Human studies/reports: 34
•Controlled Human Exposures: 11
• Dermal Toxicity: 1
• Epidemiological Studies: 13
• Hematatological: 3
• Neuro toxicity: 10
• Respiratory: 6
•Toxicokinetic Studies: 15
• Workplace Exposure studies: 10
Animal Studies: 61
• Cancer/Genotoxicity:
• Developmental: 4
•Hematological: 8
• Nephro toxicity: 1
• Neuro toxicity: 22
• Respiratory: 6
•Toxicokinetic: 19
Other supporting studies: 70
• Chemical properties: 6
•BMD modeling: 6
• Exposure assessment: 5
• Mode of action: 22
•PBPIC29
• Susceptible populations: 3
Guidance documents: 32
•EPA:24
•Non-EPA:8
Figure LS-1. Literature search and study selection strategy for TMBs
Note: Some references may provide information on more than one topic, and therefore, may
be included in more than one study type. Accordingly, the sum of the references for
subcategories of studies is not expected to equal the number of references for the larger
category.
This document is a draft for review purposes only and does not constitute Agency policy.
xxxviii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Selection of studies for inclusion in the Toxicological Review was based on consideration of
2 the extent to which the study was informative and relevant to the assessment and general study
3 quality considerations. In general, the relevance of health effect studies was evaluated as outlined
4 in the Preamble and EPA Guidance (Review of the Reference Dose and Reference Concentration
5 Processes [2002] and Methods for Derivation of Inhalation Reference Concentrations and Application
6 of Inhalation Dosimetry (1994b)]. From the list of "considered" references, 157 full text
7 publications were identified as providing relevant information for use in the development of this
8 document, and included 34 studies in humans (e.g., occupational epidemiologic studies, workplace
9 exposure studies, and controlled human exposures), 61 inhalation or oral animal studies, and 70
10 other studies (e.g., studies that provided supporting information on mode of action, chemical
11 properties, and susceptible subpopulations).
12 Although a number of industry reports or TSCA submissions regarding the toxicity of the
13 1MB isomers, or mixtures containing the isomers were located, these documents were excluded
14 from the Toxicological Review following careful consideration (Koch Industries. 1995a. bj
15 Industrial Bio-Test Laboratories. 1992: Chevron Chemical Company. 1985: Borriston Labs. 19831.
16 These reports were not peer-reviewed and they either did not use appropriate durations of
17 exposure that would support derivation of chronic human health reference values (e.g., 14 days),
18 reported minimal and difficult to interpret toxic effects, or investigated mixtures containing TMB
19 isomers. Ultimately, the decision was made to not seek external peer review for these documents
20 as these studies would not qualitatively enhance hazard identification, quantitatively enhance dose-
21 response analysis, or substantially decrease uncertainty in the assessment
22 The references that are cited in the document, as well as those that were considered but not
23 included in the Toxicological Review of TMBs, can be found within the Health and Environmental
24 Research Online (HERO) website (https://hero.epa.gov/tmb)2. This site contains HERO links to
25 lists of references, including bibliographic information and abstracts, which were considered for
26 inclusion in the Toxicological Review of TMBs.
2 HERO is a database of scientific studies and other references used to develop EPA's risk assessments aimed at
understanding the health and environmental effects of pollutants and chemicals. It is developed and managed in
EPA's Office of Research and Development (ORD) by the National Center for Environmental Assessment (NCEA).
The database includes more than 300,000 scientific articles from the peer-reviewed literature. New studies are
added continuously to HERO.
This document is a draft for review purposes only and does not constitute Agency policy.
xxxix DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1.HAZARD IDENTIFICATION
3 1.1. Synthesis of Evidence
4 1.1.1. Neurological Effects
5 There is evidence in humans and animals that inhalation exposure to trimethylbenzenes
6 (TMBs) induces neurotoxic effects. The human evidence comes from occupational studies involving
7 complex volatile organic compound (VOC) mixtures that include TMBs; thus, effects cannot be
8 attributed to any TMB isomer specifically. Prevalence rates of neuropsychological symptoms
9 increased with exposure duration in dockyard painters, with symptoms related to motor
10 coordination exhibiting the strongest association [Chenetal., 1999]: similarly, a significant
11 association between exposure and impaired performance in short term memory (symbol digit
12 substitution) and motor speed/ coordination (finger tapping) tests was observed in shipyard
13 painters exposed to TMBs (isomers were not specified) and other solvents (Lee etal., 2005). A
14 significant, positive association between exposure symptoms (e.g., abnormal fatigue) and
15 1,2,4-TMB exposure, but not exposure to lower levels of 1,2,3-TMB or 1,3,5-TMB, was reported in
16 asphalt workers (Norsethetal., 1991). Nervousness, tension, headaches, vertigo, and anxiety were
17 reported in paint shop workers exposed to 49-295 mg/m3 of a solvent mixture containing 50%
18 1,2,4-TMB, 30% 1,3,5-TMB, and unspecified amounts of 1,2,3-TMB (listed as possibly present)
19 (Battig et al. (1956). as reviewed by MOE (2006) and Baettig et al. (1958)).
20 Additional evidence suggests damage or dysfunction of the inner ear and increased
21 occurrence of vertigo following exposure to TMBs, and other organic solvents in paint and varnish
22 factories (Sulkowski etal.. 2002). Increased reaction time was significantly and consistently
23 associated with exposure in controlled, acute volunteer studies in which humans were exposed to
24 mixtures containing 1,2,4-TMB (Lammers et al., 2007], although it is unclear whether 1,2,4-TMB or
25 other constituents within the mixtures were responsible for the observed effects (for controlled
26 human exposures, see individual study summary tables in the appendices for information on
27 human subjects research ethics procedures). Uptake of TMBs was reported in human volunteers
28 exposed for 2 hours to either: 1) 300 mg/m3 white spirit (WS, corresponding to 11 mg/m31,2,4-
29 TMB), 2) 11 or 123 mg/m31,2,4-TMB, 3) 123 mg/m31,2,3-TMB, 4) or 123 mg/m31,3,5-TMB.
30 However, effects on the central nervous system (CNS) were based on measures of overt CNS
31 depression (heart rate and pulmonary ventilation) and a subjective rating of CNS symptoms (i.e.,
32 headache, fatigue, nausea, dizziness, and intoxication), and were not observed (Tarnbergetal.,
33 1997a: Tarnbergetal.. 1996).
34 In two studies examining controlled human exposures to 5-150 mg/m31,2,3-TMB, 1,2,4-
35 TMB, or 1,3,5-TMB, no neurological abnormalities in routine clinical examinations were reported
This document is a draft for review purposes only and does not constitute Agency policy.
1 -1 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 following exposure, although details regarding the specific tests performed were not provided
2 [Kostrzewskietal.. 1997: Kostrewski and Wiaderna-Brycht. 1995). Studies identifying an
3 association between occupational exposure to 1MB isomers and neurological effects are limited
4 due to an inability to attribute effects due to 1,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB individually
5 versus those due to the other isomers or additional constituents within the mixture. The studies
6 detailing controlled exposures to human volunteers are also limited for evaluating neurotoxicity to
7 TMBs due to a lack of methods to adequately assess CNS function and a lack of no-exposure
8 controls, short exposure duration, and exposure of individual subjects to different concentrations of
9 TMB isomers.
10 In animals, there is consistent evidence of neurotoxicity following inhalation or oral
11 exposure to either 1,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB; a summary of the evidence pertaining to
12 neurotoxic effects for TMBs is shown in Tables 1-1 and 1-2 for inhalation and oral exposures,
13 respectively. This information is presented graphically in Figures 1-1 to 1-4.
14
15 Pain sensitivity
16 Decreased pain sensitivity has been observed following inhalation exposure to TMBs in
17 multiple studies conducted in male Wistar rats. To test pain responses following TMB exposure,
18 animal studies have employed the hot plate test In this test, a thermal stimulus is applied to
19 determine pain sensitivity, as indicated by the animals' latency to paw-lick following introduction of
20 the stimulus. Decreases in pain sensitivity have been observed at concentrations > 492 mg/m3
21 following subchronic and short-term exposure to 1,2,4-TMB [Wiaderna et al.. 2002: Gralewicz and
22 Wiaderna, 2001: Korsak and Rydzynski, 1996] and short-term exposure to 1,3,5-TMB [Wiaderna et
23 al.. 2002: Gralewicz and Wiaderna, 2001]. Decreased pain sensitivity was also observed at
24 concentrations > 123 mg/m3 or > 492 mg/m3 following subchronic or short-term exposure to 1,2,3-
25 TMB, respectively [Wiaderna et al.. 1998: Korsak and Rydzynski. 1996]. although changes were not
26 observed at 492 mg/m31,2,3-TMB in another short-term exposure study [Gralewicz and Wiaderna,
27 20011 In the subchronic study fKorsak and Rydzynski. 19961 inhalation of 1,2,4-TMB or 1,2,3-TMB
28 resulted in reduced pain sensitivity which occurred in a concentration-dependent manner.
29 In short-term studies that examined a range of concentrations [Wiaderna et al., 2 00 2:
30 Gralewicz etal.. 1997a: Wiaderna et al.. 1998] these decreases in pain sensitivity following
31 exposure to TMB isomers were non-monotonic. Differences in experimental design (discussed
32 below] may account for the lack of monotonicity in these short-term studies, in contrast to the
33 observations in Korsak and Rydzynski [1996]. Similar to the subchronic study, acute exposures to
34 1,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB induced concentration-dependent decreases in pain sensitivity,
35 with ECso values of 5,682, 4,172 , and 5,963 mg/m3 for increased latency to paw-lick compared to
36 controls, respectively [Korsak and Rydzynski, 1996: Korsak et al., 1995].
37 A second, somewhat different measure of pain sensitivity was reported in the studies
38 evaluating performance in the hot plate test (before and after footshock] several weeks following
39 short-term (i.e., 4-week], inhalation exposure to TMB isomers (Wiaderna et al., 2002: Gralewicz and
This document is a draft for review purposes only and does not constitute Agency policy.
I -2 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Wiaderna, 2001: Wiadernaetal., 1998: Gralewiczetal., 1997a]. In these studies, treatment-related,
2 statistically significant changes in pain sensitivity at > 492 mg/m31,2,3-TMB, 1,2,4-TMB, or 1,3,5-
3 TMB were observed 24 hours after rats were given a footshock; no statistically significant effects at
4 any concentration were observed prior to or immediately following footshock. These findings
5 indicate that inhalation exposure to TMBs may prolong footshock-induced reductions in pain
6 sensitivity. It is also plausible that an amplification of responses associated with classically
7 conditioned analgesia (i.e., decreased pain sensitivity) occurs following TMBs exposure.
8 Specifically, footshock can cause contextual cues (e.g., the hot plate test apparatus) to become
9 associated with the noxious stimulus (footshock), inducing stress or fear-related responses in the
10 shocked animal such that, subsequently, both footshock itself as well as the contextual cues
11 associated with footshock, can reduce sensitivity to pain (possibly via the release of endogenous
12 opiods). Thus, exposure to the hot plate apparatus immediately following footshock may associate
13 this test environment with the footshock, such that subsequent re-exposure to the hot plate
14 apparatus can, itself, produce analgesia. From the data available, the relative contribution of these
15 behaviors to the observed effects cannot be easily distinguished.
16 The decreases in pain sensitivity measured in the subchronic (Table 1-1) and acute studies
17 were observed immediately after exposure, with no significant effects persisting 2 weeks after
18 exposures were terminated (Korsak and Rydzynski, 1996: Korsak et al., 1995). In contrast,
19 performance in the hot plate test was significantly impaired following short-term exposure to the
20 TMB isomers when tested 50-51 days after exposure (Wiaderna et al.. 1998) (Wiadernaetal..
21 2002: Gralewicz and Wiaderna. 2001: Gralewicz etal.. 1997a). indicating a persistence of these
22 effects. It is not clear why effects are observed to persist following the short-term exposures, but
23 not the subchronic exposures, although the testing paradigm between studies was substantially
24 different. Regardless, the ability of male Wistar rats to respond to a thermal stimulus in the hot
25 plate test was consistently impaired following inhalation exposure to TMBs. Although some studies
26 suggest a slightly more pronounced analgesic effect of 1,2,3-TMB as compared to the other isomers
27 (Table 1-1), the overall database does not provide sufficient support for this conclusion, indicating
28 that TMBs are similar in their capacity to decrease pain sensitivity. Pain sensitivity was not
29 examined following oral exposure.
30
31 Neuromuscularfunction and coordination
32 Human exposures to solvent mixtures containing 1,2,4-TMB (Lammers etal., 2007) or
33 multiple TMB isomers (Battig et al. (1956). as reviewed by MOE (2006) and Baettig et al.
34 (1958))(Lee etal.. 2005: Sulkowski etal.. 2002) result in effects that suggest alterations to
35 neuromuscular function and balance, including increased reaction time and vertigo. Animal studies
36 using rotarod performance, which tests motor coordination, balance, and overall neuromuscular
37 function, indicate that inhalation of TMB isomers can affect neuromuscular system function (Table
38 1-1). Significant decreases in rotarod performance were observed at 1,230 mg/m31,2,4-TMB and >
39 493 mg/m31,2,3-TMB when tested immediately after exposure for 13 weeks (Korsak and
This document is a draft for review purposes only and does not constitute Agency policy.
I -3 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Rydzynski, 1996]: significant decreases in performance were also observed at 1,230 mg/m3 after 4
2 or 8 weeks of exposure to 1,2,3-TMB or 1,2,4-TMB, respectively. This impaired function was still
3 evident at 2 weeks post-exposure and, while not statistically significant for 1,2,4-TMB, may indicate
4 long-lasting neuromuscular effects of subchronic exposures to 1,2,4-TMB and 1,2,3-TMB. Acute
5 inhalation exposure studies support this observation. Effects such as loss of reflexes and righting
6 responses, have been observed following acute inhalation exposure to 1,250-45,000 mg/m3
7 1,2,4-TMB [MOE. 2006: Henderson. 2001). Similarly, acute exposure to 1,2,3-TMB, 1,2,4-TMB, or
8 1,3,5-TMB resulted in decreased performance in rotarod tests immediately following exposure,
9 with ECso values of 4,693 mg/m3, 3,779 mg/m3,and 4,738 mg/m3, respectively [Korsak and
10 Rydzynski. 1996: Korsak etal.. 1995]. Similar to observations related to effects on pain sensitivity,
11 these results indicate that 1,2,4-TMB and 1,3,5-TMB may be similar in their ability to impair
12 neuromuscular function, balance, and coordination while 1,2,3-TMB exposure may elicit effects at
13 lower concentrations compared to the other two isomers. No studies evaluating oral exposure to
14 TMB isomers address this endpoint
15 The neurobehavioral tests administered (i.e., hot plate and rotarod) in the subchronic and
16 acute studies by Korsak and Rydzynski, [1996] and Korsak et al. [1995] appear to have been
17 conducted on the same days; however, it is unclear whether the tests were performed sequentially
18 in the same cohorts of animals. Performing the hot plate test immediately following the rotarod
19 test could introduce a potential confounder, as shock alone (such as that used as negative
20 reinforcement following rotarod failure, see Table B-29, Appendix B] can cause reductions in pain
21 sensitivity. Thus, if the tests were performed sequentially in the same animals, TMB-exposed
22 animals failing more often in the rotarod test may exhibit increases in paw-lick latency unrelated to
23 treatment, as compared to controls receiving less shock reinforcement. However, the observations
24 by Korsak and Rydzynski, [1996] and Korsak et al. [1995] are supported by 2- to 3-fold increases in
25 latency to paw-lick that, although not statistically significant, were observed up to 7 weeks after
26 termination of short-term exposures to 492 mg/m31,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB [Gralewicz
27 and Wiaderna. 2001]: increases of this magnitude were not present in the studies evaluating
28 multiple concentrations of the isomers [Wiaderna etal.. 2002.1998: Gralewicz etal.. 1997a].
29
30 Motor function and/or anxiety
31 Effects in open field testing have been consistently reported in oral and inhalation studies of
32 exposure to 1,2,4-TMB and 1,3,5-TMB, but not 1,2,3-TMB, in male rats (Table 1-1]. Altered
33 behaviors and locomotion in open field tests can be attributed to anxiety responses due to open
34 spaces and bright light, as well as changes to motor system function. Decreased anxiety and/or
35 increased motor function at > 492 mg/m31,2,4-TMB or 1,3,5-TMB has been reported in short-term
36 studies, as evidenced by increases in horizontal locomotion or grooming activities [Lutz etal., 2010:
37 Gralewicz and Wiaderna. 2001: Gralewicz etal., 1997a]. Statistically significant increases in
38 horizontal locomotion were observed in short-term studies assessing open field behavior following
39 inhalation exposure to 1,2,4-TMB or 1,3,5-TMB [Lutz etal.. 2010] [Gralewicz and Wiaderna. 2001].
This document is a draft for review purposes only and does not constitute Agency policy.
I -4 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Non-monotonic increases in grooming were reported following short-term exposure to 1,2,4-TMB,
2 although changes in horizontal locomotion were not statistically significant (increases of 3-35%
3 were also non-monotonic) [Gralewicz etal., 1997a]. No effects on open field activity have been
4 observed following short-term exposure of male rats to 1,2,3-TMB [Lutz etal., 2010: Gralewicz and
5 Wiaderna. 2001: WiadernaetaL 1998).
6 Open field locomotion following injections with the stimulant amphetamine was amplified
7 by previous short-term exposure to 1,2,3-TMB, but not 1,2,4-TMB (which actually tended to inhibit
8 amphetamine-induced increases in activity), suggesting possible effects of 1,2,3-TMB on
9 sensitization-type responses. Although contributing factors other than anxiety and motor function
10 may explain alterations in open field behavior, the experimental tests employed in the above
11 studies are insufficient to identify these factors as all but one of the studies (Lutz etal., 2010]
12 observed animals for only 5 or 10 minutes. Thus, EPA has concluded that decreased anxiety and/or
13 increased motor function are the two most likely explanations for the TMB-induced effects. As
14 open field testing was conducted 14 or 25 days after termination of exposure in these studies and
15 TMB isomers are cleared rapidly from the body following the end of inhalation exposures (Section
16 B.2, Appendix B), the results suggest persistence of the effects of 1,2,4-TMB and 1,3,5-TMB on
17 anxiety and/or motor function following clearance of the toxic moiety from the nervous system.
18 Slight, transient increases in locomotor activity were also observed in open field tests
19 immediately following acute, oral exposure to the TMB isomers (Table 1-2). Significant increases in
20 locomotor activity—measured as number of squares crossed after exposure compared with prior to
21 exposure—were observed at 3,850 mg/kg for 1,2,4-TMB and 1,2,3-TMB, and at > 1,920 mg/kg for
22 1,3,5-TMB, with minimal concentration-effect or time-effect relationships and negligible differences
23 in the magnitude of the change in activity between isomers (Tomas etal., 1999b). Increases in
24 locomotor activity were biphasic in nature. At early timepoints following exposure, increased
25 locomotor activity was associated with perturbed motor coordination and tremor, whereas after 90
26 minutes, this apparent motor ataxia progressed to hind limb paralysis, full immobility, and
27 respiratory distress (e.g., tachypnea), leading to several deaths by 24 hours (Tomas etal.. 1999b).
28 As mentioned previously, open field tests cannot easily distinguish between anxiety-related
29 responses and changes in motor activity. However, effects on motor activity were observed
30 following inhalation exposure to elevated concentrations of TMBs in several acute studies, although
31 the results are somewhat inconsistent with observations in open field tests. Decreased motor
32 activity was observed in male rats immediately after exposure to 5,000 mg/m31,2,4-TMB (McKee
33 etal.. 2010). Decreased motor activity was also reported in rats acutely exposed via inhalation to a
34 mixture containing TMB isomers (Lammers etal.. 2007). but the use of a mixture precludes a
35 determination of the toxicity specifically associated with individual isomers. As biphasic changes in
36 activity are frequently observed following exposures to solvents, it is likely that the timing of the
37 evaluations conducted in the short-term versus acute studies, as well as the differing isomer
38 concentrations, may influence the consistency of these results.
This document is a draft for review purposes only and does not constitute Agency policy.
I -5 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Overall, exposure to 1,2,4-TMB and 1,3,5-TMB affects anxiety and/or motor function at
2 concentrations above 492 mg/m3, although the exact, potentially biphasic, concentration-response
3 relationship remains unclear. The results for 1,2,3-TMB are difficult to interpret, as no effects were
4 observed following short term inhalation exposure while acute oral exposure elicited responses
5 consistent with 1,2,4-TMB and 1,3,5-TMB. Although an explanation for this disparity is lacking,
6 these data highlight a potential difference between 1,2,3-TMB and the other isomers.
7
8 Cognitive function
9 Cognitive function following exposure to TMB isomers alone has not been evaluated in
10 humans or following oral exposure in animals; controlled exposure of human volunteers to
11 mixtures containing TMBs did not indicate any effects on short-term learning and memory tests
12 [Lammers etal., 2007]. Similarly, short-term spatial memory (radial maze performance) was
13 unaffected by exposure to either 1,2,4-TMB or 1,3,5-TMB via inhalation in animal studies
14 [Wiadernaetal.. 2002: Gralewicz and Wiaderna. 2001: Gralewicz etal.. 1997a). although one study
15 indicates a significant decrement in performance following exposure to 123 mg/m3, but not higher
16 concentrations, of 1,2,3-TMB (Wiadernaetal.. 1998).
17 In contrast, effects on cognitive function in different neurobehavioral tests, observed as
18 altered conditioning behaviors, were consistently observed across multiple studies in male rats 4-8
19 weeks following short-term inhalation exposure to the TMB isomers, although clear concentration-
20 effect relationships were not observed (Table 1-1). Comparing the results of the behavioral tests
21 reveals that there are differences in neurological effects reported for each TMB isomer, as well as
22 differences in the concentrations at which the cognitive effects were observed. Decreased step-
23 down latency in passive avoidance tests were observed 35-45 days after short-term inhalation
24 exposure to > 123 mg/m31,2,3-TMB and 1,3,5-TMB or > 492 mg/m31,2,4-TMB (Wiadernaetal..
25 1998] (Wiaderna etal.. 2002: Gralewicz and Wiaderna. 2001: Gralewicz etal.. 1997a): decreased
26 step-down latency may be attributed to a reduced ability to inhibit motor reactions (or a lowered
27 motor threshold) in response to stress. These responses were consistently observed and similar in
28 magnitude across all studies at 7 days post footshock (a 30% decrease in latency following 1,2,3-
29 TMB exposure was not statistically significant in Gralewicz and Wiaderna (2001)]. At 3 days post
30 footshock, decreases in latency were inconsistent (i.e., decreased at 123 mg/m31,2,3-TMB and at
31 492 mg/m31,2,4-TMB and 1,3,5-TMB, but not at other concentrations). Statistically significant
32 changes were observed < 24 hours following footshock only after exposure to 123 mg/m31,2,3-
33 TMB, suggesting that 1,2,4-TMB and 1,3,5-TMB exposure, and possibly 1,2,3-TMB exposure, may
34 have a particular effect on adaptive behaviors associated with the persistence of stress or fear-
35 related responses. Reduced active avoidance learning was also observed in male rats following
36 short-term inhalation exposure to 492 mg/m31,2,4-TMB (Gralewicz and Wiaderna, 2001):
37 however, these changes were not observed in the other 1,2,4-TMB short-term study (Gralewicz et
38 al.. 1997a). Decreased performance in active avoidance tests was consistently observed following
39 short-term exposure to > 123 mg/m31,3,5-TMB and at 492 mg/m31,2,3-TMB (Wiadernaetal..
This document is a draft for review purposes only and does not constitute Agency policy.
I -6 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 1998: Wiadernaetal.. 2002: Gralewicz and Wiaderna. 20011. Similar to 1,2,4-TMB (Gralewicz and
2 Wiaderna. 2001). the effects of 1,2,3-TMB and 1,3,5-TMB were particular to the learning component
3 of the test (acquisition training), rather than the memory component (retention session 7 days
4 later) (Wiadernaetal., 2002: Gralewicz and Wiaderna, 2001: Wiaderna et al., 1998). It is unclear
5 whether potential alterations in locomotor activity by 1MB isomers would affect performance in
6 these tests.
7 Acute inhalation exposure studies provide some support for the observed effects of TMB
8 isomers on learned behaviors. Significant increases in response latency in psychomotor tasks,
9 observed immediately after exposure (effects did not persist to 24 hours later), were reported in
10 male rats following acute exposure to 5,000 mg/m31,2,4-TMB (McKee etal.. 2010) or to 4,800
11 mg/m3 of a mixture containing TMBs (Lammers etal., 2007). The effects on active and passive
12 avoidance behaviors indicate that learning and/or long-term memory processes are affected by
13 exposure to the TMB isomers. The data suggest that 1,3,5-TMB may be a more potent inducer of
14 toxic effects on cognitive function than 1,2,4-TMB and 1,2,3-TMB, as the effects following exposure
15 to 1,3,5-TMB were more consistent and sometimes occurred at lower concentrations than those
16 reported following exposure to the other two isomers. Overall, however, these differences were
17 slight
18 Controlled human exposure studies suggest that exposures of < 123 mg/m3 of the TMB
19 isomers do not cause overt CNS depression (measured as heart rate and respiration) (Tarnberg et
20 al.. 1996). although symptoms related to this effect (e.g., lightheadedness, fatigue) have been
21 reported in workers occupationally exposed to mixtures containing TMBs. In mice, CNS depression
22 has been observed following acute inhalation exposure to > 25,000 mg/m31,3,5-TMB, with similar
23 effect levels for 1,2,4-TMB (ACGIH. 2002).
24
25 Electrocortical activity
26 Neurophysiological evidence from short-term inhalation studies in animals, as well as
27 supportive evidence from acute oral and injection studies, suggests that exposures to TMB isomers
28 at lower concentrations (at least for 1,2,4-TMB) may affect parameters associated with brain
29 excitability. Decreases in a particular component of electrocortical arousal (i.e., spike-wave
30 discharge, SWD, bursts in recordings from cortical-hippocampal electroencephalograms, EEGs)
31 were observed in male rats 120 days after short-term exposure to > 492 mg/m31,2,4-TMB
32 (statistically significant at 1,230 mg/m3), suggesting persistent functional changes in the rat CNS
33 (Gralewicz etal., 1997b). In recordings from rats that were awake, but immobile (not exhibiting
34 pronounced exploratory activity, as determined by EEC morphology), statistically significant
35 decreases in the frequency of SWD episodes were observed at 24 hours following short-term
36 exposure to 492 mg/m31,2,4-TMB (decreases that were not statistically significant were also
37 observed at >492 mg/m31,2,4-TMB at30 and 120 days after exposure) (Gralewicz etal.. 1997b).
38 Complementing these findings, dose-related decreases in the duration and number of SWD
39 bursts (termed high-voltage spindles) were observed at > 240 mg/kg of the TMB isomers
This document is a draft for review purposes only and does not constitute Agency policy.
I -7 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 subsequent to acute oral exposure [Tomas etal., 1999a] (Table 1-2). The stronger and more
2 persistent effects on electrocortical activity followed a pattern of 1,2,3-TMB > 1,3,5-TMB > 1,2,4-
3 TMB [Tomas etal., 1999a]. Similarly, electrophysiological alterations in cortical andhippocampal
4 EEGs were more pronounced following i.p. injection of 1,2,3-TMB, with 1,2,4-TMB and 1,3,5-TMB
5 exerting lesser effects [Tomas etal., 1999c]. Although it is unclear whether these changes affect
6 related processes such as memory and seizure initiation/propagation, the observed EEC
7 abnormalities following inhalation [Gralewicz etal.. 1997b). oral [Tomas etal.. 1999a). and i.p.
8 [Tomas etal., 1999c] exposure to TMB isomers provide supportive evidence of possible acute CNS
9 depression by TMB isomers [Tomas etal.. 1999a: Tomas etal.. 1999c) and indicate persistent (up
10 to 120 days post-exposure) (Gralewicz etal., 1997b) alterations in CNS activity that may reflect an
11 adaptive response to TMB exposure.
12
This document is a draft for review purposes only and does not constitute Agency policy.
1 -8 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 1-1. Evidence pertaining to neurological effects of TMBs in
animals — inhalation exposures
Study Designa'b and Reference
Results
1,2,4-TMB
Pain sensitivity
0, 123, 492, 1,230 mg/m3, (recovery:
1,230 mg/m3 at 2 wks post-exposure)
90 d; Rat, Wistar, male, N = 10
Korsak and Rydzyriski (1996) Table B-29C
0, 492 mg/m3
4 wks; Rat, Wistar, male, N = 11
Gralewicz and Wiaderna (2001), Table B-26
0, 123, 492, 1,230 mg/m3
4 wks; Rat, Wistar, male, N = 15
Gralewicz et al. (1997a), Table B-24
Exposure-dependent increases in paw-lick latency which recovers by 2
wks post-exposure.
Response relative to control: 0, 18, 79*, 95*% (recovery = 12%)
Increased paw-lick latency 24 hrs after intermittent footshockd.
Response relative to control: 0, 191*%
Increases in paw-lick latency 24 hrs after intermittent footshockd.
Response relative to control: 0, 2, 74*, 33*%
Neuromuscular function and coordination
0, 123, 492, 1,230 mg/m3, (recovery:
1,230 mg/m3 at 2 wks post-exposure)
90 d; Rat, Wistar, male, N = 10
Korsak and Rydzyriski (1996), Table B-29
Exposure-dependent increases in rotarod failures at 13 wks which do not
recover by 2 wks post-exposure.
Response relative to control: 0, 10, 20, 40*% (recovery= 30%)
Motor function and/or anxiety
0, 123, 492, 1,230 mg/m3
4 wks; Rat, Wistar, male, N = 15
Lutz et al. (2010), Table B-34
0, 492 mg/m3
4 wks; Rat, Wistar, male, N = 11
Gralewicz and Wiaderna (2001), Table B-26
0, 123, 492, 1,230 mg/m3
4 wks; Rat, Wistar, male, N = 15
Gralewicz et al. (1997a), Table B-24
Increased horizontal locomotion (distance traveled) in an open field.
Response relative to control: 0, 100, 84, 154*%e
No overall change following single or multiple amphetamine injections.
Increased horizontal locomotion in open field tests.
Response relative to control: 0, 62*%
Increased grooming in open field tests at middle concentration; no
change in horizontal locomotion or exploration.
Response relative to control: 0, 82, 147*, 76%
This document is a draft for review purposes only and does not constitute Agency policy.
I -9 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Cognitive function
0, 492 mg/m3
4 wks; Rat, Wistar, male, N = 1
Gralewicz and Wiaderna (2001), Table B-26
0, 123, 492, or 1,230 mg/m3
4 wks; Rat, Wistar, male, N = 15
Gralewicz et al. (1997a), Table B-24
Decreased step down latency in passive avoidance tests and decreased
performance in active avoidance tests; no change in radial maze tests.
Response relative to control: 0, -43*%g ; 0, -60*%h
Decreases in step down latency in passive avoidance tests; no change in
active avoidance or radial maze tests.
Response relative to control: 0, -21, -81*, -49*%g
Electrocortical activity
0, 123, 492, 1,230 mg/m3
4 wks; Rat, Wistar, male, N = 9
Gralewicz et al. (1997b), Table B-25
Decreased spike-wave discharge bursts in EEG recordings' at
exposure; no change in global level of arousal.
Response relative to vehicle control: 0, 13, -35, -55*%
120 d post-
This document is a draft for review purposes only and does not constitute Agency policy.
1-10 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1,2,3-TMB
Pain sensitivity
0, 123, 492, 1,230 mg/m3, (recovery:
1,230 mg/m3 at 2 wks post-exposure)
90 d; Rat, Wistar, male, N = 10
Korsak & Rydzyriski (1996), Table B-29
0, 492 mg/m3
4 wks; Rat, Wistar, male, N = 11
Gralewicz and Wiaderna (2001), Table B-26
0, 123, 492, 1,230 mg/m3
4 wks; Rat, Wistar, male, N = 15
Wiaderna et al. (1998), Table B-41
Exposure-dependent increases in paw-lick latency which recovers by 2
wks post-exposure.
Response relative to control: 0, 22*, 68, 78*% (recovery = 13%)
No change in paw-lick latency.
Increased paw-lick latency 24 hrs after intermittent footshock at middle
concentration^
Response relative to control: 0, -19, 45*, 8%
Neuromuscular function and coordination
0, 123, 492, 1,230 mg/m3, (recovery:
1,230 mg/m3 at 2 wks post-exposure)
90 d; Rat, Wistar, male, N = 10
Korsak and Rydzyriski (1996), Table B-29
Exposure-dependent increases in rotarod failures at 13 wks which do not
recover by 2 wks post-exposure.
Response relative to control: 0, 20, 40*, 70*% (recovery = 50%)
Motor function and/or anxiety
0, 123, 492, 1,230 mg/m3
4 wks; Rat, Wistar, male, N = 15
Lutz et al. (2010), Table B-34
0, 492 mg/m3
4 wks; Rat, Wistar, male, N = 11
Gralewicz and Wiaderna (2001), Table B-26
0, 123, 492, 1,230 mg/m3
4 wks; Rat, Wistar, male, N = 15
Wiaderna et al. (1998), Table B-41
No change in horizontal locomotion (distance traveled) in an open field.
Increased distance traveled in 2 hrs after amphetamine injections:
Response relative to control after single injection: 0, 15, 198*, llle%
Response relative to control after multiple injections: 0, -21, 103*, 41e%
No change in horizontal locomotion in open field tests.
No change in horizontal locomotion, exploration, or grooming in open
field tests.
Cognitive function
0, 492 mg/m3
4 wks; Rat, Wistar, male, N = 1
Gralewicz and Wiaderna (2001), Table B-26
0, 123, 492, or 1,230 mg/m3
4 wks; Rat, Wistar, male, N = 15
Wiaderna et al. (1998), Table B-41
Decreased performance in active avoidance tests; no change in passive
avoidance or radial maze tests.
Response relative to control: 0, -53*%h
Decreases in step down latency in passive avoidance tests and decreased
performance in active avoidance and radial maze tests at middle
concentration and at low concentration, respectively.
Response relative to control: 0, -50*, -62*, -37%g ; 0, -3, -41*, -14%h; 0, -
30*, 16, -l%j
This document is a draft for review purposes only and does not constitute Agency policy.
1-11 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1,3,5-TMB
Pain sensitivity
0, 492 mg/m3
4 wks; Rat, Wistar, male, N = 11
Gralewicz and Wiaderna (2001). Table B-26
0, 123, 492, 1,230 mg/m3
4 wks; Rat, Wistar, male, N = 12
Wiaderna et al. (2002), Table B-42
Increased paw-lick latency 24 hrs after intermittent footshock .
Response relative to control: 0, 250*%
Increased paw-lick latency 24 hrs after intermittent footshock at
concentration .
Response relative to control: 0, -4, 70*, 17%
middle
Motor function and/or anxiety
0, 492 mg/m3
4 wks; Rat, Wistar, male, N = 11
Gralewicz and Wiaderna (2001). Table B-26
Increased horizontal locomotion in open field tests.
Response relative to control: 0, 70*%
Cognitive function
0, 123, 492, 1,230 mg/m3
4 wks; Rat, Wistar, male, N = 12
Wiaderna et al. (2002), Table B-42
0, 492 mg/m3
4 wks; Rat, Wistar, male, N = 11
Gralewicz & Wiaderna (2001), Table B-26
Decreases in step down latency in passive avoidance tests and decreased
performance in active avoidance tests; no change in radial maze tests.
Response relative to control: 0, -48*, -55*, -46*%g ; 0, -40*, -35*, -50*%h
Decreases in step down latency in passive avoidance tests and decreased
performance in active avoidance tests; no change in radial maze tests.
Response relative to control: 0, -57*%e; 0, -70*%h
*Significantly different from controls (p< 0.05).
Note: For studies other than Korsak and Rydzyriski (1996), % change from control calculated from digitized data using
Grab It! XP software.
aRotarod and hot plate tests were administered immediately after termination of exposure or following a 2 week
recovery period by Korsak and Rydzyriski (1996). EEG recordings were acquired prior to exposure and one, 30, or
120 days after exposure by Gralewicz et al. (1997b). Motor behavior in an open field (tested for 30 min) was
assessed 14 days after exposure and re-tested following single and multiple (to induce sensitization) injections with
amphetamine for 120 min by Lutz et al. (2010). For the remaining studies (Wiaderna et al., 2002; Gralewicz and
Wiaderna, 2001; Wiaderna et al., 1998; Gralewicz et al., 1997a): radial maze tests were administered prior to
exposure and on days 14-18 after exposure; open field activity (tested for 5-10 minutes) was assessed prior to
exposure and on day 25 after exposure; passive avoidance was tested on days 35-48 after exposure; hot plate
sensitivity was assessed on days 50 and 51 after exposure; and active avoidance tests were administered on or after
day 54 post-exposure.
bln instances where authors reported exposures in ppm, EPA converted these values to mg/m3. See Appendix B for
conversion factor, and individual study summary tables for ppm values.
"Tables referenced in Study Design and Reference column correspond to study summary tables in Appendix B
dThis effect was only observed 24 hours following intermittent foot shock (reported as L3); no significant effects at any
exposure were observed prior to or immediately following foot shock.
ePrior to injections (tested for 30 min and reported as Block 1); significance indicated in study text only.
'Locomotion was assessed for 120 minutes following single or multiple amphetamine exposures; a 118% increase
relative to controls was reported for 1,230 mg/m3 from 0-30 minutes following a single injection (Session 1, Block 3),
while 492 mg/m3 appeared to prevent amphetamine-induced increases.
8Decreased step down latency in passive avoidance tests at 7 days post footshock.
Increased number of trials to reach avoidance criteria.
'Electroencephalograms (EEGs) were recorded at electrodes implanted in the fronto-parietal cortex and the dorsal
hippocampus (one recording from each region was analyzed for each rat).
'increased perseveration errors at trial day 5.
This document is a draft for review purposes only and does not constitute Agency policy.
1-12 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 1-2. Evidence pertaining to neurological effects of 1,2,3-TMB,
1,2,4-TMB, or 1,3,5-TMB in animals — oral exposures
Study Designa'b and Reference
Results
1,2,4-TMB
Motor function and/or anxiety
0, 960, 1,920, 3,850 mg/kg single oral
gavage
Rat, Wag/Rij, male, N = 10
Tomas et al. (1999b), Table B-39C
Transient increases in locomotor activity in open field tests.
Response at 20 min after exposure relative to pre-injection controls: 0,
34.1, 57.8, 60.6*% (No significant changes were reported for 10, 30, 40
50, 60, or 70 minutes.)
/
Electrocortical activity
0, 240, 960, 3,850 mg/kg, single oral
gavage
Rat \A/acy/Rii mala M fi
Tomas et al. (1999a), Table B-38
Inhibition of the duration and number of high voltage spindle episodes
EEG recordingsd (response relative to vehicle control):
20 min
Duration 0, -72, -58, -83%
Number 0, -26, -44,-62*%
1,2,3-TMB
40 min
0, -80*, -97*, -45%
0, -53*,-88*,-73*%
in
60 min
0, 11, -67, -45%
0, 7, -53*, -22%
Motor function and/or anxiety
0, 960, 1,920, 3,850 mg/kg single oral
gavage
Rat, Wag/Rij, male, N = 10
Tomas et al. (1999b), Table B-39
Transient increases in locomotor activity in open field tests.
Response at 20 min after exposure relative to pre-injection controls: 0,
30.9, 26.5, 56.1*% (also increased 65.6% at 30 min in the highest
exposure group; no other significant changes were noted at 10, 40, 50,
60, or 70 minutes)
Electrocortical activity
0, 960, 3,850 mg/kg, single oral gavage
Rat, Wag/Rij, male, N = 6
Tomas et al. (1999a), Table B-38
Inhibition of the duration and number of high voltage spindle episodes
EEG recordingsd (response relative to vehicle control):
20 min
Duration 0, -86, -97*, -76*%
Number 0, -71*, -86*, -48%
40 min
0, -95, -98*, -97*%
0, -84*,-93*,-86*%
in
60 min
0, -81, -94*, -99*%
0, -70*,-99*,-96*%
This document is a draft for review purposes only and does not constitute Agency policy.
1-13 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1,3,5-TMB
Motor function and/or anxiety
0, 960, 1,920, 3,850 mg/kg single oral
gavage
Rat, Wag/Rij, male, N = 10
Tomas et al. (1999b), Table B-39
Transient increases in locomotor activity in open field tests.
Response at 20 mm after exposure relative to pre-injection controls: 0, 0,
46.7*, 42.4*% (also increased 65-70% at 40-60 min in the highest
exposure group; no other significant changes were noted at 10, 30, or 70
minutes).
Electrocortical activity
0, 240, 960, 3,850 mg/kg, single oral
gavage
Rat, Wag/Rij, male, N = 6
Tomas et al. (1999a), Table B-38
Inhibition of the duration and number of high voltage spindle episodes in
EEG recordingsd (response relative to vehicle control):
Duration
Number
20 min
0, -76*, -79,-86%
0, -57,- 67, -77%
40 min
0, -85*,-97*,-95*%
0,-52*,-93*,-91*%
60 min
0, -66*,-94*,-88*%
0,-49*,-91*, -89*%
*Significantly different from controls (p < 0.05).
Note: % change from control calculated from digitized data using Grab It! XP software.
aLocomotor activity in open field tests and electrocortical arousal were assessed prior to exposure and immediately
after exposure every 10 minutes for up to 70 minutes.
bln instances where authors reported exposures in ppm, EPA converted these values to mg/m3. See Appendix B for
conversion factor, and individual study summary tables for ppm values.
"Tables referenced in Study Design and Reference column correspond to study summary tables in Appendix B
Electroencephalograms (EEGs) were recorded prior to exposure and at 20, 40, and 60 minutes after exposure via
electrodes implanted in the fronto-parietal cortex.
This document is a draft for review purposes only and does not constitute Agency policy.
1-14 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1500
1000
oo
.§. 500
c
o
4J
£
4-1
0)
u
I 0
O
6
O
O
1,2,4-TMB
O
o
o
o
o
o
1s Paw-lick 1s Paw-lick 1s Paw-lick 1s Rotarod
latency latency latency failure
(a) (d) (b) (a)
Sub-
chronic
Short-
term
4- Pain
Sensitivity
Sub-
chronic
Impaired
Neuro-
muscular
Function
Short-
term
1s Motor
Activity and/ or 4- Anxiety
1s N/CAmph.- 1s 1s -J, Passive -J, Passive N/C Active N/C Radial N/C Radial
Horizontal Induced Horizontal Grooming and Active Avoidance Avoidance Maze Maze
Loco- Loco- Loco- Activity Avoidance (b) (b) (b) (d)
motion motion (e) motion (b) (d)
(e) (d)
Short-
term
Altered Cognitive Function
Concentrations
ONOAEL
DLOAEL
o
bursts
(c)
short-
term
Endpoint
Summary
Duration
4- Electro- Effect
cortical Summary
Activity
Solid lines represent range of exposure concentrations, (a) Korsak and Rydzyriski (1996); (b) Gralewicz et al. (1997a); (c) Gralewicz et al. (1997b); (d) Gralewicz
and Wiaderna (2001); (e) Lutz et al. (2010). Exposure concentrations (y-axis) in mg/m3. All effects are in male Wistar rats.
Figure 1-1. Exposure response array of neurological effects following inhalation exposure to 1,2,4-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
1-15 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1500
1000
oo
•^•
o
4J
to
4-1
0)
u
c
o
500
o
o
o
1s Paw-lick N/C Paw- 1s Paw-lick 1s Rotarod
latency
(a)
Sub-
chronic
lick
latency
(c)
latency
(b)
Short-
term
4- Pain
Sensitivity
failure
(a)
Sub-
chronic
Impaired
N euro-
muscular
Function
O
O
N/C
Horizontal
Loco-
motion
(e)
Induced
Loco-
motion (e)
1,2,3-TMB
O
O
N/COpen
Field
Activity
(c)
o
o
o
o
Concentrations
ONOAEL
DLOAEL
Short-
term
N/C Motor
Activity and/ or Anxiety
N/C Open N/C Passive 4- Passive 4- Active 4- Active N/C Radial
Field Avoidance Avoidance Avoidance Avoidance Maze
Activity (c) (d) (c) (d) (c)
(d)
Short-
term
Altered Cognitive Function
4- Radial
Maze
(d)
Endpoint
Summary
Duration
Effect
Summary
Solid lines represent range of exposure concentrations, (a) Korsak and Rydzyriski (1996); (b) Gralewicz et al. (1997a); (c) Gralewicz and Wiaderna (2001); (d)
Wiaderna et al. (1998): (e) Lutz et al. (2010). Exposure concentrations (y-axis) in mg/m3. All effects are in male Wistar rats.
Figure 1-2. Exposure response array of neurological effects following inhalation exposure to 1,2,3-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
1-16 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1500
oo
c
o
4J
to
4-1
0)
u
O
1000
500
0
o
1,3,5-TMB
D
T
O
O
Concentrations
ONOAEL
DLOAEL
• Paw-lick 1s Paw-lick
latency latency
(a) (b)
Short-
term
s|/ Pain
Sensitivity
1s Horizontal -J/ Passive
Loco- Avoidance
motion (a)
(a)
Short-
term
1s Motor
Activity and/
or s|/ Anxiety
s|/ Passive s|/ Active
Avoidance Avoidance
(b) (a)
s|/ Active N/C Radial
Avoidance Maze
(b) (a)
Short-
term
Altered Cognitive Function
N/C Radial Endpoint
Maze Summary
(b)
Duration
Effect
Summary
Solid lines represent range of exposure concentrations, (a) Gralewicz and Wiaderna (2001); (b) Wiaderna et al. (2002). Exposure concentrations (y-axis) in
mg/m3. All effects are in male Wistar rats.
Figure 1-3. Exposure response array of neurological effects following inhalation exposure to 1,3,5-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
1-17 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1,2,4-TMB, 1,2,3-TMB, or 1,3,5-TMB
4500
4000 .-
3500
•553000
^
"So 2500
E
o, 2000 /J
in S
0
0 1500
1000 ^
f T 1
: ,
, :
i (
: ,
500 1
rh H
> (
, i
1 (
: .
h r
Concentrations
ONOAEL
DLOAEL
1
h
l_l l_l LJ
0
t Horizontal ^ HVS Episodes t Horizontal 4> HVS Episodes t Horizontal 4> HVS Episodes Endpoint
Locomotion and Duration Locomotion and Duration Locomotion and Duration Summary
(a) (b) (a) (b) (a) (b)
Acute Acute Acute Acute Acute Acute Duration
T" Motor \|/ Electro-cortical T" Motor \|/ Electro-cortical T" Motor \|/ Electro-cortical
Activity and/ or \|/ Activity Activity and/ or \|/ Activity Activity and/ or \|/ Activity
Anxiety Anxiety Anxiety
1,2,4-TMB
1,2,3-TMB
1,3,5-TMB
Isomer
Solid lines represent range of exposure concentrations, (a) Tomas et al. (1999a); (b) Tomas et al. (1999b). Exposure concentrations (y-axis) in mg/kg. All
effects are in male WAG/Rij (Tomas et al. (1999a)) or Wistar (Tomas et al. (1999b)) rats.
Figure 1-4. Exposure response array of neurological effects following oral exposure to 1,2,3-TMB, 1,2,4-TMB,
or 1,3,5-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
1-18 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Mode of Action Analysis - Neurological Effects
1 The observation of neurotoxicity following acute-, short-term-, and subchronic-duration
3 exposure to 1MB [Lutzetal.. 2010: Lammers etal.. 2007: Wiadernaetal.. 2002: Gralewicz and
4 Wiaderna. 2001: Wiadernaetal.. 1998: Gralewicz etal.. 1997a: Gralewicz etal.. 1997b: Korsakand
5 Rydzynski. 1996: Korsak etal.. 1995] may indicate that TMB perturbs normal neurotransmission in
6 exposed animals, although the specific key events necessary for TMB-induced neurotoxicity are not
7 established. Although limited mechanistic data for TMBs exists, structurally similar compounds
8 like toluene and xylene have been more thoroughly characterized and it is hypothesized that TMBs
9 would operate through a similar mechanism in producing the resultant neurotoxicological effects.
10 Aromatic hydrocarbons are known to interact with catecholaminergic systems [Kyrklund. 1992].
11 Inhalation exposures to toluene and xylene have been shown to significantly change concentration
12 and turnover rate of both dopamine and norepinephrine in various regions of the rat brain [Rea et
13 al.. 1984: Anderssonetal.. 1983: Anderssonetal.. 1981: Anderssonetal.. 19801 These changes
14 have been hypothesized to be due to potential metabolites with affinity to catecholamine receptors
15 that would, in turn, influence the uptake and release of neurotransmitters [Andersson et al.. 1983:
16 Andersson etal.. 1981: Andersson etal.. 1980].
17 Catecholaminergic changes with toluene have been reported and are similar to that
18 observed with TMBs which would therefore increase the plausibility that the mechanisms of
19 neurotoxicity are similar between the two compounds. For example, subchronic inhalation
20 exposures of rats to low concentrations of toluene (as low as 80 ppm [300 mg/m3 ]] have been
21 shown to decrease spatial learning and memory, increase dopamine-mediated locomotor activity,
22 increase the number of dopamine D2 receptors, and increase dopamine D2 agonist receptor
23 binding [Hillefors-Berglundetal.. 1995: von Euler etal.. 1994: vonEuler etal.. 1993]. These effects
24 were observed to persist up to four weeks after the termination of the toluene exposure.
25 Activation of the dopaminergic system may also result in an inability to inhibit locomotor
26 responses normally suppressed by punishment (Tackson and Westlind-Danielsson. 1994]. Direct
27 application of dopamine to the nucleus accumbens of rats has been observed to result in
28 retardation of the acquisition of passive avoidance learning at concentrations that also stimulated
29 locomotor activity [Braes etal., 1984]. Increases in catecholaminergic neurotransmission (through
30 exposure to norepinephrine or dopamine agonists] result in dose-dependent reductions in the
31 duration of spike wave discharges in rats (Snead, 1995: Warter etal., 1988]. These observations
32 and findings are in concordance with those resulting from exposure to TMBs (Wiaderna etal.,
33 2002: Gralewicz and Wiaderna. 2001: Gralewicz etal.. 1997a: Gralewicz etal.. 1997b](Tomas etal..
34 1999a: Tomas etal.. 1999c]. Additionally, with regards to toluene and related aromatic
35 hydrocarbons, it is known that there is direct interaction with these compounds on various ion
36 channels (ligand and voltage gated] that are present in the central nervous system (Bowen etal.,
37 2006: Balster, 1998]. There is not enough information to ascertain the specific molecular sites and
This document is a draft for review purposes only and does not constitute Agency policy.
1-19 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 how the changes correlate to the observed neurotoxicological effects. However, it is widely
2 believed that the interactions with the neuronal receptors in the brain (e.g., ion channels,
3 catecholaminergic systems) may influence these changes.
4 Aromatic hydrocarbons may also affect the phospholipids in the nerve cell membrane
5 [Anderssonetal.. 1981]. Pertubation of the phospholipids on the cell membrane could indirectly
6 affect the binding of neurotransmitters to the catecholamine or other receptors and potentially lead
7 to alterations in receptor activity or uptake-release mechanisms. Uneven distribution of
8 metabolites within differing regions of the brain, or spatial variations in phospholipid composition
9 of nerve cell membranes, may explain the differential effects seen in regard to catecholamine levels
10 and turnover [Anderssonetal.. 1981]. Based on effect levels with other related solvents (e.g.,
11 toluene - see Balster (1998]], it is hypothesized that with TMBs there may be an initial interaction
12 with the neuronal receptors (e.g., catecholaminergic systems, ion channels] followed by, at much
13 higher exposures, interaction with the lipid membrane when the available sites on the neuronal
14 receptors are completely occupied.
15 Additional mechanisms that may play a role in 1MB neurotoxicity include production of
16 reactive oxygen species (ROS]. Myhre et al. (2000] observed increased respiratory burst in
17 neutrophils after 1,2,4-TMB exposure demonstrated by fluorescence spectroscopy, hydroxylation of
18 4-hydroxybenzoic acid, and electron paramagnetic resonance spectroscopy. The authors suggest
19 that the observation of solvent-induced ROS production may be relevant to brain injury, as
20 microglia cells have a respiratory burst similar to neutrophils. Stronger evidence of potential ROS-
21 related mechanisms of neurotoxicity was observed in a related study by Myhre and Fonnum (2001]
22 in which rat neural synaptosomes exposed to 1,2,4-TMB produced a dose-dependent increase in
23 reactive oxygen and nitrogen species demonstrated by the formation of the fluorescence of 2'7'-
24 dichlorofluorescein. This observation of ROS production in rat synaptosomes may potentially
25 explain the observed TMB-induced neurotoxicity in acute, short-term, and subchronic inhalation
26 studies.
27 Summary of Neurological Effects
28 Neurotoxicity is associated with exposure to TMBs based on evidence in humans and
29 animals. All three TMB isomers are taken up in humans (Tarnberg et al., 1998,1997a: Tarnberg et
30 al.. 1996}. and occupational studies involving exposure to TMBs and other VOCs show
31 neuropsychological effects (Chenetal., 1999], deficits in short term memory and reduced motor
32 speed/coordination (Lee etal., 2005], abnormal fatigue (Norsethetal., 1991], and nervousness,
33 anxiety, and/or vertigo (Battig et al. (1956], as reviewed by MOE (2006] and Baettig et al. (1958]].
34 These effects, however, cannot be attributed to any specific compound. None of the available
35 studies have addressed the potential for latent neurological effects or effects in sensitive
36 populations.
This document is a draft for review purposes only and does not constitute Agency policy.
1 -20 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 There is strong, consistent evidence of neurotoxicity in male Wistar rats exposed to any
2 1MB isomer via inhalation across multiple concentrations and multiple durations, although the
3 studies were conducted at the same institute [Wiadernaetal., 2002: Gralewicz and Wiaderna,
4 2001: Wiadernaetal.. 1998: Gralewicz etal.. 1997a: Gralewicz etal.. 1997b: KorsakandRydzynski.
5 1996: Korsak etal.. 1995]. By gavage, similar effects were observed (e.g., EEC; open field) [Tomas
6 etal.. 1999a: Tomas et al.. 1999b). although testing by this route was not as extensive as by
7 inhalation.
8 Most of the neurotoxicity tests in animals incorporated the application of footshock which,
9 depending on the procedure, can involve multiple contributing factors and can complicate
10 interpretations regarding effects on discrete neurological function. The spectrum of effects
11 suggests that TMBs affect multiple, possibly overlapping, CNS systems rather than a single brain
12 region or neuronal nuclei (suggested by the solvent activity of the compounds). Almost all tests
13 (other than pain) involve a contributing component of motor system function. Some endpoints
14 exhibited clear exposure-response relationships (e.g., pain sensitivity and rotarod), although the
15 pain sensitivity was not consistent across studies with different experimental design (i.e., varying
16 exposure durations and timing of endpoint analyses). Other endpoints did not show a clear
17 concentration-effect relationship. In summary, the evidence supports a determination that TMBs
18 are neurotoxic following inhalation or oral exposure, based on consistency and coherency of effects
19 in animals and humans, biological plausibility, and observed exposure-response relationships in
20 animals.
21 1.1.2. Respiratory Effects
22 There is evidence in humans and animals that inhalation exposure to TMBs induces
23 respiratory toxicity. The human evidence comes from occupational and residential studies
24 involving complex VOC mixtures that include TMBs; thus, effects cannot be attributed to any TMB
25 isomer specifically. TMB isomers are associated with increased measures of respiratory irritation,
26 such as laryngeal and/or pharyngeal irritation (Norseth etal.. 1991) and asthmatic bronchitis
27 ffBattig etal.. 19561. as reviewed in MOE (2006) and Baettig et al. (195811 following occupational
28 exposures. Residential exposures have demonstrated significant associations between 1,2,4-TMB
29 and asthma (Billionnetetal.. 2011). Controlled human exposures (Tones etal., 2006: Tarnbergetal.,
30 1997a: Jarnbergetal.. 1996) have failed to observe substantial irritative symptoms following acute
31 (less than 4 hours) inhalation exposures to TMB isomers of up to 25 ppm (123 mg/m3).
32 In animals, there is consistent evidence of respiratory toxicity following inhalation exposure
33 of rodents to the TMB isomers (Table 1-3). Markers of inflammation and irritation in the lungs of
34 rats have been observed following subchronic inhalation exposures of Wistar rats to 1,2,4-TMB or
35 1,2,3-TMB. Increases in immune and inflammatory cells in bronchoalveolar lavage (BAL) fluid have
36 been observed following subchronic exposures of male Wistar rats to 1,2,4-TMB at concentrations
37 > 123 mg/m3 (Korsak etal., 1997). Specifically, the number of cells in the BAL fluid of exposed rats
This document is a draft for review purposes only and does not constitute Agency policy.
I -21 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 was increased for both total cells (> 123 mg/m3) and macrophages (> 492 mg/m3). However, some
2 attenuation of these effects was observed at high concentrations (i.e., at 1,230 mg/m3) compared to
3 lower concentrations. For example, the number of macrophages was increased 2.7-fold relative to
4 control at 492 mg/m3, but only 2.2-fold at 1,230 mg/m3. This may indicate either adaptation to the
5 respiratory irritation effects of 1,2,4-TMB, saturation of metabolic pathways, or immune
6 suppression at higher doses. Subchronic exposure of male Wistar rats also significantly increased
7 the BAL fluid content of polymorphonuclear leukocytes and lymphocytes; however the specific
8 concentrations eliciting these significant increases were not reported by study authors. A small, but
9 not significant, decrease in cell viability (all cells) was observed following subchronic exposure to
10 1,2,4-TMB at > 123 mg/m3 fKorsaketal.. 19971.
11 In addition to increases in immune and inflammatory cells in BAL fluid following exposure
12 to 1,2,4-TMB, histopathological alterations characterized by increases in lymphatic tissue in the
13 lower respiratory tract have also been observed following subchronic exposures of male and female
14 Wistar rats to 1,2,4-TMB or 1,2,3-TMB fKorsaketal.. 2000a. b). Significant proliferation of
15 peribronchial lymphatic tissue was observed in male rats exposed to 123 mg/m31,2,3-TMB or 492
16 mg/m31,2,4-TMB and female rats exposed to 123 and 492 mg/m31,2,3-TMB, although trend
17 analysis demonstrated that these increases were not concentration-dependent. Non-concentration
18 dependent increases in interstitial lymphocytic infiltrations were also observed in male rats
19 exposed to 492 mg/m31,2,4-TMB. However, statistically significant increases in interstitial
20 lymphocytic infiltrations observed in male and female rats exposed to 1,230 mg/m31,2,3-TMB or
21 1,2,4-TMB, respectively, were concentration-dependent based on trend analysis.
22 In some 1,2,4-TMB or 1,2,3-TMB-exposed rats exhibiting peribronchial lymphatic
23 proliferation, the bronchial epithelium lost its cuboidal shape and formed lymphoepithelium.
24 However, this formation of lymphoepithelium was apparently non-mono tonic and not dependent
25 on concentration. Alveolar macrophages were increased in both sexes exposed to 1,230 mg/m3
26 1,2,4-TMB (significant only for males), with trend analysis demonstrating concentration-
27 dependence across the entire concentration range. Goblet cells were statistically significantly
28 increased in a concentration-dependent manner in female rats exposed to > 492 mg/m31,2,3-TMB.
29 When the incidences of all pulmonary lesions were analyzed in aggregate, lesions were significantly
30 increased in males at 492 mg/m31,2,4-TMB, but not at any concentration in females. However,
31 trend-analysis demonstrated significant increases in aggregate pulmonary lesions in both sexes
32 across the entire concentration range. In rats exposed to 1,2,3-TMB, the aggregate incidences of
33 pulmonary lesions were not statistically significantly increased at any single concentration in males
34 or females. Male rats, however, did exhibit a concentration-dependent increase in aggregate lesions
35 according to trend analysis. Studies on the respiratory effects of subchronic exposures to
36 1,3,5-TMB were not available.
37 Additional effects on clinical chemistry including increased total protein (37% increase at
38 exposures of both 123 and 492 mg/m3), decreased mucoprotein (13% decrease at 123 mg/m3
This document is a draft for review purposes only and does not constitute Agency policy.
I -22 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 exposure), increased lactate dehydrogenase (170% and 79% increase at 123 and 492 mg/m3,
2 respectively) and increased acid phosphatase activity (47-75% increase at > 123 mg/m3) were
3 observed in animals exposed to 1,2,4-TMB, suggesting pulmonary irritation or inflammation. All of
4 these effects also exhibited either some attenuation of effect at high concentrations compared to
5 lower concentrations. Therefore, some adaptation to the respiratory irritation effects of 1,2,4-TMB
6 may be occurring.
7 Decreased respiration, a symptom of sensory irritation, has been observed in male BALB/C
8 mice during acute inhalation exposures to the TMB isomers for 6 minutes. These acute exposures
9 were observed to result in dose-dependent depression of respiratory rates, with the maximum
10 decrease in respiration occurring in the first 1 or 2 minutes of exposure (Korsaketal., 1997: Korsak
11 etal.. 1995). The concentration of 1,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB that was observed to result
12 in a 50% depression in the respiratory rate (RDso) was similar between the three isomers: 578, 541,
13 or 519 ppm (2,844, 2,662, or 2,553 mg/m3), respectively.
14
This document is a draft for review purposes only and does not constitute Agency policy.
1 -23 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 1-3. Evidence pertaining to respiratory effects of TMBs in
animals — inhalation exposures
Study design3 and reference
Results
1,2,4-TMB
Pulmonary inflammation/irritation
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d,
5 d/wk)
Rat, Wistar, male, N = 6-7
Korsak et al. (1997), Table B-30b
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d,
5 d/wk)
Rat, Wistar, male, N = 6-7
Korsak et al. (1997) , Table B-30
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d,
5 d/wk)
Rat, Wistar, male and female, N = 10
Korsak et al. (2000a). Table B-31
Increased total bronchoalveolar cell count with evidence of attenuation
at high exposure.
Response relative to control: 0, 202***, 208**, 131*%
Increased macrophage count with evidence of attenuation at high
exposure.
Response relative to control: 0, 107, 170**, 116**%
Increase in number of pulmonary lesions.
Response relative to control: Incidences not reported, thus calculation of
response relative to control not possible; authors report statistically
significant increases at 492 and 1,230 mg/m3.
Clinical chemistry effect
0, 123, 492, 1,230 mg/m3, 90 ds (6 hr/d,
5 d/wk)
Rat, Wistar, male, N = 10
Korsak et al. (1997), Table B-30
Increased acid phosphatase activity with evidence of attenuation at
exposure.
Response relative to control: 0, 47*, 74*, 45*%
high
Sensory irritation (decreased respiration)
1,245, 3,178, 5,186, 6,391, 9,486 mg/m3,
6 min
Mouse, BALB/C, male, N = 8-10
Korsak et al. (1997); Korsak et al. (1995),
Tables B-30 and B-28
Decreased respiratory rate as measured during first minute of exposure.
Response relative to control: RD50 = 2,844
This document is a draft for review purposes only and does not constitute Agency policy.
I -24 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1,2,3-TMB
Pulmonary inflammation/irritation
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d,
5 d/wk)
Rat, Wistar, male and female, N = 10
Korsak et al. (2000b), Table B-32
Increase in number of pulmonary lesions.
Response relative to control: Incidences not reported, thus calculation
of response relative to control not possible; authors report statistically
significant increases at 492 and 1,230 mg/m3.
Sensory irritation (decreased respiration)
1,255, 2,514, 4,143, 7,828 mg/m3, 6 min
Mouse, BALB/C, male, N = 8-10
Korsak et al. (1997); Tables B-30
Decreased respiratory rate as measured during first minute
exposure.
Response relative to control. RD50 - 2,662
of
1,3,5-TMB
Sensory irritation (decreased respiration)
1,348, 2,160, 2,716, 3,597, 4,900 mg/m3,
6 min
Mouse, BALB/C, male, N = 8-10
Korsak et al. (1997), Table B-30
Decreased respiratory rate as measured during first minute
exposure.
Response relative to control: RD50 = 2,553
of
* , **, *** Statistically different from controls at p < 0.05, p < 0.01, and p < 0.001, respectively.
aln instances where authors reported exposures in ppm, EPA converted these values to mg/m3. See Appendix B for
conversion factor, and individual study summary tables for ppm values.
bTables referenced in Study Design and Reference column correspond to study summary tables in Appendix B
This document is a draft for review purposes only and does not constitute Agency policy.
I -25 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
10000
oo
~c
o
4J
£
4-1
0)
u
c
o
1000
100
1s BAL Total
Cell Count
(a)
O
1s BAL
Macrophage
Count
(a)
Subchronic
Pulmonary
Inflammation/
Irritation
O
T" Pulmonary
Inflammatory
Lesions
(b)
1s Acid
Phosphatase
Activity
(a)
Subchronic
Clinical
Chemistry
Effects
\|/ Respiratory
Rate
(a,c)
Acute
Sensory
Irritation
O
T" Pulmonary
Inflammatory
Lesions
(d)
Subchronic
Pulmonary
Inflammation/
Irritation
Concentrations
ONOAEL
DLOAEL
ARD50
\|/ Respiratory
Rate
(a)
Acute
Sensory
Irritation
1,2,4-TMB
1,2,3-TMB
\|/ Respiratory
Rate
Acute
Sensory
Irritation
1,3,5-TMB
Endpoint
Summary
Duration
Effect
Summary
Isomer
Solid lines represent range of exposure concentrations, (a) Korsak et al. (1997): (b) Korsak et al. (2000a); (c) Korsak et al. (1995): (d) Korsak (2000b).
Concentrations (y-axis) in mg/m3; y-axis is displayed on a logarithmic scale. All subchronic effects are in male Wistar rats, except for increased pulmonary
lesions, which occur in both male and female Wistar rats; acute effects are in Balb/C mice.
Figure 1-5. Exposure response array of respiratory effects following inhalation exposure to 1,2,3-TMB, 1,2,4-
TMB, or 1,3,5-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
1 -26 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Mode of Action Analysis - Respiratory Effects
1 Data regarding the potential mode of action for the respiratory effects resulting from TMB
3 inhalation exposures are limited and the key events for TMB-induced respiratory toxicity are not
4 established. However, the available toxicity data suggest that TMB isomers act as potent acute
5 respiratory irritants and induce inflammatory responses following longer exposures (i.e.,
6 subchronic) in animals. Korsak et al. [1995] and Korsak et al. [1997] have suggested that
7 decreased respiratory rate following TMB inhalation exposure is indicative of irritation, and
8 proposed that respiratory irritants such as TMB may activate a "sensory irritant receptor" on the
9 trigeminal nerve ending in the nasal mucosa leading to an inflammatory response. Korsak et al.
10 [1997: 1995] further suggested that activation of this irritant receptor follows either adsorption of
11 the agonist, or adsorption and chemical reaction with the receptor. The authors referenced a
12 proposed model for the receptor protein that includes two main binding sites for benzene moieties
13 and a thiol group. Further, they suggested that in the case of organic solvents (i.e., toluene, xylene,
14 and TMB], a correlation between the potency of the irritating effect and the number of methyl
15 groups is likely given the observation that RDso values for depressed respiratory rates following
16 exposure to TMB isomers is approximately 8-fold lower than toluene and 4-fold lower than xylene.
17 Following subchronic inhalation exposure of rats to 1,2,4-TMB, inflammatory cell (i.e.,
18 macrophages, polymorphonuclear leukocytes, and lymphocytes] numbers were increased along
19 with markers of their activation (i.e., total lactate dehydrogenase and acid phosphatase activity in
20 BAL] [Korsak et al.. 1997]. further indicating the inflammatory nature of responses in the
21 respiratory tract of TMB-exposed animals. Inflammatory pulmonary lesions were also observed
22 following subchronic inhalation exposures in rats. However, many of these effects were not
23 observed to be concentration-dependent in repeat exposure studies (i.e., no progression of effect
24 over an order of magnitude of concentrations], suggesting that there may be adaptation to
25 respiratory irritation that occurs following extended inhalation exposure to TMB. The processes
26 responsible for the respiratory inflammatory responses observed in subchronically exposed
27 animals are unknown. However, a major inflammatory mediator, interleukin 8 [IL-8], was
28 increased following exposure of porcine and human macrophages to secondary organic aerosol
29 (SOA] particles derived from 1,3,5-TMB [Gaschenetal.. 2010]. The observation that IL-8 levels
30 increase following exposure to 1,3,5-TMB-derived SOA is noteworthy as a major function of IL-8 is
31 to recruit immune cells to sites of inflammation. Therefore, the observation of inflammatory
32 lesions involving immune cells (i.e., macrophages and leukocytes] may be partially explained by
33 increases in inflammatory cytokines following TMB exposures. Additionally, ROS-generation has
34 been observed in cultured neutrophil granulocytes and rat neural synaptosomes exposed to TMB
35 [Myhre and Fonnum, 2001: Myhre etal., 2000], and the related compounds benzene and toluene
36 have been shown to induce oxidative stress in cultured lung cells [Mogel etal., 2011]. Although
37 pulmonary ROS-generation has not been observed following in vivo or in vitro TMB exposures,
This document is a draft for review purposes only and does not constitute Agency policy.
I -27 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 there is suggestive evidence that it could play a role in the irritative and inflammatory responses
2 seen in exposed animals.
3 In a study investigating jet fuel-induced cytotoxicity in human epidermal keratinocytes
4 (HEK), aromatic hydrocarbons were more potent inducers of cell death than aliphatic constituents,
5 even though the aromatic compounds only accounted for less than one-fourth of aliphatic
6 constituents [Chou etal.. 2003). Of the single aromatic ring hydrocarbons, 1,2,4-TMB and xylene
7 were the most lethal to HEK. Increased cytotoxicity may explain the small, but insignificant,
8 decrease in BAL cell viability observed in Korsak et al. [1997].
9 Summary of Respiratory Effects
10 Respiratory toxicity is associated with inhalation exposure to TMBs based on evidence in
11 humans and animals. All three TMB isomers are taken up by humans (Jarnbergetal., 1998,1997a:
12 Tarnbergetal.. 1996}. and occupational and residential studies involving exposure to TMBs and
13 other VOCs suggest an association between TMB exposure and asthmatic symptoms [Billionnet et
14 al., 2011: Battigetal., 1956] and sensory irritation [Norsethetal., 1991]. These effects, however,
15 cannot be attributed to any specific compound.
16 There is strong, consistent evidence of respiratory toxicity in male and female Wistar rats
17 exposed to any TMB isomer via inhalation across multiple concentrations and multiple durations,
18 although the studies were conducted at the same institute [Korsak etal., 2000a, bj Korsak etal.,
19 1997: Korsak etal.. 1995]. Some endpoints (i.e., BAL macrophages and alkaline phosphatase]
20 showed concentration-dependence at low- and mid-exposures, all effects were observed to exhibit
21 some attenuation of effect at high doses, potentially indicating either adaptation to the respiratory
22 irritation effects, saturation of metabolic and/or toxicity pathways, or immune suppression at
23 higher doses. In summary, the evidence supports a determination that TMBs are respiratory
24 toxicants following inhalation exposure, based on consistency and coherency of effects observed in
25 humans and animals, biological plausibility, and observed exposure-response relationships.
26 1.1.3. Reproductive and Developmental Effects
27 There are no studies in humans that investigated the reproductive or maternal toxicity of
28 the TMB isomers by any route of exposure. Maternal toxicity in the form of decreased corrected
29 body weight (i.e., maternal body weight minus the weight of the gravid uterus] was observed in
30 Sprague-Dawley rat dams following inhalation exposure during gestation to 1,2,4-TMB or 1,3,5-
31 TMB (Saillenfait etal.. 2005]. Dams exposed to 2,952 mg/m31,2,4-TMB gained only 50% of the
32 weight gained by control animals, whereas dams exposed to 2,952 mg/m31,3,5-TMB gained only
33 25% of the weight gained by controls. Decreased maternal food consumption (across GD 6-21] was
34 also observed at > 2,952 mg/m31,2,4-TMB and > 1,476 mg/m31,3,5-TMB, although the magnitude
35 of the difference compared to controls (88-83% and 92-75% of controls, respectively] was modest
36 relative to the observed decreases in maternal weight gain. The decrease in food consumption at
This document is a draft for review purposes only and does not constitute Agency policy.
1 -28 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 1,476 mg/m31,3,5-TMB (92% relative to controls) was not considered to be a marker of adversity
2 given no accompanying decrease in maternal weight gain was observed at that concentration.
3 There are no studies in humans that investigated the developmental toxicity of either
4 1,2,4-TMB or 1,3,5-TMB by any route of exposure. Developmental toxicity (reported as decreased
5 fetal body weight) has been observed in male and female rats following gestational exposure to
6 1,2,4-TMB and 1,3,5-TMB on gestational days 6 through 20 via inhalation for 6 hours a day
7 (Saillenfaitetal.. 2005) (Table 1-4). Fetal body weights were decreased (statistically significantly)
8 by 5-13% at concentrations of > 2,952 mg/m3 of 1,2,4-TMB and 1,3,5-TMB. No adverse effects
9 were noted on embryo/fetal viability and no increase in skeletal, visceral, or external morphology
10 (i.e., teratogenesis) was observed up to the highest concentrations for either isomer. Studies on the
11 developmental or reproductive effects of 1,2,3-TMBby any route of exposure were not available.
This document is a draft for review purposes only and does not constitute Agency policy.
I -29 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 1-4. Evidence pertaining to reproductive and developmental
effects of 1,2,4-TMB and 1,3,5-TMB in animals — inhalation exposures
Study Design3 and Reference
Results
1,2,4-TMB
Developmental toxicity
0, 492, 1,476, 2,952, 4,428 mg/m3, GD 6-
20 (6 hr/d)
Rat, Sprague-Dawley, female and male0
Saillenfait et al. (2005), Table B-37b
Decreased fetal body weight of male and female fetuses.
Response relative to control:
Male: 0, -1, -2, -5*, -11**%
Female: 0, -1, -3, -5*, -12**%
Maternal toxicity
0, 492, 1,476, 2,952, 4,428 mg/m3, GD 6-
20 (6 hr/d)
Rat, Sprague-Dawley, female, N = 24-25
dams
Saillenfait et al. (2005), Table B-37
Decreased corrected maternal weight gain.
Response relative to control: 0, +7, -7, -51**, -100**%
(weight gain = Og)
1,3,5-TMB
Developmental toxicity
0, 492, 1,476, 2,952, 5,904 mg/m3, GD 6-
20 (6 hr/d)
Rat, Sprague-Dawley, female and malea'c
Saillenfait et al. (2005), Table B-37
Decreased fetal body weight of male and female.
Response relative to control:
Male: 0, -1, -5, -7*, -12**%
Female: 0, -1, -4, -6, -13**%
Maternal Toxicity
0, 492, 1,476, 2,952, 5,904 mg/m3, GD 6-
20 (6 hr/d)
Rat, Sprague-Dawley, female, N = 24-25
dams
Saillenfait et al. (2005), Table B-37
Decreased corrected maternal weight gain.
Response relative to control: 0, +3, -31,- 76**, -159
(weight gain =-12 g)
**%
* , ** Statistically significantly different from controls at p < 0.05 and p < 0.01, respectively.
aln instances where authors reported exposures in ppm, EPA converted these values to mg/m3. See Appendix B for
conversion factor, and individual study summary tables for ppm values.
bTables referenced in Study Design and Reference column correspond to study summary tables in Appendix B
cl\lumber of fetuses analyzed not reported.
This document is a draft for review purposes only and does not constitute Agency policy.
1-30 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
6000
5000
4000
oo
-=• 3000
c
s
g 2000
u
c
o
Concentrations
ONOAEL
DLOAEL
1000
Males
Females
Males
Females
v|/ Fetal
Body Weight
Developmental
Toxicity
\|/ Maternal
Body Weight
Gain
Maternal
Toxicity
v|/ Fetal
Body Weight
Developmental
Toxicity
\|/ Maternal
Body Weight
Gain
Maternal
Toxicity
1,2,4-TMB
1,3,5-TMB
Sex
Endpoint
Summary
Effect
Summary
Isomer
Solid lines represent range of exposure concentrations. All effects from Saillenfait et al. (2005). Concentrations (y-axis) in mg/m .
Figure 1-6. Exposure response array of reproductive and developmental effects following inhalation exposure
to 1,2,4-TMB or 1,3,5-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
1 -31 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Summary of Reproductive and Developmental Effects
1 The database for reproductive and developmental toxicity following inhalation exposure to
3 1,2,4-TMB and 1,3,5-TMB is limited to one animal developmental study; no studies in humans are
4 available. Thus, these isomers may cause developmental toxicity, although this is based on only one
5 study that demonstrated clear, exposure-related effects on fetal and maternal body weights.
6 1.1.4. Hematological and Clinical Chemistry Effects
7 There is limited evidence in humans, and stronger evidence in animals, that exposure to
8 TMB isomers via inhalation induces hematological toxicity. Alterations in blood clotting and
9 anemia in workers exposed to a paint solvent containing 50% 1,2,4-TMB, 30% 1,3,5-TMB, and
10 unspecified amounts of 1,2,3-TMB (listed as possibly present) was reported by Battig et al. [1956],
11 as reviewed by MOE [2006]: effects observed at 295 mg/m3. However, as workers were exposed to
12 a solvent mixture containing multiple TMB isomers and other VOCs, effects cannot be attributed to
13 any TMB isomer specifically.
14 In animals, there is evidence of hematological toxicity following subchronic inhalation
15 exposure to 1,2,4-TMB or 1,2,3-TMB and short-term inhalation exposure to 1,3,5-TMB (Table 1-5].
16 Subchronic exposures to 1,2,4-TMB or 1,2,3-TMB have been shown to result in hematological
17 effects and changes in serum chemistry in rats [Korsak et al., 2000a, b]. In male rats exposed to
18 1,230 mg/m31,2,4-TMB or 1,2,3-TMB, red blood cells (RBC] counts were significantly decreased 23
19 and 15%, respectively. The observed alterations in RBCs were concentration-dependent as
20 determined by trend analysis. Exposure to 1,2,4-TMB or 1,2,3-TMB did not significantly decrease
21 RBCs in female rats, but trend analysis demonstrated that decreases in RBC counts in female rats
22 exposed to 1,2,3-TMB were concentration dependent, with a maximum decrease of 9% at 1,230
23 mg/m3. RBCs in both sexes were observed to still be depressed relative to controls 2 weeks
24 following termination of exposure to both isomers, but these decreases were not statistically
25 significant.
26 White blood cell [WBC] counts were significantly increased 80% in male rats and increased
27 30% (not statistically significant] in female rats exposed to 1,230 mg/m31,2,4-TMB. After a two-
28 week follow-up after termination of exposure, WBC counts had returned to normal in female rats
29 and were slightly depressed [18%] in male rats. WBC numbers were unchanged in male rats
30 exposed to 1,2,3-TMB, but were increased (not statistically significant] 22% in female rats exposed
31 to 1,230 mg/m3. After two weeks following termination of exposure, WBC counts in male and
32 female rats had fallen to roughly 60% of controls.
33 Significant decreases in reticulocytes (71% decrease relative to controls] and clotting time
34 (37% decrease relative to controls] were observed in female rats exposed to 1,230 mg/m3 and 492
35 mg/m31,2,4-TMB, respectively. Both of these effects were concentration-dependent across the
36 entire-range of concentrations as determined by trend-analysis; animals fully recovered within 2
This document is a draft for review purposes only and does not constitute Agency policy.
1-32 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 weeks after termination of exposure. Reticulocyte numbers were statistically significantly
2 increased 60% in male rats exposed to 1,230 mg/m31,2,3-TMB, with reticulocyte numbers even
3 further increased (150%) two weeks following the termination of exposure. Reticulocyte numbers
4 in females exposed to 1,2,3-TMB were significantly increased 77% and 100% at 123 and 492
5 mg/m3, and increased 69% (not statistically significant) at 1,230 mg/m3. Reticulocyte numbers
6 were still increased in males and females 2 weeks after the termination of exposure to 1,2,3-TMB.
7 Segmented neutrophils were statistically significantly decreased 29% in male rats exposed to 1,230
8 mg/m31,2,3-TMB; statistically significant decreases of 29% and 48% were observed in female rats
9 exposed to 492 and 1,230 mg/m31,2,3-TMB. Lymphocytes were statistically increased 11% and
10 15% in male and female rats exposed to 1,230 mg/m3, respectively. Numbers of segmented
11 neutrophils and lymphocytes returned to control values 2 weeks after termination of exposure.
12 Sorbitol dehydrogenase was increased at > 123 mg/m3 in male rats exposed to 1,2,4-TMB
13 (18-23% relative to controls) and at 1,230 mg/m3 in male rats exposed to 1,2,3-TMB (69% relative
14 to controls)(Korsak et al.. 2000a. b). However, the increases following exposure to 1,2,4-TMB were
15 not concentration-dependent. Sorbitol dehydrogenase activity was also higher in female rats
16 exposed to 1,2,4-TMB (19-23% relative to controls) butthe increases in activity were not
17 significantly higher when compared to controls. Sorbitol dehydrogenase activity was not affected
18 in female rats exposed to 1,2,3-TMB. Alanine aminotransferase was decreased (23% relative to
19 controls) and alkaline phosphatase was increased (42-45% relative to controls) at 1,230 mg/m3
20 and > 492 mg/m3 (respectively) in female rats exposed to 1,2,3-TMB.
21 An increase (30% relative to controls) in aspartate aminotransferase, but no other
22 substantial hematological effects, was observed in rats 14 days following short-term exposure (6
23 hours/day, 6 days/week for 5 weeks) (Wiglusz etal., 1975a: Wiglusz etal., 1975b). The adversity of
24 aspartate aminotransferase is unclear given the lack of a clear pattern in temporality (effects at
25 some days post-exposure, but not others) and the lack of accompanying liver histopathology.
26 Acute inhalation exposures of male Wistar rats to 1,500-6,000 mg/m31,3,5-TMB for 6
27 hours did not result in substantial effects on hemoglobin or RBC or WBC count (Wiglusz etal..
28 1975a). However, the number of segmented neutrophilic granulocytes was increased in 1,3,5-TMB-
29 exposed rats up to 28 days following exposure (statistics not reported). The greatest increase in
30 granulocyte numbers (100%) was observed the day of exposure and 1 day following in rats
31 exposed to 6,000 mg/m3, although attenuation was seen 7-28 days following exposure, possibly
32 indicating induction of metabolizing enzymes or saturation of toxicity pathways. Investigation of
33 clinical chemistry parameters in rats acutely exposed to 300-3,000 mg/m3 for 6 hours did not
34 reveal any consistent pattern in the levels of aspartate or alanine aminotransferases, although
35 alkaline phosphatase was statistically increased 84% in rats 7 days following exposure to 3,000
36 mg/m3 (Wiglusz etal.. 1975b).
37
38
This document is a draft for review purposes only and does not constitute Agency policy.
1-33 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 1-5. Evidence pertaining to hematological and clinical chemistry
effects of 1,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB in animals — inhalation
exposures
Study Design3 and Reference
Results
1,2,4-TMB
Hematological toxicity
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000a), Table B-31b
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000a), Table B-31
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000a), Table B-31
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000a), Table B-31
Decreased red blood cells in males only.
Response relative to control: 0, 1, 15, 23**%
(recovery - 24% decrease)
Increased white blood cells in males only.
Response relative to control: 0, 2, 4, 80**%
(recovery - 18% decrease)
Decreased reticulocytes in females only.
Response relative to control: 0, 51, 49, 71*%
(recovery - 65% increase)
Decreases in clotting time in females only.
Response relative to control: 0, 23, 37**, 27*%
(recovery - 60% increase)
Clinical chemistry effects
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000a), Table B-31
Non-monotonic increases in sorbitol dehydrogenase in males
only.
Response relative to control: 0, 73**, 74*,73**%
This document is a draft for review purposes only and does not constitute Agency policy.
1-34 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1,2,3-TMB
Hematological toxicity
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000b), Table B-32
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000b), Table B-32
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000b), Table B-32
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000b), Table B-32
Decreased red blood cells in males only.
Response relative to control: 0, 8, 6, -15*%
(recovery - 9% decrease)
Decreased segmented neutrophils in males and females.
Response relative to control:
Males: 0, 2, -17, -29*% (recovery = 11% increase)
Females: 0, -15, -29*, -48*% (recovery = 15% decrease)
Increased lymphocytes in males and females.
Response relative to control:
Males: 0, 1, 6, 11**% (recovery = 11% decrease)
Females: 0, 6, 10, 15**% (recovery = 3% increase)
Increased reticulocytes in males and females (non-monotonic).
Response relative to control:
Males: 0, -25, 36, 61*% (recovery = 146**% increase)
Females: 0, 77*, 100**, 69% (recovery = 162**% increase)
Clinical chemistry effects
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000b), Table B-32
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000b), Table B-32
0, 123, 492, 1,230 mg/m3, 90 d (6 hr/d, 5 d/wk)
Rat, Wistar, female and male, N = 10
Korsak et al. (2000b), Table B-32
Decreased alanine aminotransferase in females only.
Response relative to control: 0, -1, -6, -23*%
Increased alkaline phosphatase in females only.
Response relative to control: 0, 20, 45*, 42*%
Increased sorbitol dehydrogenase in males only.
Response relative to control: 0, 44, 56, 69*%
This document is a draft for review purposes only and does not constitute Agency policy.
I -3 5 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1,3,5-TMB
Hematological toxicity
1,500-6,000 mg/m3, 6 hr
Samples collected 0, 1, 7, 14, and 28 d post exposure
Rat, Wistar, male, N = 5.8
Wiglusz et al. (1975a), Table B-43
Increased segmented neutrophilicgranulocytes (1-28 d post-
exposure).
Response relative to control: Increased across all days of
exposure.
Clinical chemistry effects
3,000 mg/m3, 5 weeks (6 hr/day, 6 d/wk)
Samples collected 1, 3, 7, 14, and 28 d during
exposure
Rat, Wistar, male, N = 6
Wiglusz et al. (1975b), Table B-44
300-3,000 mg/m3, 6 hr, Samples collected 0, 2, 7, 14
and 28 d post exposure
Rat, Wistar, male, N = 6
Wiglusz et al. (1975b), Table B-44
Increased aspartate aminotransferase on d 14.
Response relative to control (d 14): 12*%
Increased alkaline phosphatase on d 7 post-exposure.
Response relative to control (on d 7 :0, -0.1, 0.03, 84*%
*, ** Statistically different from controls at p < 0.05 and p < 0.01, respectively.
aln instances where authors reported exposures in ppm, EPA converted these values to mg/m3. See Appendix B for
conversion factor, and individual study summary tables for ppm values.
bTables referenced in Study Design and Reference column correspond to study summary tables in Appendix B
This document is a draft for review purposes only and does not constitute Agency policy.
1-36 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
10000
00
g 1000
43
to
4-1
0)
u
c
o
100
—/— / • • •
r_
c
(
p c
) (
> <
p c
) (
I <
; ;
> c
Concentrations
m ONOAEL
; :
) L
DLOAEL
i Y
i
•
:
Males Males Females Females Males Males Males Males Sex
4-F
Blood
fed 1s White 4, Reticulocytes 4- Clotting 1s So
Cells Blood Cells (a) Time Dehydr
(a) (a)
Hematological Clin
Toxicity Chen
Effe
Subchronic
1,2,4-TMB
rbitol 1s Aspartate 1s Segmented 1s Alkaline
Dgenase Aminotransferase Neutrophilic Phosphatase _• , . .
) on day 14 granulocytes on day "
post-exposure (b) 7 post-exposure jUmmary
(c) (c)
ical Clinical Hematological Clinical Effect
nistry Chemistry Toxicity Chemistry Summary
;cts Effects Effects
Short-term Acute Duration
1,3,5-TMB ,somer
Solid lines represent range of exposure concentrations, (a) Korsak et al. (2000a); (b) Wiglusz et al. (1975a); (c) Wiglusz et al. (1975b). Concentrations (y-axis) in
mg/m3; y-axis is displayed on a logarithmic scale.
Figure 1-7. Exposure response array of hematological and clinical chemistry effects following inhalation
exposure to 1,2,4-TMB or 1,3,5-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
1-37 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1500
— 1000
I
c
o
I 500
c
0)
u
c
o
1,2,3-TMB
:
c
<
p r_
) c
> (
P <
) [
• d
> [j
] (
) <
p :
) C
» (
p :
) (
• (
ji <
1 •
> [j
• •
i i j
p <
) [
> c!
> c
] (
) <
Concentrations
ONOAEL
DLOAEL
P
)
»
Males Males Females Males Females Males Females Females Females Males Sex
4-
Blc
Ce
(«
ted 4- s|
od Segmented Segm
Us Neutrophils Neutr
i (a) (c
/ 1s Lymp
ented (<
ophils
hocytes ^Lymphocytes ^Reticulocytes 1s Retic
i (a) (a) (c
jlocytes 4- ^Alkaline ^ So
Alanine Phosphatase dehydr
amino- (a)
transferase
(a)
rbitol Endpoint
>genase Summary
Hematological Clinical Effect
Toxicity Chemistry Summary
Effects
Subchronic
Duration
Solid lines represent range of exposure concentrations, (a) Korsak et al. (2000b). Concentrations (y-axis) in mg/m .
Figure 1-8. Exposure response array of hematological and clinical chemistry effects following inhalation
exposure to 1,2,3-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
1 -3 8 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Mode of Action Analysis - Hematological and Clinical Chemistry Effects
1 The mode of action for TMB-induced hematological and clinical chemistry effects has not
3 been established. Increased sorbitol dehydrogenase activity is a marker for hepatic injury
4 (Ramaiah, 2007) and therefore, underlying hepatotoxicity could explain its increase in rats exposed
5 to 1,2,4-TMB or 1,2,3-TMB. However, absolute and relative liver weights were not observed to
6 increase with exposure to 1,2,4-TMB, and microscopic histopathological analysis of the liver did not
7 demonstrate any observable changes following exposure to either isomer. The increases in WBC
8 counts in exposed animals could be secondary to the observed respiratory irritative and
9 inflammatory effects of 1,2,4-TMB exposure in Korsak et al. (2000a; 1997).
10 Summary of Hematological and Clinical Chemistry Effects
11 Hematological and clinical chemistry toxicity was observed following inhalation exposure to
12 TMBs based on evidence in humans and animals. The information regarding hematological toxicity
13 in humans is limited to one study involving exposure to a complex VOC mixture containing both
14 1,2,4-TMB and 1,3,5-TMB [Battigetal.. 1956). as reviewed in MOE [2006) and Baettig et al. [1958).
15 Although this study reported hematological effects (alterations in clotting and anemia), exposure
16 was to a mixture of TMB isomers and other VOCs. Therefore, it is impossible to attribute the effects
17 to any TMB isomer. There is evidence of hematological effects in male and female Wistar rats
18 following inhalation exposure [Korsak etal.. 2000a. b), that are roughly analogous to those
19 observed in humans.
20 In summary, the evidence supports a determination that 1,2,4-TMB and 1,2,3-TMB result in
21 hematological toxicity following inhalation exposure, based on consistency and coherency of effects
22 across species (human and rats). The general lack of data on hematological effects following
23 exposure to 1,3,5-TMB precludes a determination of hazard to humans for this isomer, although it
24 is reasonably anticipated given the observed effects following 1,2,4-TMB or 1,2,3-TMB exposure.
25 1.1.5. Carcinogenicity
26 One animal study was identified that investigated the association of chronic oral exposure
27 (via gavage) to 1,2,4-TMB and cancer endpoints [Maltonietal., 1997). Male and female Sprague-
28 Dawley rats were exposed to a single dose of 800 mg/kg-day of 1,2,4-TMB in olive oil by stomach
29 tube for 4 days/week starting at 7 weeks of age. Exposures were terminated at the end of 104
30 weeks (i.e., at 111 weeks of age) and the animals were kept under observation until natural death.
31 The authors report that chronic oral exposure to 1,2,4-TMB resulted in an "intermediate" reduction
32 of survival in male rats and a "slight" reduction in females (no quantitative information on survival
33 was reported). A slight increase in total malignant tumors in both sexes of rats was observed, with
34 the incidence of head cancers being specifically increased in male rats. The predominant type of
35 head cancer identified was neuroesthesioepithelioma, which arises from the olfactory
This document is a draft for review purposes only and does not constitute Agency policy.
1-39 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 neuroepithelium and is normally rare in Sprague-Dawley rats. Other head cancers observed
2 included those in the Zymbal gland, ear duct, and nasal and oral cavities. No tests of statistical
3 significance were reported for these data. When EPA performed the Fisher's exact test on the
4 incidences calculated from the reported percentages of animals bearing tumors in the control and
5 exposed animals, no statistically significant elevations in tumor incidence relative to controls were
6 observed.
7 Janik-Spiechowicz et al. [1998] investigated the genotoxicity of TMB isomers by measuring
8 three genotoxic endpoints: mutation frequency in bacteria, micronucleus formation in mice, and
9 sister chromatid exchanges in mice. Neither 1,2,4-TMB or 1,3,5-TMB induced gene mutations in
10 any Salmonella typhimurium strain tested (TA102, TA100, TA98, and TA97a). However, 1,2,3-TMB
11 induced gene mutations in all four strains in absence of rat S9 fraction. When cells were incubated
12 in the presence of S9,1,2,3-TMB did not induce gene mutation, indicating possibly that 1,2,3-TMB
13 itself is the primary mutagen. No isomer induced the formation of micronuclei in Imp:BALB/c mice
14 following i.p. injection. Males in the high-dose groups for 1,2,4-TMB and 1,3,5-TMB, but not 1,2,3-
15 TMB, exhibited a statistically significant reduction in the ratio of polychromatic erythrocytes to
16 normochromatic erythrocytes, indicating bone marrow cytotoxicity. All three isomers significantly
17 increased the frequency of sister chromatid exchanges (SCEs) in Imp:BALB/c mice following i.p.
18 injection, with 1,2,4-TMB eliciting the more significant response. These results appear to have
19 occurred at doses that did not induce significant bone marrow cytotoxicity.
20 In summary, very little genotoxicity data are available on TMBs. Janik-Spiechowicz et al.
21 [1998] observed varying results in the Ames mutation assay in Salmonella, with 1,2,3-TMB, but not
22 1,2,4-TMB or 1,3,5-TMB, inducing gene mutations. Results for the in vivo assays for micronucleus
23 and SCE formation were consistent across isomers: TMB isomers were observed to induce SCEs,
24 but not micronuclei in mouse bone marrow cells. Increased frequency of SCEs indicates that DNA
25 damage has occurred as a result of exposure to these isomers, but it does not provide a specific
26 indication of mutagenic potential, as there is no known mechanistic association between SCE
27 induction and a transmissible genotoxic effect With only one isomer (1,2,3-TMB] demonstrating a
28 positive result for gene mutation and positive SCE results for all three isomers, there is inadequate
29 evidence to conclude that any isomer is directly genotoxic.
30 1.1.6. Similarities Among TMB Isomers Regarding Observed Inhalation and Oral
31 Toxicity
32 In the existing toxicological database for 1,2,3-TMB, 1,2,4-TMB, and 1,3,5-TMB, important
33 similarities have been observed in the potency and magnitude of effect resulting from exposure to
34 these three isomers in male and female Wistar rats, although some important differences also exist
35 In acute studies investigating respiratory irritative effects, the RDso of the three isomers
36 were very similar [Korsaketal., 1997]. Measures of neurotoxicity, namely ECso values for
37 decreases in rotarod performance and pain sensitivity, following acute inhalation exposures were
This document is a draft for review purposes only and does not constitute Agency policy.
I -40 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 similar for 1,2,4-TMB and 1,3,5-TMB [Korsak and Rydzynski. 19961. However, the EC50 values for
2 both measures were lower following exposure to 1,2,3-TMB. The observation that 1,2,3-TMB may
3 be slightly more neurotoxic than 1,2,4-TMB or 1,3,5-TMB was also observed following acute, oral,
4 and injection exposures. Although all three isomers were observed to result in altered EEC
5 readings, stronger and more persistent effects followed a pattern of 1,2,3-TMB > 1,3,5-TMB > 1,2,4-
6 TMB following oral exposures [Tomasetal.. 1999a] and 1,2,3-TMB > 1,2,4-TMB > 1,3,5-TMB
7 following i.p. injections [Tomas etal.. 1999c]. Acute exposure to both 1,2,4-TMB and 1,2,3-TMB
8 affected motor function and/or anxiety at similar exposure levels, whereas 1,3,5-TMB appeared to
9 be slightly more potent, although the magnitude of the response across isomers suggests that this
10 difference is negligible [Tomas etal.. 1999b].
11 In short-term neurotoxicity studies, a qualitatively similar pattern of effects (inability to
12 learn passive and/or active avoidance and decreased pain sensitivity) indicating altered
13 neurobehavioral function was observed for TMBs, although some quantitative differences were
14 noted fWiadernaetal.. 19981 fWiaderna etal.. 2002: Gralewicz and Wiaderna. 2001: Gralewicz et
15 al., 1997a]. Exposure to any isomer resulted in statistically significant decreases in pain sensitivity
16 at the same concentration, although the magnitude of effect was greater for 1,3,5-TMB and 1,2,4-
17 TMB compared to 1,2,3-TMB [Wiaderna etal.. 2002: Gralewicz and Wiaderna. 2001: Wiaderna et
18 al.. 1998: Gralewicz etal., 1997a]. 1,2,4-TMB and 1,3,5-TMB were also observed to change motor
19 function and/or anxiety, whereas 1,2,3-TMB was observed to have no effect on this parameter
20 [Lutz etal.. 2010: Wiaderna etal.. 2002.1998: Gralewicz etal.. 1997a]. In contrast, motor activity
21 and/or anxiety responses elicited by amphetamine were amplified following exposure to 1,2,3-
22 TMB, but not 1,2,4-TMB [Lutz etal., 2010]. All three isomers elicited effects on cognitive function
23 as measured by the ability to learn either passive or active avoidance tasks [Wiaderna etal., 2002:
24 Gralewicz and Wiaderna. 2001: Wiaderna etal.. 1998: Gralewicz etal.. 1997a]. 1,3,5-TMB was
25 observed to be the most potent isomer in this regard, eliciting effects on both passive and active
26 avoidance at > 123 mg/m3. 1,2,3-TMB and 1,2,4-TMB affected the ability to learn passive avoidance
27 at > 123 and > 492 mg/m3, respectively, and both 1,2,3-TMB and 1,2,4-TMB affected the ability to
28 learn active avoidance at 492 mg/m3.
29 Following subchronic exposure to either 1,2,4-TMB or 1,2,3-TMB, both decreased pain
30 sensitivity and decreased rotarod performance were observed. With regard to decreased pain
31 sensitivity, although 1,2,3-TMB was observed to decrease pain sensitivity at a lower concentration
32 than 1,2,4-TMB, the magnitude of effect was similar between isomers at every concentration
33 [Korsak and Rydzynski, 1996]. 1,2,3-TMB was more potent than 1,2,4-TMB in reducing rotarod
34 performance, both in the concentrations eliciting an effect as well as the magnitude of effect at each
35 concentration [Korsak and Rydzynski. 1996].
36 Lastly, similarities were observed in 1,2,4-TMB- and 1,3,5-TMB-induced developmental and
37 maternal effects [Saillenfaitetal., 2005]. Male fetal weights were significantly reduced in animals
38 exposed gestationally to 2,952 mg/m31,2,4-TMB (5% decrease] or 1,3,5-TMB (7% decrease].
This document is a draft for review purposes only and does not constitute Agency policy.
1 -41 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 1,2,4-TMB also significantly decreased female fetal weights by approximately 5% in animals
2 exposed to the same concentration. Although, 1,3,5-1MB significantly reduced female fetal weights
3 by 13% in animals exposed to 5,904 mg/m3, female fetal weights were decreased at 2,952 mg/m3
4 to a similar degree (6%) as animals exposed to the same concentration of 1,2,4-TMB. Maternal
5 toxicity, measured as decreased corrected maternal weight gain, was significantly decreased in
6 animals exposed to 2,952 mg/m31,2,4-TMB or 1,3,5-TMB. However, 1,3,5-TMB exposure resulted
7 in a 75% reduction of maternal weight gain compared to controls, whereas 1,2,4-TMB exposure
8 reduced maternal weight gain by 50%. A summary of these comparisons across isomers is
9 presented below in Table 1-6.
10
11 Table 1-6. Similarities between 1,2,3-TMB, 1,2,4-TMB, and 1,3,5-TMB
12 regarding observed inhalation and oral toxicity
Health Outcome Measure
Pain Sensitivity
Neuromuscular Function
Motor Function /Anxiety
Sensitization
Cognitive Function
Electrocortical activity
Respiratory Effects
Developmental Effects
Hematological Effects
Exposure Duration
acute
short-term
subchronic
acute
subchronic
short-term
short-term
short-term
acute
acute
gestational
subchronic
TMB Isomer Potency
1,2,3-TMB > 1,2,4-TMB * 1,3,5-TMB
1,2,4-TMB * 1,3,5-TMB > 1,2,3-TMB
1,2,4-TMB * 1,2,3-TMB
1,2,3-TMB > 1,2,4-TMB * 1,3,5-TMB
1,2,3-TMB > 1,2,4-TMB
1,2,4-TMB * 1,3,5-TMB » 1,2,3-TMB
1,2,3-TMB > 1,2,4-TMB
1,3,5-TMB > 1,2,4-TMB * 1,2,3-TMB
1,2,3-TMB » 1,3,5-TMB > 1,2,4-TMB
1,2,4-TMB * 1,3,5-TMB * 1,2,3-TMB
1,2,4-TMB = 1,3,5-TMB
1,2,4-TMB * 1,2,3-TMB
13 1.2. Summary and Evaluation
14 1.2.1. Weight of Evidence for Effects Other than Cancer
15 In both humans and animals, inhalation exposure to TMBs has been shown to result in
16 toxicity in multiple organ systems, including the nervous, respiratory, and hematological systems.
17 In addition, developmental toxicity has been observed in animals exposed to either 1,2,4-TMB or
18 1,3,5-TMB. Generally, the information regarding inhalation toxicity in humans is limited for a
19 number of reasons, including that the majority of human studies involved exposure to complex VOC
20 mixtures containing several TMB isomers and other VOCs, and not the individual isomers
21 themselves. Therefore, the observed health effects cannot be attributed to specific TMB isomers.
22 However, these studies observe effects in exposed human populations that are generally analogous
23 to effects observed in animal toxicity studies, and provide qualitative, supportive evidence for
24 hazard identification. Currently, no human studies exist that investigate the oral toxicity of any
This document is a draft for review purposes only and does not constitute Agency policy.
1 -42 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 1MB isomer. Potential limitations in the animal inhalation and oral toxicity database for TMBs
2 include the lack of a chronic study and the fact that all of the available inhalation animal studies
3 were conducted by the same research group: The Nofer Institute of Occupational Medicine, Lodz
4 Poland.
5 The most strongly and widely supported manifestation of toxicity in humans and animals
6 following inhalation exposure to 1,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB is neurotoxicity. In humans
7 exposed to TMB-containing VOC mixtures, a multitude of effects, including neuropsychological
8 effects [Chen etal., 1999], deficits in short-term memory and reduced motor speed/coordination
9 [Lee etal.. 2005). abnormal fatigue [Norsethetal.. 1991). dysfunction of the inner ear/vertigo
10 [Sulkowski etal.. 2002]. and nervousness, anxiety, and/or vertigo (Battig et al. [1956]. as reviewed
11 by MOE [2006] and Baettig et al. [1958], have been observed. None of the available human studies
12 have addressed the potential for latent neurological effects or effects in sensitive populations.
13 Although the reported human symptoms do not directly parallel the animal data, exposure of male
14 Wistar rats to the TMB isomers has been shown to consistently result in a multitude of neurotoxic
15 effects, including decreased pain sensitivity, impaired neuromuscular function and coordination,
16 altered cognitive function, decreased anxiety and/or increased motor function, and
17 neurophysiological effects (e.g., decreased electrocortical activity] across multiple concentrations
18 and durations [Wiadernaetal., 2002: Gralewicz and Wiaderna, 2001: Wiaderna et al., 1998:
19 Gralewicz etal..!997a: Gralewicz etal.. 1997b: Korsak and Rydzynski. 1996: Korsak etal.. 1995].
20 The effects observed in the animal neurotoxicity studies are recognized in the U.S. EPA's Guidelines
21 for Neurotoxicity Risk Assessment [U.S. EPA. 1998] as possible indicators of neurotoxicity. The
22 neurotoxic effects are biologically plausible and analogous to effects that could occur in humans.
23 The evidence for TMBs identifies neurotoxicity as a toxicity hazard based on consistency and
24 coherency of effect across multiple studies and durations of exposure.
25 Three acute oral studies [Tomas etal.. 1999a: Tomas et al.. 1999b: Tomas etal.. 1999c] exist
26 that observe similar effects as observed in the available inhalation neurotoxicity studies (i.e.,
27 increased motor activity and altered brain wave activity]. However, these studies are limited with
28 regard to their duration (i.e., acute] and nature of endpoints investigated, and as such, no weight of
29 evidence determination can be made regarding the oral toxicity of the TMB isomers.
30 In addition to neurotoxicity, both respiratory and hematological toxicity have been
31 observed in human populations and animals exposed to TMBs, or to mixtures containing the three
32 isomers. In humans, occupational and residential exposure to VOC mixtures containing TMB
33 isomers have resulted in number of effects characterized as respiratory toxicity, including
34 asthmatic bronchitis [[Battig etal.. 1956]. as reviewed in MOE [2006] and Baettig et al. [1958]].
35 asthma [Billionnetetal.. 2011]. or laryngeal/pharyngeal irritation [Norsethetal.. 1991].
36 Additionally, workers exposed to a VOC mixture containing 1,2,4-TMB and 1,3,5-TMB, and possibly
37 1,2,3-TMB, were reported to exhibit hematological effects including alterations in clotting time and
38 anemia [[Battig etal.. 1956]. as reviewed in MOE [2006] and Baettig et al. [1958]]. Again, as
This document is a draft for review purposes only and does not constitute Agency policy.
I -43 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 workers were exposed to complex VOC mixtures containing 1MB isomers, the observed health
2 effects cannot be attributed to any single 1MB isomer.
3 The observation of respiratory irritation and inflammation in Wistar rats and BALB/C mice
4 following exposure to 1,2,4-TMB was consistent across multiple concentrations, and subchronic
5 and acute exposure durations [Korsaketal.. 2000a: Korsaketal.. 1997: Korsaketal.. 1995].
6 Respiratory toxicity was also observed in multiple studies involving exposure to 1,2,3-TMB [Korsak
7 etal.. 2000b: Korsak et al.. 1995). Although the reported symptoms in humans (laryngeal and/or
8 pharyngeal irritation, asthmatic bronchitis, and asthma) do not directly parallel the effects
9 observed in animal studies, the observation of irritative and/or inflammatory responses in multiple
10 species (including humans) demonstrates a consistency in TMB-induced respiratory toxicity.
11 Additionally, multiple measures of hematological toxicity have been observed in rats subchronically
12 exposed to 1,2,4-TMB or 1,2,3-TMB, including decreased RBCs, increased WBCs, decreased clotting
13 time, and decreased reticulocytes (1,2,4-TMB) and decreased RBCs, decreased segmented
14 neutrophils, increased lymphocytes and increased reticulocytes (1,2,3-TMB) (Korsaketal.. 2000a.
15 b). At least two of these effects, decreased RBCs and decreased clotting time, are roughly analogous
16 to the hematological effects (alterations in clotting and anemia) observed in occupationally exposed
17 humans, thereby demonstrating a consistency and coherency of effect across species. Therefore,
18 the respiratory and hematological effects observed in animals are biologically plausible and
19 analogous to effects that could occur in exposed human populations. The available weight of
20 evidence for 1,2,4-TMB and 1,2,3-TMB identified respiratory and hematological toxicity as a hazard.
21 Currently, no human studies exist that investigate the reproductive or developmental
22 toxicity of 1,2,3-TMB, 1,2,4-TMB, or 1,3,5-TMB. However, one animal study (Saillenfaitetal.. 2005)
23 observed effects on fetal body weights and maternal body weight gains due to gestational exposure
24 to 1,2,4-TMB or 1,3,5-TMB. Although the weight of evidence regarding developmental toxicity is
25 not as strong compared to other measures of toxicity in the TMB database, these effects observed in
26 animals are considered biologically plausible and potentially analogous to effects that could occur
27 in humans. The available evidence for 1,2,4-TMB and 1,3,5-TMB identifies maternal and
28 developmental toxicity as a hazard.
29 1.2.2. Weight of Evidence for Carcinogenicity
30 Under the Guidelines for Carcinogen Risk Assessment (2005), the database for the TMBs
31 provides "inadequate information to assess carcinogenic potential" of these isomers. This
32 characterization is based on the fact that there is no information regarding the carcinogenicity of
33 TMB in humans and that the only animal study available on the carcinogenicity of 1,2,4-TMB
34 observed no statistically significant carcinogenic effects. No studies regarding the carcinogenicity
35 of 1,2,3-TMB or 1,3,5-TMB were identified in the available scientific literature.
36 In the animal carcinogenicity study identified (Mai toni etal., 1997), involving exposure to
37 1,2,4-TMB by oral gavage, an increased incidence of total malignant tumors in both sexes and head
This document is a draft for review purposes only and does not constitute Agency policy.
I -44 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 cancers (predominantly neuroethesioepithelioma) in males was observed in exposed rats, no
2 statistical analyses were reported. When EPA independently performed the Fisher's exact test on
3 the reported data, no statistically significant effects were observed.
4 Additionally, in the only study investigating the genotoxicity of TMB isomers, Janik-
5 Spiechowicz et al. [1998] observed negative results in in vitro genotoxicity assays (i.e., Ames
6 mutation assay in Salmonella) involving 1,2,4-TMB and 1,3,5-TMB. However, 1,2,3-TMB was
7 observed to induce gene mutations in all Salmonella typhimurium strains tested. All three isomers
8 failed to induce micronuclei in mouse bone marrow cells. Janik-Spiechowicz et al. [1998] observed
9 an increased incidence of SCE in mice exposed to all three TMB isomers (individually); however,
10 this observation does not provide a specific indication of mutagenic potential. Given the findings
11 regarding the in vitro genotoxicity of the TMB isomers, and the uncertainty regarding the
12 interpretation of the SCE results, the evidence is inadequate to conclude that any TMB isomer is
13 genotoxic.
14 1.2.3. Susceptible Populations and Lifestages
15 Although there are no chemical-specific data that would allow for the identification of
16 susceptible populations and lifestages, the reduced metabolic and elimination capacities in children
17 relative to adults may be a source of susceptibility (Ginsberg etal., 2004). TMB isomers are
18 metabolized following inhalation and oral exposure via side-chain oxidation to form alcohols and
19 aromatic carboxylic/mercapturic acids or by hydroxylation to form phenols, which are then
20 conjugated with glucuronic acid, glycine, or sulfates for urinary excretion. The activities of multiple
21 cytochrome P450 (GYP P450) mono-oxygenase isozymes have been shown to be reduced in
22 children up to 1 year of age compared to adult activities (Ginsberg etal., 2004). Additionally, the
23 rate of glucuronidation and sulfation is decreased in children. Therefore, as both GYP P450 mono-
24 oxygenase activities and the rate of glucuronidation and sulfation appear to be decreased in early
25 life, newborns and young infants may experience higher and more persistent blood concentrations
26 of the TMB isomers, and/or their respective metabolites compared with adults at similar exposure
27 levels. Reduced renal clearance in children may be another important source of potential
28 susceptibility. TMB isomers and their metabolites are excreted in the urine of exposed laboratory
29 animals and occupationally exposed humans. Data indicating reduced renal clearance for infants
30 up to 2 months of age (Ginsberg etal., 2004] may suggest a potential to affect TMB excretion, thus
31 possibly prolonging its toxic effects. Additionally, those with pre-existing respiratory diseases (e.g.,
32 asthma) may be more sensitive to the respiratory irritative and inflammatory effects of TMB
33 isomers.
34
This document is a draft for review purposes only and does not constitute Agency policy.
I -45 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
2. DOSE-RESPONSE ANALYSIS
3 2.1. Inhalation Reference Concentration for Effects Other Than Cancer for 1,2,4-
4 TMB
5 2.1.1. Identification of Candidate Principal Studies and Critical Effects for 1,2,4-TMB
6 The nervous, respiratory, and hematological systems are the primary targets of inhaled
7 1,2,4-TMB in humans and experimental animals, and effects in these systems have been identified
8 as hazards following inhalation exposure to 1,2,4-TMB.
9 The selection of studies and general procedures for dose-response analysis are discussed in
10 sections 6 and 7 of the Preamble. Human data are preferred over animal data for deriving
11 reference values when possible because the use of human data is more relevant in the assessment
12 of human health and avoids the uncertainty associated with interspecies extrapolation introduced
13 when animal data serve as the basis for the reference value. In this case, while literature exists on
14 the effects of 1,2,4-TMB exposure in humans, including neurological, respiratory, and hematological
15 toxicities, no human studies are available that would allow for dose-response analysis. The human
16 studies evaluated TMB exposures occurring as complex solvents or VOC mixtures, and this
17 confounding along with other uncertainties including high imprecision in effect measures due to
18 low statistical power, lack of quantitative exposure assessment, and lack of control for co-
19 exposures, limit their utility in derivation of quantitative human health toxicity values. However,
20 these studies provide supportive evidence for the neurological, respiratory, and hematological
21 toxicity of TMB isomers in humans and indicate a coherency of effects in both humans and
22 laboratory animals.
23 Several studies investigating 1,2,4-TMB effects in experimental animal models were
24 identified in the literature. No chronic studies were available, although acute, short-term, and
25 subchronic studies were identified. 1,2,4-TMB-induced toxicity was observed across several organ
26 systems in three subchronic studies by Korsak et al., [2000a: 1997] and Korsak and Rydzyhski
27 [1996]. These were the only subchronic studies identified in the peer-reviewed literature. Data
28 from these studies pertaining to the primary hazards observed in humans and animals identified in
29 Chapter 1 (neurological, respiratory, and hematological toxicity] were considered as candidate
30 critical effects for the purpose of determining the point of departure (POD] for derivation of the
31 inhalation RfC for 1,2,4-TMB. Neurotoxicity was also observed in both acute and short-term
32 inhalation studies and respiratory toxicity was also observed in acute studies. However, the high
33 concentrations used in acute studies and the short exposure durations of both acute and short-term
34 studies limit their utility for the quantitation of chronic human health effects. Nevertheless, as with
This document is a draft for review purposes only and does not constitute Agency policy.
2-1 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 the human mixture studies, these studies provide qualitative information regarding hazard
2 identification, especially the observation of the consistency and coherency of these effects across
3 the 1,2,4-TMB database.
4 The three subchronic studies by Korsak et al., [2000a: 1997] and Korsak and Rydzyhski
5 [1996] are adequate for dose response analysis. All three studies used rats as an appropriate
6 laboratory animal species, and utilized appropriate sham-exposed controls. Animals were exposed
7 to 1,2,4-TMB reported as > 97% pure (impurities not reported]. These studies utilized an
8 appropriate route [inhaled air] and duration [subchronic] of exposure. The studies used a
9 reasonable range of appropriately-spaced exposure levels to facilitate dose-response analysis. An
10 appropriate latency between exposure and development of toxicological outcomes was used, and
11 the persistence of some outcomes after termination of exposure was investigated. Adequate
12 numbers of animals per exposure group were used, and appropriate statistical tests including pair-
13 wise and trend analyses were performed. With regard to reporting of exposure methodologies,
14 Korsak et al. [2000a] reported actual concentrations, as measured by gas chromatography, to be
15 within 10% of target concentrations. This increases the confidence in the overall evaluation and
16 adequacy of this study. Although Korsak and Rydzyhski [1996] and Korsak et al. [1997] do not
17 report actual, measured concentrations, these studies use the same exposure methodology as
18 Korsak et al. [2000a]: suggesting that it is likely that the actual concentrations in these studies are
19 within 10% of target concentrations. Target and actual concentrations, as well as internal blood
20 dose metrics calculated using the PBPK model, are listed in Table 2-1.
21
This document is a draft for review purposes only and does not constitute Agency policy.
2-2 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 2-1. Internal blood dose metrics calculated using the available
rat PBPK model (Hissinketal.. 2007)
Reference
Korsak and
Rydzyriski
(1996)
Korsak et al.
(1997)
Korsak et al.
(2000a)
Species/
sex
Rat, male
Rat, male
Rat, male
Rat, female
Body weight
(kg)a
0.387
0.404
0.403
0.383
0.409
0.416
0.390
0.399
0.389
0.243
0.230
0.229
Exposure concentration
(mg/m3)b
123
492
1,230
123
492
1,230
129
492
1,207
129
492
1,207
Internal dose - average weekly
venous blood concentration
(mg/L)
0.1272
0.8666
5.4424
0.1272
0.8661
5.4274
0.1339
0.8671
5.2481
0.1335
0.8899
5.5189
J
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
aFor Korsak et al. (2000a; 1997), exposure group-specific terminal body weights from those studies were used to
calculate internal dose metrics; for Korsak and Rydzyriski (1996) the average of the exposure group-specific body
weights reported in Korsak et al. (2000a; 1997) were used in internal dose metric calculations.
bFor Korsak and Rydzyriski (1996) and Korsak et al. (1997) exposure concentrations are target concentrations, for
Korsak (2000a) exposure concentrations are actual concentrations as measured by gas chromatography.
These subchronic studies examined 1,2,4-TMB-induced toxicity in multiple organ systems
and neurological, respiratory, and hematological endpoints that demonstrated statistically
significant pair-wise increases or decreases relative to control were considered for the derivation of
the RfC for 1,2,4-TMB (Table 2-2). The endpoints included decreased pain sensitivity in male rats
[Korsak and Rydzynski. 1996}. increased BAL total cells in male rats [Korsak etal.. 1997}. increased
inflammatory lung lesions, decreased RBCs, and increased WBCs in male rats and decreased
reticulocytes and clotting time in female rats [Korsak etal.. 2000a]. Increases in BAL
polymorphonuclear leukocytes and lymphocytes observed in the Korsak et al. [1997] study were
not considered for RfC derivation due to a lack of reporting of exposures in which statistically
significant increases occurred. Additionally, Korsak et al. [1997] reported that 123 mg/m3 was the
LOAEL for increased BAL total cells, but the NOAEL for increased BAL macrophages. Therefore,
increased BAL macrophages were not considered for RfC derivation as these effects were not
observed at concentrations that elicited an increase in total BAL cells. Changes in BAL protein and
enzyme activity level were not considered due to non-monotonically increasing dose-responses,
and increases in sorbitol dehydrogenase were not further considered due to the lack of
accompanying hepatocellular histopathological alterations in exposed animals.
This document is a draft for review purposes only and does not constitute Agency policy.
2-3 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 2-2. Endpoints resulting from subchronic inhalation exposure to
1,2,4-TMB considered for the derivation of the RfC
Endpoint
Species/ sex
Exposure concentration (mg/ms)a
0
123
492
1,230
Neurological endpoints
Decreased pain sensitivity (measured as
latency to paw-lick in seconds)b
Rat, male
15.4 ±5.8
(n = 9)
18.2 ±5.7
(n = 10)
27.6 ±3.2**
(n = 9)
30.1 ±7.9**
(n = 10)
Hematological endpoints
Decreased RBCs (106/cm3)c
Increased WBCs (106/cm3)c
Decreased reticulocytes (%)c
Decreased clotting time (s)c
Rat, male
Rat, female
9.98 ±1.68
(n = 10)
8.68 ±2.89
(n = 10)
3.5 ±2.6
(n = 10)
30 ±10
(n = 10)
9.84 ± 1.82
(n = 10)
8.92 ± 3.44
(n = 10)
1.7 ±2.0
(n = 10)
23 ±4
(n = 10)
8.50 ±1.11
(n = 10)
8.30 ±1.84
(n = 10)
1.8 ±0.9
(n = 10)
**
19 ±5
(n = 10)
7.70+1.38**
(n = 10)
15.89 ± 5.74**
(n = 10)
1.0 ±0.6*
(n = 10)
22 ±7*
(n =10)
Respiratory endpoints
Increased BAL total cells (106/cm3)d
Increased inflammatory lung lesions0
Rat, male
1.93 ±0.79
(n = 6)
e
(n = 10)
5.82 ± 1.32*"
(n = 6)
e
(n =10)
5.96 ± 2.80**
(n = 7)
e
(n = 10)
4.45 ± 1.58*
(n = 7)
e
(n = 10)
3
4
5
6
7
8
9
10
11
12
13
14
*p < 0.05; **p < 0.01; *** p < 0.001.
aValues are expressed as mean ± 1SD.
b Adapted from Korsak and Rydzyriski (1996)
"Adapted from Korsak et al. (2000a)
d Adapted from Korsak et al. (1997)
Incidences for individual exposure groups not reported; however, based on qualitative information reported in the
study (i.e., that female rats exhibited a statistically significant increase in inflammatory lung lesions at 492 mg/m3),
a NOAEL of 123 mg/m3 was identified.
Impaired neuromuscular function and coordination, measured as performance on the
rotarod apparatus, was also observed in rats exposed to 1,2,4-TMB. The use of rotarod data from
Korsak and Rydzyhski [1996] was initially considered as a candidate critical effect for 1,2,4-TMB.
However, upon critical evaluation of the exposure-response information in the study, it was
determined that the endpoint was reported in a manner that reduced the confidence in the
observed effect levels. The primary limitation noted for these data relates to the presentation of
rotarod performance, which is best represented as a continuous variable, as opposed to a quantal
variable such as that presented by Korsak and Rydzyhski [1996]. In contrast to the percent failures
reported by the study authors, the most widely used and accepted measure of rotarod performance
in rodents is latency to fall from the rotating rod [Brooks andDunnett. 2009: Kaspar etal.. 2003:
Bogo etal., 1981], typically with an arbitrary upper limit on the maximum latency allowed to
prevent confounding by fatigue. Although the quantal percent failures data can provide useful
This document is a draft for review purposes only and does not constitute Agency policy.
2-4 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 information, these measures require an arbitrary selection of the length of time required for
2 successful performance; there is no scientific consensus on an optimal time for this parameter. In
3 addition, when identifying effect levels based on the data presented by Korsak and Rydzyhski
4 [1996], latencies on the rod of 1 versus 119 seconds would be treated identically as failures when,
5 in fact, they indicate very different levels of neurological dysfunction [Bogo etal., 1981]. This adds
6 uncertainty when trying to extrapolate to a concentration associated with a minimally adverse
7 effect Finally, this quantal presentation of data does not allow for interpretations related to intra-
8 rat and intra-group variability in performance. Due to these reporting limitations, impaired
9 neuromuscular function and coordination, measured as performance on the rotarod apparatus, was
10 excluded from consideration for derivation of the RfC for 1,2,4-TMB.
11 Additionally, although the Saillenfait et al. [2005] study was a well conducted
12 developmental toxicity study, data from this study were not considered for identification of
13 candidate critical effects for 1,2,4-TMB due to the fact that maternal and developmental toxicities
14 were observed at concentrations 6- to 24-fold higher than the concentrations that resulted in the
15 neurological, respiratory, and hematological effects observed in the subchronic Korsak studies.
16 Endpoints carried forward for derivation of an RfC for 1,2,4-TMB, along with their exposure
17 ranges and NOAEL/LOAEL values (identified by EPA] are graphically presented in Figure 2-1.
18
This document is a draft for review purposes only and does not constitute Agency policy.
2-5 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
10000 -,
1
2
3
4
5
6
"55
.2 1000
E
01
u
100
Concentrations
ONOAEL
DLOAEL
o
d
D
O
O
o
o
D
o
pain sensitivity BAL pulmonary RBCs
(a) total cells lesions (c)
(b) (b)
WBCs reticulo-
(c) cytes
(c)
4, t t
clotting maternal fetal
time weight gain weight
(c) (d) (d)
9
10
11
12
13
14
15
16
17
18
19
20
21
Solid lines represent range of concentrations, (a) Korsak and Rydzyriski (1996); (b) Korsak et al. (1997); (c)
Korsak et al. (2000a); (d) Saillenfait et al. (2005).
Figure 2-1. Exposure response array of endpoints resulting from
inhalation exposure to 1,2,4-TMB considered for the derivation of the
RfC.
2.1.2. Methods of Analysis for 1,2,4-TMB
This assessment uses PBPK model estimates of internal blood dose metrics coupled with
the benchmark dose (BMD) approach, when possible, to estimate a POD for the derivation of an RfC
for 1,2,4-TMB (see Section B.2 of Appendix B and Section C.I of Appendix C for details regarding
PBPK model estimates and BMD modeling, respectively). As dosimetry can often be non-linear due
to metabolic saturation, and internal dose metrics are expected to correlate more closely to toxic
response than external concentrations [Mclanahanetal.. 2012]. the order of analysis employed in
this assessment is calculation of internal dose metrics with the available PBPK model first, followed
by BMD modeling using the PBPK model-estimated internal dose metrics.
For 1,2,4-TMB, the available deterministic PBPK rat model fHissinketal.. 20071 was used to
convert non-continuous external inhalation concentrations (in mg/m3) of 1,2,4-TMB to the internal
blood dose metric of average weekly venous blood concentration (in mg/L) of 1,2,4-TMB (see Table
2-1). Weekly average venous blood 1,2,4-TMB concentration was chosen as the internal dose
metric on which to base the RfC as it is assumed that the parent compound is the toxic moiety of
interest and that average venous blood concentration of 1,2,4-TMB is assumed to adequately
This document is a draft for review purposes only and does not constitute Agency policy.
2-6 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 represent the target tissue dose across the multiple tissues of interest. The use of concentration of
2 parent compound in venous blood as the relevant dose metric in non-metabolizing, non-first pass
3 organs is recommended by Aylward et al. [2011]. Furthermore, toluene-induced neurological
4 effects in the brain are provided by Aylward et al. [2011] as an example of a chemically induced
5 toxic endpoint for which this dose metric is relevant. As discussed in Section 1 (Mode of Action
6 Analysis - Neurotoxic Effects], 1,2,4-TMB is reasonably expected to have a mode of action for
7 neurotoxic effects similar to toluene, further supporting the selection of venous blood
8 concentration as the relevant internal dose metric.
9 After calculation of internal blood dose metrics, those dose metrics were used as the dose
10 inputs for BMD modeling. The BMD approach involves fitting a suite of mathematical models to the
11 observed dose-response data using EPA's Benchmark Dose Software (BMDS, version 2.2]. Each
12 fitted model estimates a BMD and its associated 95% lower confidence limit (BMDL] corresponding
13 to a selected benchmark response (BMR]. For continuous data (i.e., decreased pain sensitivity,
14 increased BAL total cells, decreased RBCs, decreased reticulocytes, and decreased clotting time]
15 from the Korsak and Rydzyhski [1996] and Korsak et al. [2000a: 1997] studies, no information is
16 available regarding the change in these responses that would be considered biologically significant,
17 thus a BMR equal to a change in the mean equal to 1 standard deviation of the model estimated
18 control mean was used in modeling these endpoints, consistent with EPA's draft Benchmark Dose
19 Technical Guidance Document [U.S. EPA. 2000]. The estimated BMDL is then used as the POD for
20 deriving the RfC (Table 2-3].
21 The suitability of the above methods to determine a POD is dependent on the nature of the
22 toxicity database for a specific chemical. Some endpoints for 1,2,4-TMB were not modeled for a
23 variety of reasons, including equal responses at all exposure groups (e.g., increased BAL total cells
24 and decreased reticulocytes], responses only in the high exposure group with no changes in
25 responses in lower exposure groups (e.g., increased WBCs], and absence of incidence data (e.g.,
26 increased inflammatory lung lesions]. Additionally, some datasets were modeled, but no model
27 provided estimated BMDLs that were considered to be biologically plausible (e.g., decreased
28 clotting time]. In cases where BMD modeling was not feasible or modeling failed to appropriately
29 describe the dose-response characteristics, the NOAEL/LOAEL approach was used to identify a
30 POD. Detailed modeling results are provided in Section C.I of Appendix C. Additionally, detailed
31 modeling results for maternal and fetal endpoints observed in Saillenfait et al. [2005] are provided
32 in Appendix C for comparison to endpoints observed in the Korsak et al. [2000a: 1997] and Korsak
33 and Rydzyhski [1996] studies.
34
This document is a draft for review purposes only and does not constitute Agency policy.
2-7 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 2-3. Summary of dose-response analysis and point of departure
estimation for endpoints resulting from subchronic inhalation
exposure to 1,2,4-TMB
Reference
Endpoint
Species/sex
POD
basis
Best-fit model; BMR
Candidate
PODADJa
(mg/L)
Neurological endpoints
Korsakand
Rydzyriski (1996)
Decreased pain
sensitivity
Rat, male
BMDL
Exponential 4; ISO
0.086
Hematological endpoints
Korsak et al. (2000a)
Decreased RBCs
Increased WBCs
Decreased
reticulocytes
Decreased clotting
time
Rat, male
Rat, male
Rat, female
Rat, female
BMDL
NOAEL
NOAEL
NOAEL
Linear; ISO
n/ab
n/ab
n/ab
0.499
0.867
0.890
0.134
Respiratory endpoints
Korsak et al. (1997)
Korsak et al. (2000a)
Increased BAL total
cells
Increased
inflammatory lung
lesions
Rat, male
Rat, male
LOAEL
NOAEL
n/ab
n/ab
0.127
0.134
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
aWeekly average venous blood 1,2,4-TMB concentration (mg/L). See Appendix B for details on PBPK
modeling.
bNo model was able to fit data adequately, or data were not modeled.
One consequence of using PBPK model-estimated internal dose metrics as the dose inputs
for BMD modeling was the necessity of dropping the high exposure group in all datasets modeled.
During the validation and optimization of the animal PBPK model [Hissinketal., 2007] against
available animal toxicokinetic datasets, the model accurately reproduced venous blood
concentrations of 1,2,4-TMB following repeated (6 hours/day, 5 days/week, 4 weeks) exposures to
123 or 492 mg/m3 (see Section B.3.3.2, Appendix B). However, the PBPK model consistently
overpredicted venous blood concentrations following exposure to 1,230 mg/m3. It was concluded
that the optimized animal PBPK model produces acceptable simulations of venous blood 1,2,4-TMB
concentrations for chronic exposures to 100 ppm [492 mg/m3] in rats following inhalation
exposure to 1,2,4-TMB (Section B.3.3.2, Appendix B). Therefore, as the model-estimated internal
blood dose metrics at the high concentration are not representative of empirically observed blood
concentrations, using the high-dose model estimates as dose inputs for BMD modeling is not
appropriate. The decision to drop the high concentration results in a loss of information regarding
dose-response characteristics at high concentrations and a reduction in the number of available
dose-response models to fit to the data (due to the number of model parameters > exposure
groups). However, this methodology is preferred over inclusion of demonstrably inaccurate
This document is a draft for review purposes only and does not constitute Agency policy.
2-8 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 internal blood dose metrics that result from high concentrations. Additionally, this methodology
2 still allows for BMD modeling of these endpoints, which is preferred over use of the NOAEL/LOAEL
3 approach.
4 2.1.3. Derivation of the Reference Concentration for 1,2,4-TMB
5 For the derivation of an RfC based upon animal data, the calculated POD values are
6 converted to human equivalent concentrations (HECs) using the available human PBPK model
7 fHissinketal.. 20071 (Table 2-4).
9
10
Table 2-4. PODADj values, human equivalent concentrations (HECs),
uncertainty factors, and candidate RfCs for 1,2,4-TMB
Reference
Endpoint
PODADJ
(mg/L)
HEC
(mg/m3)3
Uncertainty factors (UF)
UFA
UFH
UFL
UFS
UFD
UFcoMPOSITE
Candidate
RfC
. , 3.b
(mg/m )
Neurological endpoints
Korsakand
Rydzyriski
(1996)
Decreased pain
sensitivity
0.086
15.8
3
10
1
10
3
1,000
1.58 x 10"2
Hematological endpoints
Korsaket al.
(2000a)
Decreased RBCs
Increased WBCs
Decreased
reticulocytes
Decreased
clotting time
0.499
0.867
0.890
0.134
83.9
131.5
134.0
24.4
3
3
3
3
10
10
10
10
1
1
1
1
10
10
10
10
3
3
3
3
1,000
1,000
1,000
1,000
8.39 x 10"2
1.31 x 10"1
1.34 x 10"1
2.44 x 10"2
Respiratory endpoints
Korsak et al.
(1997)
Korsaket al.
(2000a)
Increased BAL
total cells
Increased
inflammatory
lung lesions
0.127
0.134
23.2
24.4
3
3
10
10
10
1
10
10
3
3
10,000
1,000
n/ac
2.44 x 10"2
11
12
13
14
Human equivalent concentration.
bAs calculated by application of uncertainty factors, not rounded to 1 significant digit.
°Endpoint excluded for further consideration due to a UFCOMPOSITE of 10,000. In the report, "A Review of the
Reference Dose and Reference Concentration Processes" (U.S. EPA, 2002) the RfD/RfC Technical Panel concluded
that, in cases where maximum uncertainty exists in four or more areas of uncertainty, or when the composite
uncertainty factor is 10,000 or ore, it is unlikely that the database is sufficient to derive a reference value.
Therefore, a candidate RfC based on the data for increased BAL total cells was not derived.
As stated above, the HECs were derived using a human PBPK model [Hissinketal.. 2007] to
account for interspecies differences in toxicokinetics. The human PBPK model was run (as
described in Appendix B), assuming a continuous (24 hours/day, 7 days/week) exposure, to
estimate a human PODHEc that would result from the same weekly average venous blood
This document is a draft for review purposes only and does not constitute Agency policy.
2-9 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 concentration reflected in the PODADj in animals (Table 2-3). Then, dividing this PODHEc by the
2 composite UF yields a candidate RfC.
3 Neurotoxicity is the most consistently observed endpoint in the toxicological database for
4 1,2,4-TMB. According to EPA's Guidelines for Neurotoxicity Risk Assessment [U.S. EPA, 1998], many
5 neurobehavioral changes are regarded as adverse, and the observation of correlated and replicated
6 measures of neurotoxicity strengthen the evidence for a hazard. Decreased pain sensitivity, as
7 measured as latency to paw-lick, is a measure of nociception (i.e., decreased pain sensitivity), and
8 therefore this endpoint represents an alteration in neurobehavioral function (U.S. EPA, 1998).
9 Decreased pain sensitivity was observed in multiple studies across multiple exposure durations
10 (Gralewicz and Wiaderna. 2001: Gralewicz etal.. 1997a: Korsak and Rydzynski. 1996: Korsaketal..
11 1995). and in the presence of other measures of altered neurobehavior, including impaired
12 neuromuscular function and coordination and altered cognitive function. Additionally,
13 neurotoxicological endpoints (hand tremble, weakness) were observed in worker populations
14 exposed to complex VOC mixtures containing 1,2,4-TMB, indicating a consistency and coherency of
15 effects in humans and animals following exposure to 1,2,4-TMB.
16 The U.S. EPA's Guidelines for Neurotoxicity Risk Assessment (U.S. EPA. 19981 note that effects
17 that are reversible in minutes, hours, or days after the end of exposure and appear to be associated
18 with the pharmacokinetics of the agent and its presence in the body may be of less concern than
19 effects that persist for longer periods of time after the end of exposure. Pain sensitivity was
20 observed to return to control levels 2 weeks after termination of subchronic 1,2,4-TMB exposure in
21 one study (Korsak and Rydzynski. 1996). However, in several short-term studies of TMBs, there is
22 evidence indicating that decreased pain sensitivity associated with exposure to TMBs is not rapidly
23 reversible and not associated with clearance of the chemical from the body. TMB isomers have
24 been observed to clear rapidly from blood and nervous tissues (Section B.2, Appendix B), and
25 decreased pain sensitivity persisted for up to 50-51 days after termination of short-term exposures
26 (Wiaderna etal., 2002: Gralewicz and Wiaderna, 2001: Gralewicz etal., 1997a). Taken as a whole,
27 the database does not support the characterization of decreased pain sensitivity associated with
28 exposure to 1,2,4-TMB as rapidly reversible upon clearance from the body. Given the consistency
29 of decreased pain sensitivity across independent studies and multiple durations of exposure in
30 animal studies, and the consistency of observed neurotoxicity in animals and humans, there is
31 strong evidence that neurotoxicity is a hazard associated with exposure to 1,2,4-TMB. Further,
32 decreased pain sensitivity is an adverse neurotoxic effect and thus is an appropriate effect on which
33 to base the RfC. Therefore, the candidate RfC for neurotoxicity based on decreased pain
34 sensitivity was selected as the RfC for 1,2,4-TMB.
35 A PODHEC of 15.8 mg/m3 for decreased pain sensitivity (Korsak and Rydzynski. 1996] was
36 used as the POD from which to derive the chronic RfC for 1,2,4-TMB (see Table 2-4). The
37 uncertainty factors (UFs), selected and applied in accordance with the procedures described in
38 EPA's A Review of the Reference Dose and Reference Concentration Processes (U.S. EPA, 2002]
This document is a draft for review purposes only and does not constitute Agency policy.
2-10 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 (Section 4.4.5 of the report), address five areas of uncertainty resulting in a composite UF of 1,000.
2 The selected POD was divided by this composite UF to derive the RfC.
3 An interspecies uncertainty factor, UFA, of 3 (101/2 = 3.16, rounded to 3) was applied to
4 account for uncertainty in characterizing the toxicokinetic and toxicodynamic differences between
5 rats and humans following inhalation exposure to 1,2,4-TMB. In this assessment, the use of a PBPK
6 model to convert internal doses in rats to administered doses in humans reduces toxicokinetic
7 uncertainty in extrapolating from the rat to humans, but does not account for interspecies
8 differences due to toxicodynamics. A default UFA of 3 was thus applied to account for this
9 remaining toxicodynamic and any residual toxicokinetic uncertainty not accounted for by the PBPK
10 model.
11 An intraspecies uncertainty factor, UFn, of 10 was applied to account for potentially
12 susceptible individuals in the absence of data evaluating variability of response in the human
13 population following inhalation of 1,2,4-TMB. No information is currently available to predict
14 potential variability in human susceptibility, including variability in the expression of enzymes
15 involved in 1,2,4-TMB metabolism.
16 A LOAEL to NOAEL uncertainty factor, UFu of 1 was applied because the current approach is
17 to address this factor as one of the considerations in selecting a BMR for BMD modeling. In this
18 case, a BMR equal to a change in the mean equal to 1 standard deviation of the model estimated
19 control mean for decreased pain sensitivity was selected under the assumption that this BMR
20 represents a minimal, biologically significant change for this endpoint.
21 A subchronic to chronic uncertainty factor, UFs, of 10 was applied to account for
22 extrapolation from a subchronic exposure duration study to derive a chronic RfC. The 10-fold
23 uncertainty factor is applied to the POD identified from the subchronic study on the assumption
24 that effects observed in a similar chronic study would be observed at lower concentrations for a
25 number of possible reasons, including potential cumulative damage occurring over the duration of
26 the chronic study or an increase in the magnitude or severity of effect with increasing duration of
27 exposure.
28 A database uncertainty factor, UFD, of 3 (101/2 = 3.16, rounded to 3) was applied to account
29 for database deficiencies. Strengths of the database include the three well-designed subchronic
30 studies that observe exposure-response effects in multiple organ systems (nervous, respiratory,
31 and hematological systems) in Wistar rats exposed to 1,2,4-TMB via inhalation. An additional
32 strength of the database is the well-designed developmental toxicity study that investigated
33 standard measures of maternal and fetal toxicity in a different strain of rat (Sprague-Dawley).
34 However, the lack of a multi-generation reproductive/developmental toxicity study or a
35 developmental neurotoxicity study investigating effects due to 1,2,4-TMB exposure is a weakness of
36 the database.
37 Although a multi-generation reproductive/developmental study does not exist for 1,2,4-
38 TMB, there is a multi-generation reproductive/developmental study for high flash naphtha, of
This document is a draft for review purposes only and does not constitute Agency policy.
2-11 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 which 1,2,4-TMB is a constituent. This study demonstrates effects on postnatal growth at lower
2 exposures in the F3 generation (2,460 mg/m3) compared to the F2 or Fi generation (7,380 mg/m3)
3 (McKee etal., 1990], but did not observe a consistent effect on reproductive parameters. This
4 raises some concern that addition of a multi-generation reproductive/developmental toxicity study
5 of 1,2,4-TMB might result in the identification of a lower POD.
6 EPA's Review of the Reference Dose and Reference Concentration Processes (U.S. EPA. 2002]
7 also recommends that the database uncertainty factor take into consideration whether there is
8 concern from the available toxicology database that the developing organism may be particularly
9 susceptible to effects in specific organ systems. TMBs (unspecified isomer] are able to cross the
10 placenta (Cooper etal.. 2001: Dowty etal.. 1976]: therefore, as neurotoxicity is observed in adult
11 animals, there is the concern that exposure to 1,2,4-TMB may result in neurotoxicity in the
12 developing organism. EPA's Guidelines for Neurotoxicity Risk Assessment (U.S. EPA, 1998] identifies
13 specific effects observed in adult animals (e.g., cognitive and motor function] that can also affect the
14 developing organism exposed in utero. The Neurotoxicity Guidelines (U.S. EPA. 1998] also indicate
15 that neurotoxicants may have greater access to the nervous system in developing organisms due to
16 an incomplete blood-brain barrier and immature metabolic detoxifying pathways. Therefore, there
17 is some concern that the lack of a developmental neurotoxicity study is a deficiency in the database
18 and that inclusion of such a study would potentially result in a lower POD than the POD for
19 neurotoxicity identified from the available 1,2,4-TMB toxicity database. In summary, a 3-fold
20 database UF was applied to account for the lack of both a multi-gene ration
21 reproductive/developmental toxicity study and a developmental neurotoxicity study in the
22 available database for 1,2,4-TMB.
23 Application of the composite UF of 1,000 to the PODHEc yields the following chronic RfC for
24 1,2,4-TMB:
25 RfC = PODHEC * UF = 15.8 mg/m3 4- 1,000 = 0.02 mg/m3 = 2 x 10 2 mg/m3 (rounded to
26 one significant digit)
27 2.1.4. Uncertainties in the Derivation of the Reference Concentration for 1,2,4-TMB
28 As presented above, the UF approach, following EPA practices and RfC guidance (U.S. EPA.
29 2002.1994b], was applied to the PODHEc in order to derive the chronic RfC for 1,2,4-TMB. Factors
30 accounting for uncertainties associated with a number of steps in the analyses were adopted to
31 account for extrapolation from animals to humans, a diverse human population of varying
32 susceptibilities, duration of exposure, POD determination methodologies (NOAEL, LOAEL, or
33 BMDL], and database deficiencies.
34 The critical effect selected, decreased pain sensitivity, does not introduce substantial
35 uncertainty into the RfC calculation as selection of alternative hematological or respiratory effects
36 would result in similar RfCs that would be equivalent when rounding to one significant digit (i.e., 2
This document is a draft for review purposes only and does not constitute Agency policy.
2-12 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 x ID-2 mg/m3, see Figure 2-2). Some uncertainty exists regarding the selection of the BMRs for use
2 in BMD modeling due to the absence of information to determine the biologically significant level of
3 response associated with the endpoints. However in cases such as this, the selection of a BMR of 1
4 standard deviation for continuous endpoints is supported by EPA guidance [U.S. EPA, 2000].
5 Uncertainty regarding the selection of particular models for individual endpoints exists as selection
6 of alternative models could decrease or increase the estimated POD and consequently, the RfC. The
7 selection criteria for model selection was based on a practical approach as described in EPA's
8 Benchmark Dose Technical Guidance Document [U.S. EPA, 2000]. Uncertainty may exist in the PBPK
9 model estimates of internal blood dose metrics for the rat, and subsequent HEC calculations for
10 humans, including parameter uncertainty, but such uncertainties would apply equally to all
11 endpoints.
12 2.1.5. Confidence Statement for 1,2,4-TMB
13 A confidence level of high, medium, or low is assigned to the study used to derive the RfC,
14 the overall database, and the RfC itself, as described in Section 4.3.9.2 of EPA's Methods for
15 Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry [U.S. EPA.
16 1994b].
17 Confidence in the study from which the critical effect was identified, Korsak and
18 Rydzynski (1996) is medium. The study is a well-conducted peer-reviewed study that utilized
19 three dose groups plus untreated controls, an appropriate number of animals per dose group, and
20 performed appropriate statistical analyses.
21 One area of uncertainty regarding this study is the lack of reported actual concentrations.
22 However, as the methods by which the test atmosphere was generated and analyzed were reported
23 in sufficient detail, and given the fact that this laboratory has used this methodology in subsequent
24 studies [Korsak et al., 2000a, b] and achieved appropriate actual concentrations (i.e., within 10% of
25 target concentrations], the concern regarding the lack of reported actual concentrations is minimal.
26 The critical effect on which the RfC is based is well-supported as the weight of evidence for
27 1,2,4-TMB-induced neurotoxicity is coherent across species (i.e., human and rat] and consistent
28 across multiple exposure durations (i.e., acute, short-term, and subchronic] (Gralewicz and
29 Wiaderna. 2001: Chen etal.. 1999: Wiadernaetal.. 1998: Gralewicz etal.. 1997a: Gralewicz etal..
30 1997b: Korsak and Rydzynski. 1996: Norseth etal.. 1991].
31 The database for 1,2,4-TMB includes acute, short-term, subchronic, and developmental
32 toxicity studies in rats and mice. However, confidence in the database is low to medium because
33 it lacks chronic, multi-generation reproductive/developmental, and developmental neurotoxicity
34 studies, and the studies supporting the critical effect predominantly come from the same research
35 institute. The overall confidence in the RfC for 1,2,4-TMB is low to medium.
This document is a draft for review purposes only and does not constitute Agency policy.
2-13 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1
2
3
4
5
6
7
2.1.6. Comparison of Candidate Reference Concentrations for 1,2,4-TMB
The predominant effect observed following acute, short-term, and subchronic inhalation
exposures to 1,2,4-TMB is neurotoxicity. Respiratory toxicity is observed at similar doses following
acute and subchronic exposures, while hematological effects are observed at similar doses after
subchronic exposures. Figure 2-2 provides a graphical display of all candidate PODs and RfCs
derived from the three subchronic studies considered in the selection of the POD for derivation of
the inhalation RfC for 1,2,4-TMB.
100
0.01
Decreased
Increased
pain sensitivity inflammatory
lung lesions
Neurological
Effects
Respiratory
Effects
Decreased
RBCs
Increased
WBCs
Decreased
reticulocytes
Decreased
clotting time
Hematological
Effects
D Database Uncertainty
• Interspecies Extrapolation
• Subchronic to Chronic
ORfC
D Intraspecies Variability
OPOD
9
10
11
Figure 2-2. Array of candidate PODHEC values with applied UFs and
candidate RfCs for neurological, respiratory, and hematological effects
resulting from inhalation exposure to 1,2,4-TMB.
12 2.2. Inhalation Reference Concentration for Effects Other Than Cancer for 1,2,3-
13 TMB
14 2.2.1. Identification of Candidate Principal Studies and Critical Effects for 1,2,3-TMB
15 The nervous, hematological, and respiratory systems are the primary targets of inhaled
16 1,2,3-TMB in humans and experimental animals, and effects in these systems have been identified
This document is a draft for review purposes only and does not constitute Agency policy.
2-14 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 as hazards following inhalation exposure to 1,2,3-TMB. Human data are preferred over animal data
2 for deriving reference values when possible because the use of human data is more relevant in the
3 assessment of human health and avoids the uncertainty associated with interspecies extrapolation
4 introduced when animal data serve as the basis for the RfC. In this case, while literature exists on
5 the effects of 1,2,3-TMB exposure in humans, including neurological, hematological, and respiratory
6 toxicities, no human studies are available that would allow for dose-response analysis. The human
7 studies evaluated TMB exposures occurring as complex solvents or VOC mixtures, and this
8 consideration along with other uncertainties including high imprecision in effect measures due to
9 low statistical power, lack of quantitative exposure assessment, and lack of control for co-
10 exposures, limit their utility in derivation of quantitative human health toxicity values. However,
11 these studies provide supportive evidence for the neurological, hematological, and respiratory
12 toxicity of TMB isomers in humans and indicate a coherency of effects in both humans and
13 laboratory animals.
14 Several studies investigating 1,2,3-TMB effects in experimental animal models were
15 identified in the literature. No chronic studies were available, although several acute, short-term,
16 and subchronic studies were identified. 1,2,3-TMB-induced toxicity was observed across several
17 organ systems in two subchronic studies by Korsak et al. [2000b] and Korsak and Rydzyhski
18 [1996]. These were the only subchronic studies identified in the peer-reviewed literature. Data
19 from these studies pertaining to the primary hazards observed in humans and animals identified in
20 Chapter 1 (neurological, hematological, and respiratory toxicity) were considered as candidate
21 critical effects for the purpose of determining the point of departure (POD) for derivation of the
22 inhalation RfC for 1,2,3-TMB. Neurotoxicity was also observed in both acute and short-term
23 inhalation studies and respiratory toxicity was also observed in acute studies. However, the high
24 concentrations used in acute studies and the short exposure durations of both acute and short-term
25 studies limit their applicability for quantitation of chronic human health effects. Nevertheless, as
26 with the human mixture studies, these studies provide qualitative information regarding the
27 consistency and coherency of these effects across the 1,2,3-TMB database..
28 The two subchronic studies by Korsak et al. (2000b) and Korsak and Rydzyhski (1996) are
29 adequate for dose-response analysis. Both studies used rats as an appropriate laboratory animal
30 species, and utilized appropriate sham-exposed controls. Animal were exposed tol,2,3-TMB
31 reported as > 97% pure (impurities not reported). The studies utilized an appropriate route
32 [inhaled air] and duration [subchronic] of exposure. The studies used a reasonable range of
33 appropriately-spaced exposure levels to facilitate dose-response analysis. An appropriate latency
34 between exposure and development of toxicological outcomes was used, and the persistence of
35 some outcomes after termination of exposure was investigated. Adequate numbers of animals per
36 exposure group were used, and appropriate statistical tests including pair-wise and trend analyses
37 were performed. With regard to reporting of exposure methodologies, Korsak et al. (2000b)
38 reported actual concentrations, as measured by gas chromatography, to be within 10% of target
This document is a draft for review purposes only and does not constitute Agency policy.
2-15 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1
2
3
4
5
6
7
concentrations. This increases the confidence in the overall evaluation and adequacy of this study.
Although Korsak and Rydzyhski [1996] do not report actual, measured concentrations, this study
uses the same exposure methodology as Korsak et al. [2000b]: suggesting that it is likely that the
actual concentrations in this study are within 10% of target concentrations. Target and actual
concentrations for these studies are listed in Table 2-5.
Table 2-5. Target and actual exposure concentrations used in BMD
modeling of 1,2,3-TMB endpoints considered for the derivation of the
RfC
Reference
Korsak and
Rydzyriski (1996)
Korsak et al.
(2000b)
Species/
sex
Rat, male
Rat, male
Rat, female
Target exposure concentration
(mg/m3)
123
492
1,230
123
492
1,230
123
492
1,230
Actual exposure concentration
(mg/m3)
n/a
n/a
n/a
128
523
1,269
128
523
1,269
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
These subchronic studies examined 1,2,3-TMB-induced toxicity in multiple organ systems
and the neurological, hematological, and respiratory endpoints that demonstrated statistically
significant pair-wise increases or decreases relative to control were considered for the derivation of
the RfC for 1,2,3-TMB (Table 2-6). These endpoints included decreased pain sensitivity in male rats
[Korsak and Rydzynski, 1996], and decreased RBCs and increased reticulocytes in male rats,
decreased segmented neutrophils and increased lymphocytes in male and female rats, and
increased inflammatory lung lesions in female rats [Korsak etal., 2000b]. Changes in liver organ
weights and clinical chemistry parameters from Korsak et al. [2000b] were not further considered
due to the lack of accompanying hepatocellular histopathological alterations in exposed animals.
Changes in splenic organ weights were similarly not considered further due to a lack of any
observed histopathological changes in that organ. Increases in reticulocytes in females were not
further considered due to non-monotonicity in response (increases in high concentration animals,
not statistically significant]. Increased lymphocytes were excluded from further consideration due
to the unusually high standard deviations reported in the high-concentration group.
This document is a draft for review purposes only and does not constitute Agency policy.
2-16 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 2-6. Endpoints resulting from subchronic inhalation exposure to
1,2,3-TMB considered for the derivation of the RfC
Endpoint
Species/sex
Exposure concentration (mg/ms)a
0
123
492
1,230
Neurological endpoints
Decreased pain sensitivity (measured as
latency to paw-lick in seconds)b
Rat, male
9.7 ±2.1
(n = 30)
11.8 ±3.8
(n = 20)
16.3 ± 6.3°
(n = 10)
17.3 ±3.4
(n = 10)
Hematological endpoints
Decreased RBCs (106/cm3)d
Decreased segmented neutrophils (%)d
Increased reticulocytes (%)d
Rat, male
Rat, male
Rat, female
Rat, male
9.49 ± 2.03
(n = 10)
24.8 ±4.5
(n = 10)
23.1 ±6.1
(n = 10)
2.8 ±1.3
(n = 10)
10.2 ± 1.29
(n = 10)
25.4 ±5.8
(n = 10)
19.7 ±3.4
(n = 10)
2.1 ±1.7
(n = 10)
10.11 ±1.27
(n = 10)
20.7 ±5.8
(n = 10)
16.4 ±4.2
(n = 10)
3.8 ±2.1
(n = 10)
8.05 ± 1.38
(n = 10)
17.7 ±8.3
(n = 10)
11.9 ±7.1
(n = 10)
4.5 ±1.8
(n = 10)
Respiratory Endpoints
Increased inflammatory lung lesionsd
Rat, female
e
(n = 10)
e
(n =10)
e
(n = 10)
e
(n = 10)
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
p<0.05; p<0.01.
aValues are expressed as mean ± 1SD.
b Adapted from Korsak and Rydzyriski (1996)
c Level of significance not reported in Table 1 from Korsak and Rydzyriski (1996), however the results of an ad-hoc t-
test (performed by EPA) indicated significance at p < 0.01.
d Adapted from Korsak et al. (2000b)
e Incidences for exposure groups not reported; however, based on qualitative information reported in the study (i.e.,
that female rats exhibited a statistically significant increase in inflammatory lung lesions at 492 mg/m3), a NOAEL of
123 mg/m3 was identified.
Impaired neuromuscular function and coordination, measured as performance on the
rotarod apparatus, was also observed in rats exposed to 1,2,3-TMB. The use of rotarod data from
Korsak and Rydzyhski [1996] was initially considered as a candidate critical effect for 1,2,3-TMB.
However, upon critical evaluation of the exposure-response information in the study it was
determined that the endpoint was reported in a manner that reduced the confidence in the
observed effect levels. The primary limitation noted for these data relates to the presentation of
rotarod performance, which is best represented as a continuous variable, as opposed to a quantal
variable such as that presented by Korsak and Rydzyhski [1996]. In contrast to the percent failures
reported by the study authors, the most widely used and accepted measurement for rotarod
performance in rodents is latency to fall from the rotating rod [Brooks and Dunnett. 2009: Kaspar
etal.. 2003: Bogo etal., 1981], typically with an arbitrary upper limit on the maximum latency
allowed to prevent confounding by fatigue. Although the quantal percent failures data can provide
useful information, these measures require an arbitrary selection of the length of time required for
successful performance; there is no scientific consensus on an optimal time for this parameter. In
addition, when identifying effect levels based on the data presented by Korsak and Rydzyhski
[1996], latencies on the rod of 1 and 119 seconds would be treated identically as failures when, in
This document is a draft for review purposes only and does not constitute Agency policy.
2-17 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
fact, they indicate very different levels of neurological dysfunction [Bogo etal., 1981]. This adds
uncertainty when trying to extrapolate to a concentration associated with a minimally adverse
effect Finally, quantal presentation of data does not allow for interpretations related to intra-rat
and intra-group variability in performance. Due to these reporting limitations, impaired
neuromuscular function and coordination, measured as performance on the rotarod apparatus, was
excluded from consideration for derivation of the RfC for 1,2,3-TMB.
Endpoints carried forward for derivation of an RfC for 1,2,3-TMB, along with their exposure
ranges and NOAEL/LOAEL values (identified by EPA), are graphically represented in Figure 2-3.
10000
Concentrations
ONOAEL
DLOAEL
oo
£
c
o
"5
to
0)
1000
100
pain sensitivity
(a)
RBCs
males
(b)
lung
lesions
(b)
Seg.
Neutrophils
males
(b)
Seg.
Neutrophils
females
(b)
reticulo-
cytes
(b)
Solid lines represent range of exposure concentrations, (a) Korsak and Rydzyriski (1996); (b) Korsak et al.
(2000b).
Figure 2-3. Exposure response array for endpoints resulting from
inhalation exposure to 1,2,3-TMB considered for the derivation of the
RfC.
2.2.2. Methods of Analysis for 1,2,3-TMB
As discussed above in Section 2.2.1, endpoints observed in Korsak et al. (2000b] and Korsak
and Rydzyhski (1996] that demonstrated statistically significant (p < 0.05 level] pair-wise increases
or decreases relative to control for at least one exposure group were considered for the derivation
of the RfC for 1,2,3-TMB; these effects are listed in Table 2-5. This assessment used the BMD
approach, when possible, to estimate a POD for the derivation of an RfC for 1,2,3-TMB (see Section
This document is a draft for review purposes only and does not constitute Agency policy.
2-18 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 C.I of Appendix C for detailed methodology). The BMD approach involves fitting a suite of
2 mathematical models to the observed dose-response data using EPA's BMDS (version 2.2). Each
3 fitted model estimates a BMD and its associated BMDL corresponding to a selected BMR. For
4 continuous data (i.e., decreased pain sensitivity, decreased RBCs, decreased segmented neutrophils,
5 increased reticulocytes) from the Korsak and Rydzyhski (1996) and Korsak et al. (2000b) studies,
6 no information is available regarding the change in these responses that would be considered
7 biologically significant, and thus a BMR equal to a change in the mean equal to 1 standard deviation
8 of the model estimated control mean was used in modeling the endpoints, consistent with the
9 Benchmark Dose Technical Guidance Document (U.S. EPA. 2000). The estimated BMDL is then used
10 as the POD for deriving the RfC (Table 2-7).
11 The suitability of the above methods to determine a POD is dependent on the nature of the
12 toxicity database for a specific chemical. Some endpoints for 1,2,3-TMB were not modeled for a
13 variety of reasons, including responses only in the high exposure group with no changes in
14 responses in lower exposure groups (e.g., decreased RBCs) and absence of incidence data (e.g.,
15 increased inflammatory lung lesions). In cases where BMD modeling was not feasible, the
16 NOAEL/LOAEL approach was used to identify a POD. Additionally, for decreased pain sensitivity,
17 the reported SD of 3.4 in the high exposure group resulted in an inability of the variance power
18 model to fit the data adequately. For this reason, the high exposure group was dropped in order to
19 facilitate model fitting. Detailed modeling results are provided in Section C.I of Appendix C.
20 Because an RfC is a toxicity value that assumes continuous human inhalation exposure over
21 a lifetime, data derived from inhalation studies in animals need to be adjusted to account for the
22 noncontinuous exposures used in these studies. In the Korsak et al. (2000b) and Korsak and
23 Rydzyhski (1996) studies, rats were exposed to 1,2,3-TMB for 6 hours/day, 5 days/week for 3
24 months. Because no PBPK model exists for 1,2,3-TMB, the duration-adjusted PODs for effects in
25 rats were calculated as follows:
26 PODADJ (mg/m3) = POD (mg/m3) x hours exposed per day/24 hours x days exposed per
27 week/7 days
28 Therefore, for example, for decreased pain sensitivity from Korsak and Rydzyhski (1996),
29 the PODADj would be calculated as follows:
30 PODADJ (mg/m3) = 97.19 mg/m3x 6 hours/24 hours x 5 days/7 days
31 PODADj (mg/m3) = 17.36 mg/m3
32 The calculated PODADj (mg/m3) values for all neurological, hematological, and respiratory
33 endpoints considered for RfC derivation are presented in Table 2-7.
This document is a draft for review purposes only and does not constitute Agency policy.
2-19 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 2-7. Summary of dose-response analysis and point of departure
estimation for endpoints resulting from subchronic inhalation
exposure to 1,2,3-TMB
Reference
Endpoint
Species/
sex
POD
basis
Best-fit
model; BMR
Candidate
POD
(mg/m3)
Candidate
PODAD/
(mg/m3)
Neurological endpoints
Korsak and
Rydzyriski (1996)
Decreased pain
sensitivity
Rat, male
BMDL
Linear; ISO
97.19
17.36
Hematological endpoints
Korsak et al. (2000b)
Decreased RBCs
Decreased
segmented
neutrophils
Increased
reticulocytes
Rat, male
Rat, male
Rat, female
Rat, male
NOAEL
BMDL
BMDL
BMDL
n/ab
Exponential
2; ISO
Hill;lSD
Linear; ISO
523
534.81
99.21
652.90
93.39
95.50
17.72
116.58
Respiratory endpoints
Korsak et al. (2000b)
Increased
inflammatory
lung lesions
Rat, female
NOAEL
n/ab
128
22.86
"Duration adjusted PODADJ (mg/m ) = POD x (6 hours/24 hours) x (5 days/7 days) (U.S. EPA, 2002).
bNo model was able to fit data adequately, or data were not modeled.
4 2.2.3. Derivation of the Reference Concentration for 1,2,3-TMB
5 Because the majority of the selected endpoints for consideration as the critical effect
6 (decreased pain sensitivity, decreased RBCs, decreased segmented neutrophils, increased
7 reticulocytes) result primarily from systemic distribution of 1,2,3-TMB, and no available PBPK
8 model exists for 1,2,3-TMB, the human equivalent concentration (HEC) for 1,2,3-TMB was
9 calculated by the application of the dosimetric adjustment factor (DAF) for systemically acting gases
10 (i.e., Category 3 gases), in accordance with the U.S. EPA RfC Methodology (U.S. EPA. 1994b).
11 Additionally, although the observation of lung lesions would normally indicate portal-of-entry
12 effects, the observation that the overwhelming majority of 1,2,3-TMB-induced effects are systemic
13 in nature supports the determination that 1,2,3-TMB is a Category 3 gas. Other factors also support
14 that 1,2,3-TMB is a systemically-acting toxicant, including the isomer's relatively low water-
15 solubility and non-reactivity. Gases with these properties are expected to preferentially distribute
16 to the lower regions of the respiratory tract where larger surface areas and thin alveolar-capillary
17 boundaries facilitate uptake. Respiratory absorption of 1,2,3-TMB into the bloodstream has been
18 observed to be relatively high (~60%) following inhalation exposures to humans (Tarnbergetal..
19 1996). Therefore, increased inflammatory lung lesions are assumed to result from systemic
This document is a draft for review purposes only and does not constitute Agency policy.
2-20 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 distribution of 1,2,3-TMB in the bloodstream of exposed animals. DAFs are ratios of animal and
2 human physiologic parameters, and are dependent on the nature of the contaminant (particle or
3 gas) and the target site (e.g., respiratory tract or remote to the portal-of-entry [i.e., systemic]) (U.S.
4 EPA. 1994b). For gases with systemic effects, the DAF is expressed as the ratio between the animal
5 and human blood:air partition coefficients:
6 DAF = (Hb/g)A/(Hb/g)H
7 DAF = 62.6/66.5
8 DAF = 0.94
9 where:
10 (Hb/g)A = the animal bloochair partition coefficient
11 (Hb/g)H = the human bloochair partition coefficient
12 In cases where the animal blood:air partition coefficient is lower than the human value
13 (Meulenberg and Vijverberg, 2000: Tarnberg and Tohanson, 1995), resulting in a DAF < 1, the
14 calculated value is used for dosimetric adjustments (U.S. EPA. 1994b). For example, the HEC for
15 decreased pain sensitivity reported in Korsak and Rydzyhski (1996) is calculated as follows:
16 PODHEc = PODADj (mg/m3) x DAF
17 PODHEC = PODADj (mg/m3) x 0.94
18 PODHEC = 17.36 mg/m3 x 0.94
19 PODHEC = 16.32 mg/m3
20 Table 2-8 presents the calculated HECs for the candidate critical effects, selected
21 uncertainty factors (UFs), and the resulting derivation of candidate RfCs from the two subchronic
22 toxicity studies (Korsak etal., 2000b: Korsak and Rydzynski, 1996).
23
This document is a draft for review purposes only and does not constitute Agency policy.
2-21 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 2-8. PODADj values, human equivalent concentrations (HECs),
uncertainty factors, and candidate RfCs for 1,2,3-TMB
Reference
Endpoint
PODADJ
(mg/m3)
HEC
(mg/m3)3
Uncertainty factors (UF)
UFA
UFH
UFL
UFS
UFD
UFcoMPOSITE
Candidate
RfC
(mg/m3)b
Neurological endpoints
Korsak and
Rydzyriski
(1996)
Decreased pain
sensitivity
17.36
16.32
3
10
1
10
3
1,000
1.63 x 10"2
Hematological effects
Korsak et al.
(2000b)
Decreased RBCs
Decreased
segmented
neutrophils, males
Decreased
segmented
neutrophils,
females
Increased
reticulocytes
93.39
95.50
17.72
116.58
87.79
89.77
16.66
109.58
3
3
3
3
10
10
10
10
1
1
1
1
10
10
10
10
3
3
3
3
1,000
1,000
1,000
1,000
8.78 x 10"2
8.98 x 10"2
1.67 x 10"2
1.10 x 10 1
Respiratory effects
Korsak et al.
(2000b)
Increased
inflammatory lung
lesions
22.86
21.49
3
10
1
10
3
1,000
2.15 x 10"2
3
4
5
6
7
8
9
10
11
12
13
14
15
aHuman equivalent concentration.
bAs calculated by application of uncertainty factors, not rounded to 1 significant digit.
Neurotoxicity is the most consistently observed endpoint in the toxicological database for
1,2,3-TMB. According to EPA's Guidelines for Neurotoxicity Risk Assessment [U.S. EPA, 1998], many
neurobehavioral changes are regarded as adverse, and the observation of correlated and replicated
measures of neurotoxicity strengthen the evidence for a hazard. Decreased pain sensitivity, as
measured as latency to paw-lick, is a measure of nociception (i.e., decreased pain sensitivity) and
therefore this endpoint represents an alteration in neurobehavioral function [U.S. EPA, 1998].
Decreased pain sensitivity was observed in two studies investigating short-term and subchronic
exposure durations [Wiadernaetal.. 1998: Korsak and Rydzynski. 1996] and in the presence of
other metrics of altered neurobehavior, including impaired neuromuscular function and
coordination and altered cognitive function. Additionally, neurotoxicological endpoints (hand
tremble, weakness] are observed in human worker populations exposed to complex VOC mixtures
containing 1,2,3-TMB, indicating a consistency and coherency of effects in humans and animals
following exposure to 1,2,3-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
2-22 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 The U.S. EPA's Guidelines for Neurotoxicity Risk Assessment [U.S. EPA. 1998] note that effects
2 that are reversible in minutes, hours, or days after the end of exposure and appear to be associated
3 with the pharmacokinetics of the agent and its presence in the body may be of less concern than
4 effects that persist for longer periods of time after the end of exposure. Pain sensitivity was
5 observed to return to control levels 2 weeks after termination of subchronic inhalation exposure in
6 one study [Korsak and Rydzynski. 1996). However, in short-term studies of TMBs, there is
7 evidence indicating that decreased pain sensitivity associated with exposure to TMBs is not rapidly
8 reversible and not associated with clearance of the chemical from the body. TMB isomers have
9 been observed to clear rapidly from blood and nervous tissues (Section B.2, Appendix B), and
10 decreased pain sensitivity persisted for up to 50-51 days after termination of short-term exposures
11 to 1,2,3-TMB fWiadernaetal.. 19981. Short-term neurotoxicity studies of the related 1,2,4-TMB
12 isomer also reported a persistence of decreased pain sensitivity after termination of exposure
13 [Gralewicz and Wiaderna. 2001: Gralewicz etal.. 1997}. Taken as a whole, the database does not
14 support the characterization of decreased pain sensitivity associated with exposure to 1,2,3-TMB as
15 rapidly reversible upon clearance from the body. Given the consistency of decreased pain
16 sensitivity across independent studies and multiple durations of exposure in animal studies, and
17 the consistency of observed neurotoxicity in animals and humans, there is strong evidence that
18 neurotoxicity is a hazard associated with exposure to 1,2,3-TMB. Further, decreased pain
19 sensitivity is an adverse neurotoxic effect and thus is an appropriate effect on which to base the RfC.
20 Therefore, the candidate RfC for neurotoxicity based on decreased pain sensitivity was
21 selected as the RfC for 1,2,3-TMB.
22 A PODnEc of 16.3 mg/m3 for decreased pain sensitivity [Korsak and Rydzynski, 1996] was
23 used as the POD to derive the chronic RfC for 1,2,3-TMB. The uncertainty factors (UFs], selected
24 and applied in accordance with the procedures described in EPA's A Review of the Reference Dose
25 and Reference Concentration Processes [U.S. EPA. 2002] (Section 4.4.5 of the report], address five
26 areas of uncertainty resulting in a composite UF of 1,000. This composite UF was applied to the
27 selected POD to derive an RfC.
28 An interspecies uncertainty factor, UFA, of 3 (101/2 = 3.16, rounded to 3] was applied to
29 account for uncertainty in characterizing the toxicokinetic and toxicodynamic differences between
30 rats and humans following inhalation exposure to 1,2,3-TMB. In this assessment, the use of a DAF
31 to extrapolate external concentrations from rats to humans reduces toxicokinetic uncertainty in
32 extrapolating from the rat data, but does not account for the possibility that humans may be more
33 sensitive to 1,2,3-TMB than rats due to toxicodynamic differences. A default UFA of 3 was thus
34 applied to account for this remaining toxicodynamic and residual toxicokinetic uncertainty not
35 accounted for in the DAF.
36 An intraspecies uncertainty factor, UFn, of 10 was applied to account for potentially
37 susceptible individuals in the absence of data evaluating variability of response in the human
38 population following inhalation of 1,2,3-TMB. No information is currently available to predict
This document is a draft for review purposes only and does not constitute Agency policy.
2-23 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 potential variability in human susceptibility, including variability in the expression of enzymes
2 involved in 1,2,3-1MB metabolism.
3 A LOAEL to NOAEL uncertainty factor, UFu of 1 was applied because the current approach is
4 to address this factor as one of the considerations in selecting a BMR for BMD modeling. In this
5 case, a BMR equal to a change in the mean equal to 1 standard deviation of the model estimated
6 control mean for decreased pain sensitivity was selected under the assumption that this BMR
7 represents a minimal, biologically significant change for this endpoint.
8 A subchronic to chronic uncertainty factor, UFs, of 10 was applied to account for
9 extrapolation from a subchronic exposure duration study to derive a chronic RfC. The 10-fold
10 uncertainty factor is applied to the POD identified from the subchronic study on the assumption
11 that effects observed in a similar chronic study would be observed at lower concentrations for a
12 number of possible reasons, including potential cumulative damage occurring over the duration of
13 the chronic study or an increase in the magnitude or severity of effect with increasing duration of
14 exposure.
15 A database uncertainty factor, UFD, of 3 (101/2 = 3.16, rounded to 3) was applied to account
16 for database deficiencies. Strengths of the database include the two well-designed subchronic
17 studies that observe exposure-response effects in multiple organ systems (i.e., neurological,
18 hematological, and respiratory effects) in Wistar rats exposed to 1,2,3-TMB via inhalation.
19 However, the lack of a either a multi-generational reproductive/developmental toxicity study or a
20 developmental toxicity study investigating effects due to 1,2,3-TMB exposure is a weakness of the
21 database. Normally, the lack of both of these types of studies in a toxicity database would warrant
22 the application of a full, 10-fold UFo in accordance with EPA's Review of the Reference Dose and
23 Reference Concentration Processes [2002]. Although there is no developmental toxicity study for
24 1,2,3-TMB, Saillenfait et al. [2005] investigates the developmental toxicity of the other two TMB
25 isomers (1,2,4-TMB and 1,3,5-TMB] and observes developmental toxicity at levels much higher
26 than those eliciting neurotoxicity, hematotoxicity, and respiratory toxicity in adult animals (Korsak
27 studies]. Given that toxic effects were observed at lower concentrations in adult animals exposed
28 1,2,4-TMB and 1,3,5-TMB compared with rats exposed in utero and the similarities in toxicity
29 profiles amongst the three isomers, it is unlikely that the inclusion of a developmental toxicity study
30 for 1,2,3-TMB would result in a POD that is lower than the POD associated with neurotoxicity for
31 this isomer. Thus, the application of an UF to account for the lack of a developmental toxicity study
32 is not warranted.
33 Although a multi-generation reproductive/developmental study does not exist for 1,2,3-
34 TMB, there is a multi-generation reproductive/developmental study for high flash naphtha, of
35 which 1,2,3-TMB is a constituent. This study demonstrates effects on postnatal growth at lower
36 exposures in the Fs generation (2,460 mg/m3] compared to the ¥2 or Fi generation (7,380 mg/m3]
37 (McKee etal., 1990], but did not observe a consistent effect on reproductive parameters. This
This document is a draft for review purposes only and does not constitute Agency policy.
2-24 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 raises some concern that addition of a multi-generation reproductive/developmental toxicity study
2 for 1,2,3-TMB might result in the identification of a lower POD.
3 EPA's Review of the Reference Dose and Reference Concentration Processes [U.S. EPA, 2002]
4 also recommends that the database uncertainty factor take into consideration whether there is
5 concern from the available toxicology database that the developing organism may be particularly
6 susceptible to effects in specific organ systems. TMBs (unspecified isomer) are able to cross the
7 placenta [Cooper etal.. 2001: Dowtyetal.. 1976]: therefore, as neurotoxicity is observed in adult
8 animals, there is concern that exposure to 1,2,3-TMB may result in neurotoxicity in the developing
9 organism. EPA's Guidelines for Neurotoxicity Risk Assessment [U.S. EPA. 1998] identifies specific
10 effects observed in adult animals (e.g., cognitive and motor function] that can also affect the
11 developing organism exposed in utero. The Neurotoxicity Guidelines (U.S. EPA, 1998] also indicate
12 that neurotoxicants may have greater access to the nervous system in developing organisms due to
13 an incomplete blood-brain barrier and immature metabolic detoxifying pathways. Therefore, there
14 is some concern that the lack of a developmental neurotoxicity study is a deficiency in the database
15 and that the inclusion of such a study would potentially result in a lower POD than the POD for
16 neurotoxicity identified from the available 1,2,3-TMB toxicity database. In summary, a 3-fold
17 database UF was applied to account for the lack of both a multi-gene ration
18 reproductive/developmental toxicity study and a developmental neurotoxicology study in the
19 available database for 1,2,3-TMB.
20 Application of this composite UF of 1000 to the PODHEc yields the following chronic RfC for
21 1,2,3-TMB:
22 RfC = PODHEc * UF = 16.3 mg/m3 4- 1,000 = 0.02 mg/m3 = 2 x 10 2 mg/m3 (rounded to
23 one significant digit)
24 2.2.4. Uncertainties in the Derivation of the Reference Concentration for 1,2,3-TMB
25 As presented above, the UF approach following EPA practices and RfC guidance (U.S. EPA.
26 2002.1994b], was applied to the PODHEc in order to derive the chronic RfC for 1,2,3-TMB. Factors
27 accounting for uncertainties associated with a number of steps in the analyses were adopted to
28 account for extrapolation from animals to humans, a diverse human population of varying
29 susceptibilities, duration of exposure, POD determination methodologies (NOAEL, LOAEL, or
30 BMDL], and database deficiencies.
31 The critical effect selected, decreased pain sensitivity, does not introduce substantial
32 variability into the RfC calculation as selection of alternative hematological or respiratory effects
33 would result in similar RfCs that would be equivalent when rounding to one significant digit (i.e., 2
34 x 10'2 mg/m3, see Figure 2-4]. Some uncertainty exists regarding the selection of the BMRs for use
35 in BMD modeling due to the absence of information to determine the biologically significant level of
36 response associated with the endpoints. However in cases such as this, the selection of a BMR of 1
This document is a draft for review purposes only and does not constitute Agency policy.
2-25 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 standard deviation for continuous endpoints is supported by EPA guidance [U.S. EPA, 2000].
2 Uncertainty regarding the selection of particular models for individual endpoints exists as selection
3 of alternative models could decrease or increase the estimated POD and consequently, the RfC. The
4 criteria for model selection was based on a practical approach as described in EPA's Benchmark
5 Dose Technical Guidance Document [U.S. EPA. 2000]. Uncertainty may exist in the default dosimetry
6 methods used to calculate HEC estimates, but such uncertainties would apply equally to all
7 endpoints.
8 2.2.5. Confidence Statement for 1,2,3-TMB
9 Confidence in the study from which the critical effect was identified, Korsak and
10 Rydzynski (1996) is medium. The study is a well-conducted, peer-reviewed study that utilized
11 three dose groups plus untreated controls, an appropriate number of animals per dose group, and
12 appropriately performed statistical analyses.
13 One area of uncertainty regarding this study is the lack of reported actual concentrations.
14 However, as the methods by which the test atmosphere was generated and analyzed were reported
15 in sufficient detail, and given the fact that this laboratory has used this methodology in subsequent
16 studies [Korsak et al.. 2000a. b] and achieved appropriate actual concentrations (i.e., within 10% of
17 target concentrations], the concern regarding the lack of reported actual concentrations is minimal.
18 The critical effect on which the RfC is based is well-supported as the weight of evidence for 1,2,3-
19 TMB-induced neurotoxicity is coherent across multiple animals species (i.e., mouse, and rat] and
20 consistent across multiple exposure durations (i.e., acute, short-term, and subchronic] (Lutz etal..
21 2010: Wiadernaetal.. 1998: Korsak and Rydzynski. 1996].
22 The database for 1,2,3-TMB includes acute, short-term, and subchronic toxicity studies in
23 rats and mice. However, confidence in the database is low to medium because it lacks chronic,
24 multi-generation reproductive/developmental, developmental toxicity, or developmental
25 neurotoxicity studies, and the studies supporting the critical effect predominantly come from the
26 same research institute. The overall confidence in the RfC for 1,2,3-TMB is low to medium.
27 2.2.6. Comparison of Candidate Reference Concentrations for 1,2,3-TMB
28 The predominant effect observed following acute, short-term, and subchronic inhalation
29 exposures to 1,2,3-TMB is neurotoxicity. Respiratory toxicity is observed at similar doses following
30 acute and subchronic exposures, while hematological effects are observed at similar doses after
31 subchronic exposures. Figure 2-4 provides a graphical display of all candidate PODs and RfCs
32 derived from the two subchronic studies considered in the selection of the POD for the inhalation
33 RfC for 1,2,3-TMB.
34
This document is a draft for review purposes only and does not constitute Agency policy.
2-26 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1
2
o
6
4
100
0.01
Decreased Increased
pain sensitivity inflammatory
lung lesions
Neurological
Effects
Respiratory
Effects
Decreased
RBCs
Decreased
Segmented
Neutrophils
males
Decreased
Segmented
Neutrophils
females
Decreased
Reticulocytes
D Database Uncertainty •Subchronic to Chronic
• Interspecies Extrapolation ORfC
Hematological
Effects
D Intraspecies Variability
CPOD
Figure 2-4. Array of candidate PODHEC values with applied UFs and
candidate RfCs for neurological respiratory, and hematological effects
resulting from inhalation exposure to 1,2,3-TMB.
5 2.3. Inhalation Reference Concentration for Effects Other Than Cancer for 1,3,5-
6 TMB
7 2.3.1. Identification of Candidate Principal Studies and Critical Effects for 1,3,5-TMB
8 The nervous, hematological, and respiratory systems are the primary targets for inhaled
9 1,3,5-TMB in humans, whereas the nervous system in adults, pregnant females, and developing
10 organism are the primary targets of inhaled 1,3,5-TMB in experimental animals. Effects in these
11 systems have been identified as hazards following inhalation exposures to 1,3,5-TMB. Human data
12 are preferred over animal data for deriving reference values when possible because the use of
13 human data is more relevant in the assessment of human health and avoids the uncertainty
14 associated with interspecies extrapolation introduced when animal data serve as the basis for the
15 RfC. In this case, while literature exists on the effects of 1,3,5-TMB exposure in humans, including
16 neurological, hematological, and respiratory toxicities, no human studies are available that would
17 allow for dose-response analysis. The human studies evaluated TMB exposures occurring as
This document is a draft for review purposes only and does not constitute Agency policy.
2-27 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 complex solvents or VOC mixtures, and this consideration along with similar uncertainties as
2 discussed for 1,2,4-TMB and 1,2,3-TMB limit their utility in derivation of quantitative human health
3 toxicity values. As for the other two isomers, the human studies provide supportive evidence for
4 the neurological toxicity of 1,3,5-TMB in humans and indicate a consistency and coherency of this
5 effect in humans and laboratory animals.
6 Several studies investigating 1,3,5-TMB effects in experimental animals models were
7 identified in the literature. No chronic or subchronic inhalation studies were identified. However,
8 1,3,5-TMB-induced toxicity was observed in two short-term inhalation studies [Wiadernaetal.,
9 2002: Gralewicz and Wiaderna. 2001] investigating neurotoxicity outcomes in adult animals and in
10 one developmental toxicity study investigating maternal and fetal toxicity [Saillenfaitetal.. 2005].
11 Data from these studies pertaining to the primary hazards observed in humans (neurological
12 effects] and animals (neurological and maternal/developmental effects] were considered as
13 candidate critical effects for the purpose of determining the point of departure (POD] for derivation
14 of the inhalation RfC for 1,3,5-TMB. Neurotoxicity and respiratory toxicities were also observed in
15 acute inhalation studies. However, the high concentrations used in acute studies limit their
16 applicability for quantitation of chronic human health effects. Nevertheless, as with the human
17 mixture studies, these studies provide qualitative information regarding the consistency and
18 coherency of these effects across the 1,3,5-TMB database.
19 The two short-term studies by Gralewicz and Wiaderna (2001) and Wiaderna et al. (2002],
20 and the developmental toxicity study by Saillenfait et al. (2005] are adequate for dose-response
21 analysis. Both studies used rats as an appropriate laboratory animal species, and utilized
22 appropriate sham-exposed controls. Animals were exposed to 1,3,5-TMB reported as 99% pure
23 (impurities not reported]. These studies utilized an appropriate route [inhaled air] and duration
24 [short-term and gestational] of exposure. Although the duration for short-term studies was not
25 optimal, in that studies of this duration are not usually considered for derivation of chronic
26 reference values, these studies were considered appropriate for derivation of an RfC for 1,3,5-TMB
27 given the lack of any subchronic inhalation studies in adult rats. The studies used a reasonable
28 range of appropriately-spaced exposure levels to facilitate dose-response analysis. An appropriate
29 latency between exposure and development of toxicological outcomes was used, and the
30 persistence of some outcomes (neurotoxicity effects] after termination of exposure was
31 investigated. Adequate numbers of animals per exposure group were used, and appropriate pair-
32 wise statistical tests were performed. With regard to reporting of exposure methodologies,
33 Saillenfait et al. (2005] reported actual concentrations, as measured by gas chromatography, to be
34 within 10% of target concentrations. This increases the confidence in the overall evaluation and
35 adequacy of this study. Although neither Wiaderna et al. (2002] nor Gralewicz and Wiaderna
36 (2001] explicitly report actual concentration, they cite previous work from the same research
37 institute that demonstrated the methodology was capable of achieving target concentrations;
This document is a draft for review purposes only and does not constitute Agency policy.
2-28 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1
2
3
4
5
6
suggesting that it is likely that the actual concentrations in this study are within 10% of target
concentrations. Target and actual concentrations are listed in Table 2-9.
Table 2-9. Target and actual exposure concentrations used in BMD
modeling of 1,3,5-TMB endpoints considered for the derivation of the
RfC
Reference
Gralewicz and
Wiaderna (2001);
Wiaderna et al.
(2002)
Saillenfait et al.
(2005)
Species/
sex
Rat, male
Rat, female
(pregnant dam);
male and female
(fetuses)
Target exposure concentration
(mg/m3)
123
492
1,230
492
1,476
2,952
5,904
Actual exposure concentration
(mg/m3)
n/a
n/a
n/a
497
1,471
2,974
5,874
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Gralewicz and Wiaderna [2001] and Wiaderna et al. [2002] both observed altered cognitive
function, decreased pain sensitivity, and decreased anxiety and/or increased motor function
following inhalation exposure to 1,3,5-TMB (see Table 2-10]. Wiaderna et al. [2002] reported that
123 mg/m3 was the LOAEL for altered cognitive function and the NOAEL for decreased pain
sensitivity. As altered cognitive function was observed at a lower concentration than decreased
pain sensitivity, only altered cognitive function was further considered for derivation of an RfC for
1,3,5-TMB from the Wiaderna etal. [2002] study. All three neurotoxic effects (altered cognitive
function, decreased pain sensitivity, and decreased anxiety and/or increased motor function] were
observed at the only concentration utilized in the Gralewicz and Wiaderna [2001] (i.e., 492
mg/m3]; these LOAELs were further considered for derivation of an RfC for 1,3,5-TMB. From the
Saillenfait et al. [2005] study, decreased male and female fetal weights and decreased corrected
maternal weight gain were considered for derivation of the RfC. Changes in serum chemistry
parameters in rats exposed to 1,3,5-TMB in a short-term (5 weeks] inhalation study [Wiglusz etal..
1975b] were not considered for derivation of the RfC due to inconsistent temporal patterns of
effect and the lack of accompanying histopathology. Endpoints carried forward for derivation of an
RfC for 1,3,5-TMB, along with their NOAEL and LOAEL values, are graphically represented in Figure
2-5.
This document is a draft for review purposes only and does not constitute Agency policy.
2-29 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1
2
Table 2-10. Endpoints resulting from inhalation exposure to 1,3,5-TMB
considered for the derivation of the RfC
Endpoint
Species/sex
Exposure concentration (mg/m3)
0
492 1,476 2,952 5,904
Developmental endpoints
Decreased fetal weight
(g)a
Rat, male
Rat, female
5.80 ±
0.41b'c
5.50 ±0.32
5.76 ±0.27 5.50 ±0.31 5.39 ± 0.55* j^°»
5.74 ±0.21 5.27 ±0.47 5.18 ±0.68 ^«
Maternal endpoints
Decreased maternal
weight gain (g)a
Rat, female
29 ±14
(n = 21)d
30 ±9 20 ±12 7 ±20* -12 ± 19**
(n = 22) (n = 21) (n = 17) (n = 18)
Neurological endpoints
Endpoint
Altered cognitive
function6
Altered cognitive
function8
Decreased pain
sensitivity8
Decreased anxiety and/or
increased motor function8
Species/sex
Rat, male
Rat, male
Rat, male
Rat, male
Exposure concentration (mg/m3)
0
0'
(n = 12)
0
(n = 11)
0
(n = 11)
0
(n = 11)
123
40
(n = 12)
--
--
--
492
35
(n = 12)
70
(n = 11)
250*
(n = 11)
70
(n = 11)
1230
50
(n = 12)
--
--
--
p<0.05; p<0.01; p < 0.001.
-- Gralewicz and Wiaderna (2001) only utilized a control group and one exposure group of 492 mg/m3
"Adapted from Saillenfait et al. (2005).
bNumbers of live fetuses not explicitly reported.
"Values are expressed as mean ± 1SD.
dl\lumber of dams with live litters.
eAdapted from Wiaderna et al. (2002).
Values expressed as response relative to control, percentage.
8Adapted from Gralewicz and Wiaderna (2001).
This document is a draft for review purposes only and does not constitute Agency policy.
2-30 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
10000 -i
I
c
.2 1000
4-1
to
4-1
c
0)
u
c
o
100
Altered Altered 4, Pain 4, Anxiety 4, ^
cognitive cognitive sensitivity and/or 1" maternal fetal
function... function... (a) motor activity... weight gain... weight...
2 Solid lines represent range of exposure concentrations, (a) Gralewicz and Wiaderna (2001); (b) Wiaderna et al.
3 (2002): (c) Saillenfait et al. (2005)
4 Figure 2-5. Exposure response array for endpoints resulting from
5 inhalation exposure to 1,3,5-TMB considered for the derivation of the
6 RfC
7 2.3.2. Methods of Analysis for 1,3,5-TMB
8 As discussed above in Section 2.3.1, endpoints observed in Saillenfait et al. [2005] that
9 demonstrated statistically significant (p < 0.05) pair-wise increases or decreases relative to control
10 for at least one exposure group were considered for the derivation of the RfC for 1,3,5-TMB; these
11 effects are listed in Table 2-10. Additionally, altered cognitive function, decreased pain sensitivity,
12 and decreased anxiety and/or increased motor function observed in Gralewicz et al. [2001] and
13 Wiaderna et al. [2002] were also considered as the basis for the derivation of the RfC for 1,3,5-TMB.
14 This assessment used the BMD approach, when possible, to estimate a POD for the derivation of an
15 RfC for 1,3,5-TMB (see Section C.I of Appendix C for detailed methodology]. The BMD approach
16 involves fitting a suite of mathematical models to the observed dose-response data using EPA's
17 BMDS (version 2.2], and then selecting the best fitting model. Each best-fit model estimates a BMD
18 and its associated BMDL (i.e., a 95% lower bound on the BMD] corresponding to a selected BMR.
19 For the decreased male and female fetal body weight endpoints identified from the
20 Saillenfait et al. [2005] study, a BMR of 5% relative deviation from the control mean was selected.
21 A 5% decrease in fetal body weight relative to control was determined to be a minimal, biologically
22 significant response. This determination is based on the fact that decreased body weight gain in
This document is a draft for review purposes only and does not constitute Agency policy.
2-31 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 fetuses and/or pups is considered indicative of altered growth, which has been identified by EPA as
2 one of the four major manifestations of developmental toxicity [U.S. EPA. 1991). In addition, a 10%
3 decrease in adult body weight in animals is generally recognized as a biologically significant
4 response associated with identifying a maximum tolerated dose, but since fetuses and/or pups are
5 generally recognized as a susceptible lifestage, and thus are assumed to be more greatly affected by
6 decreases in body weight than adult animals, a 5% decrease in fetal body weight is considered a
7 biologically significant response. Finally, in humans, reduced birth weight is associated with a
8 series of adverse effects including neonatal and postnatal mortality, coronary heart disease, arterial
9 hypertension, chronic renal insufficiency, and diabetes mellitus [Barker. 2007: Reyes and Manalich.
10 2005]. For these reasons, the selection of a BMR of 5% for decreased fetal body weight was
11 considered reasonable. Additionally, as recommended by EPA's Benchmark Dose Technical
12 Guidance Document [2000], a BMR equal to a change in the mean of 1 standard deviation of the
13 model estimated control mean was also selected for the BMD modeling of both fetal body weight
14 and maternal body weight gain to facilitate comparisons across assessments. The estimated BMDL
15 is then used as the candidate POD (Table 2-11].
16 The suitability of the above methods to determine a POD is dependent on the nature of the
17 toxicity database for a specific chemical. The data for neurotoxicity (i.e., altered cognitive function,
18 decreased pain sensitivity, and decreased anxiety and/or increased motor function] for 1,3,5-TMB
19 were not modeled. Gralewicz and Wiaderna (2001] only employed one concentration when
20 investigating the neurotoxic effects of 1,3,5-TMB following short-term inhalation exposures. For
21 altered cognitive function (as measured as decreased passive and active avoidance] reported in
22 Wiaderna et al. (2002], responses were observed to be equal in all exposure groups, and this lack of
23 a dose-response relationship precluded BMD modeling. In the Saillenfait et al. (2005] study,
24 although decreased fetal body weight in males and females was considered appropriate for BMD
25 modeling, BMDS was unable to adequately model the variance in response for this endpoint In
26 cases where BMD modeling is not feasible or modeling failed to appropriately describe the dose-
27 response characteristics, the NOAEL/LOAEL approach was used to identify a POD. Detailed
28 modeling results are provided in Section C.I of Appendix C.
29 Because an RfC is a toxicity value that assumes continuous human inhalation exposure over
30 a lifetime, data derived from inhalation studies in animals need to be adjusted to account for the
31 noncontinuous exposures used in these studies. In the Gralewicz and Wiaderna (2001] and
32 Wiaderna et al. (2002] studies, rats were exposed to 1,3,5-TMB for 6 hours/day, 5 days/week for 4
33 weeks. Because no PBPK model exists for 1,3,5-TMB, the duration-adjusted PODs for
34 neurobehavioral effects in rats were calculated as follows:
35
36 PODADJ (mg/m3) = POD (mg/m3) x hours exposed per day/24 hours x days exposed
37 per week/7 days
38
This document is a draft for review purposes only and does not constitute Agency policy.
2-32 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Therefore, for altered cognitive function from Gralewicz and Wiaderna [2001], the PODADj
2 would be calculated as follows:
3 PODADJ (mg/m3) = 492 mg/m3x 6 hours/24 hours x 5 days/7 days
4 PODADj (mg/m3) = 87.9 mg/m3
5 In the Saillenfait et al. [2005] study, rats were exposed to 1,3,5-TMB for 6 hours/day for 15
6 consecutive days (CDs 6-20]. Therefore, the duration-adjusted PODs for developmental/ maternal
7 effects were calculated as follows:
8 PODADJ (mg/m3) = POD (mg/m3) x hours exposed per day/24 hours
9 For example, for decreased fetal weight in males, the PODADj would be calculated as follows:
10 PODADj (mg/m3) = 2,974 mg/m3 x 6 hours/24 hours
11 PODADj (mg/m3) = 744 mg/m3
12 The calculated PODADj (mg/m3] values for all neurotoxicity and developmental endpoints
13 considered for RfC derivation are presented in Table 2-11.
14
This document is a draft for review purposes only and does not constitute Agency policy.
2-33 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 2-11. Summary of dose-response analysis and point of departure
estimation for endpoints resulting from short-term and gestational
inhalation exposures to 1,3,5-TMB
Reference
Endpoint
Species/
sex
POD
basis
Best-fit model;
BMR
Candidate
POD
(mg/m3)
POD^
(mg/mV
Neurological endpoints
Gralewicz
and
Wiaderna
(2001)
Wiaderna et
al. (2002)
Altered cognitive function
Decreased pain sensitivity
Decreased anxiety and/or
increased motor function
Altered cognitive function
Rat, male
Rat, male
Rat, male
Rat, male
LOAEL
LOAEL
LOAEL
LOAEL
n/ab
n/ab
n/ab
/ b
n/a
492
492
492
123
87.9
87.9
87.9
22.0
Developmental endpoints
Saillenfait et
al. (2005)
Decreased fetal body weight
Rat, male
Rat, female
NOAEL
NOAEL
n/ab
n/ab
2,974
2,974
744
744
Maternal endpoints
Saillenfait et
al. (2005)
Decreased maternal weight
body gain
Rat, female
BMDL
Power; 1 SD
1,302
326.0
4
5
6
7
8
9
10
11
12
13
14
15
aDuration adjusted PODADJ (mg/m3) = POD x (6 hours/24 hours) for developmental/maternal endpoints, or POD x (6
hours/24 hours) x (5 days/7 days) (U.S. EPA, 2002).
bNo model was able to fit data adequately, or data were not modeled.
2.3.3. Derivation of the Reference Concentration for 1,3,5-TMB
Because the selected endpoints for consideration as the critical effect (i.e., altered cognitive
function, decreased pain sensitivity, decreased anxiety and/or increased motor function, decreased
fetal body weight, and maternal body weight gain) are assumed to result primarily from systemic
distribution of 1,3,5-TMB, and no available PBPK model exists for 1,3,5-TMB, the human equivalent
concentration (HEC) for 1,3,5-TMB was calculated by the application of the appropriate dosimetric
adjustment factor (DAF) for systemically acting gases (i.e., Category 3 gases), in accordance with the
EPA's RfC Methodology (U.S. EPA. 1994b). DAFs are ratios of animal and human physiologic
parameters, and are dependent on the nature of the contaminant (i.e., particle or gas) and the target
site (i.e., respiratory tract or remote to the portal-of-entry [i.e., systemic]) (U.S. EPA, 1994b). For
gases with systemic effects, the DAF is expressed as the ratio between the animal and human
blood:air partition coefficients:
This document is a draft for review purposes only and does not constitute Agency policy.
2-34 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 DAF = (Hb/g)A/(Hb/g)H
2 DAF = 5 5.7/43
3 DAF = 1.3
4 where:
5 (Hb/g)A = the animal bloochair partition coefficient
6 (Hb/g)H = the human bloochair partition coefficient
7 In cases where the animal blood:air partition coefficient is higher than the human value
8 [Meulenberg and Vijverberg, 2000: Tarnberg and Tohanson, 1995], resulting in a DAF > 1, a default
9 value of 1 is substituted [U.S. EPA. 1994b]. For example, the HEC for altered CNS function (reported
10 in Wiaderna et al. [2002]] is calculated as follows:
11
12
13
14
PODHEc =
PODHEc =
(mg/m3) x DAF
(mg/m3) x 1.0
PODHEC = 22 mg/m3 x 1.0
PODHEC = 22 mg/m3
15 Table 2-12 presents the calculated HECs for the candidate critical effects, selected
16 uncertainty factors (UFs], and the resulting derivation of candidate RfCs from the two short-term
17 and one developmental toxicity studies [Saillenfaitetal., 2005: Wiaderna etal., 2002: Gralewicz and
18 Wiaderna. 2001].
19
This document is a draft for review purposes only and does not constitute Agency policy.
2-35 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table 2-12. PODADj values, human equivalent concentrations (HECs),
uncertainty factors, and candidate RfCs for 1,3,5-TMB
Reference
Endpoint
PODADJ
(mg/m3)
HEC
(mg/m3)3
Uncertainty factors (UF)
UFA
UFH
UFL
UFS
UFD
UFcoMPOSITE
Candidate
RfC
(mg/m3)b
Neurological Endpoints
Gralewicz
and
Wiaderna
(2001)
Wiaderna et
al. (2002)
Altered cognitive
function
Decreased pain
sensitivity
Decreased anxiety
and/or increased
motor function
Altered cognitive
function
87.9
87.9
87.9
22.0
87.9
87.9
87.9
22.0
3
3
3
3
10
10
10
10
10
10
10
10
10
10
10
10
3
3
3
3
10,000
10,000
10,000
10,000
n/ac
n/ac
n/ac
n/ac
Developmental Endpoints
Saillenfait et
al. (2005)
Decreased fetal
body weight, male
Decreased fetal
body weight,
female
744
744
744
744
3
3
10
10
1
1
1
1
3
3
100
100
7.44
7.44
Maternal Endpoints
Saillenfait et
al. (2005)
Decreased
maternal weight
body gain
326.0
326.0
3
10
1
10
3
1,000
3.26 x 10"1
Human equivalent concentration
b As calculated by application of uncertainty factors, not rounded to 1 significant digit.
c Endpoint excluded for further consideration due to a UFCOMPOSITE of 10,000. The 2002 report "A Review of the
Reference Dose and Reference Concentration Processes" (U.S. EPA, 2002) recommends a maximum composite UF
of 3000 for derivation of an RfC.
3 The magnitude of the composite uncertainty factors associated with the neurotoxicological
4 endpoints from Gralewicz and Wiaderna [2001] and Wiaderna et al. [2002] indicate that these
5 endpoints cannot support the derivation of an RfC for 1,3,5-TMB. The composite UF for 1,3,5-TMB
6 for the neurotoxicological endpoints from Gralewicz and Wiaderna [2001] and Wiaderna et al.
7 [2002] would be 10,000. In the report, A Review of the Reference Dose and Reference Concentration
8 Processes [U.S. EPA. 2002] the RfD/RfC Technical Panel concluded that, in cases where maximum
9 uncertainty exists in four or more areas of uncertainty, or when the composite uncertainty factor is
10 10,000 or more, it is unlikely that the database is sufficient to derive a reference value. Therefore,
11 consistent with the recommendations in this report [U.S. EPA, 2002], the available neurotoxicity
12 data following short-term inhalation exposure to 1,3,5-TMB were considered insufficient to support
13 reference value derivation and a candidate RfC for 1,3,5-TMB was not derived based on these data.
This document is a draft for review purposes only and does not constitute Agency policy.
2-36 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 Of the remaining effects considered for derivation of the RfC, decreased maternal weight
2 gain was identified as the most sensitive endpoint A PODHEc of 326.0 mg/m3 for decreased
3 maternal weight gain from Saillenfait et al. [2005] was used to derive a candidate chronic RfC for
4 1,3,5-TMB as shown in Table 2-11. Uncertainty factors, selected and applied in accordance with the
5 procedures described in based on EPA's A Review of the Reference Dose and Reference Concentration
6 Processes [U.S. EPA. 2002). address five areas of uncertainty resulting in a composite UF of 1,000.
7 This composite UF was applied to the selected POD to derive an RfC.
8 An interspecies uncertainty factor, UFA, of 3 (101/2 = 3.16, rounded to 3) was applied to
9 account for uncertainty in characterizing the toxicokinetic and toxicodynamic differences between
10 rats and humans following inhalation exposure to 1,3,5-TMB. In this assessment, the use of a DAF
11 to extrapolate external concentrations from rats to humans reduces toxicokinetic uncertainty in
12 extrapolating from the rat data, but does not account for the possibility that humans may be more
13 sensitive to 1,3,5-TMB than rats due to toxicodynamic differences. A default UFA of 3 was thus
14 applied to account for this remaining toxicodynamic uncertainty and any residual toxicokinetic
15 uncertainty.
16 An intraspecies uncertainty factor, UFH, of 10 was applied to account for potentially
17 susceptible individuals in the absence of data evaluating variability of response in the human
18 population following inhalation of 1,3,5-TMB. No information is currently available to predict
19 potential variability in human susceptibility, including variability in the expression of enzymes
20 involved in 1,3,5-TMB metabolism.
21 A LOAEL to NOAEL uncertainty factor, UFi, of 1 was applied because a NOAEL was used as
22 the POD.
23 A subchronic to chronic uncertainty factor, UFs, of 10 was applied to account for
24 extrapolation from a subchronic exposure duration study to derive a chronic RfC. The 10-fold
25 uncertainty factor is applied to the POD identified from the subchronic study on the assumption
26 that effects observed in a similar chronic study would be observed at lower concentrations for a
27 number of possible reasons, including potential cumulative damage occurring over the duration of
28 the chronic study or an increase in the magnitude or severity of effect with increasing duration of
29 exposure.
30 A database uncertainty factor, UFo, of 3 (101/2 = 3.16, rounded to 3) was applied to account
31 for database deficiencies. Strengths of the database include two well-designed short-term studies
32 that observed exposure-response effects in the central nervous system of Wistar rats exposed to
33 1,3,5-TMB. An additional strength of the database is the well-designed developmental toxicity
34 study that investigated standard measures of maternal and fetal toxicity in a different strain of rat
35 (Sprague-Dawley). However, the lack of a multi-generational reproductive/developmental toxicity
36 study investigating effects due to 1,3,5-TMB exposure is a weakness of the database. Although a
37 multi-generation reproductive/developmental toxicity study does not exist for 1,3,5-TMB, there is a
38 multi-generation reproductive/developmental toxicity study for high flash naphtha, of which 1,3,5-
This document is a draft for review purposes only and does not constitute Agency policy.
2-37 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 1MB is a constituent. This study demonstrates effects on postnatal growth at lower exposures in
2 the F3 generation (2,460 mg/m3) compared to the F2 or Fi generation (7,380 mg/m3) (McKee etal..
3 1990]. but did not observe a consistent effect on reproductive parameters. This raises some
4 concern that addition of multi-generation reproductive/developmental toxicity study for 1,3,5-TMB
5 might result in the identification of a lower POD.
6 EPA's Review of the Reference Dose and Reference Concentration Processes (U.S. EPA. 2002]
7 also recommends that the database uncertainty factor take into consideration whether there is
8 concern from the available toxicology database that the developing organism may be particular
9 susceptible to effects in specific organ systems. TMBs (unspecified isomer] are able to cross the
10 placenta (Cooper etal.. 2001: Dowty etal.. 1976]: therefore, as neurotoxicity is observed in adult
11 animals, there is concern that exposure to 1,3,5-TMB may result in neurotoxicity in the developing
12 organism. EPA's Guidelines for Neurotoxicity Risk Assessment (U.S. EPA, 1998] identifies specific
13 effects observed in adult animals (e.g., cognitive and motor function] that can also affect the
14 developing organism exposed in utero. The Neurotoxicity Guidelines (U.S. EPA. 1998] also indicate
15 that neurotoxicants may have greater access to the nervous system in developing organisms due to
16 an incomplete blood-brain barrier and immature metabolic detoxifying pathways. Therefore, there
17 is some concern that the lack of a developmental neurotoxicity study is a deficiency in the database
18 and that the inclusion of such a study would potentially result in a lower POD than the POD for
19 maternal effects identified from the available 1,3,5-TMB toxicity database. In summary, a 3-fold
20 database UF was applied to account for the lack of both a multi-gene ration
21 reproductive/developmental toxicity study and a developmental neurotoxicity study in the
22 available database for 1,3,5-TMB.
23 Application of this 1,000-fold composite UF yields the calculation of the chronic RfC for
24 1,3,5-TMB as follows:
25 RfC = PODHEc * UF = 326 mg/m3 4- 1,000 = 0.326 mg/m3 = 3 x l(H mg/m3 (rounded to one
26 significant digit)
27
28 While Saillenfait et al. (2005] is a well-conducted developmental toxicity study that utilizes
29 appropriate study design, group sizes, and statistical analysis, and evaluates a wide range of fetal
30 and maternal endpoints resulting from 1,3,5-TMB inhalation exposure, a number of other factors
31 lessens its suitability for use in deriving an RfC for 1,3,5-TMB. First, although maternal and fetal
32 toxicities were observed in this study, it is important to note that the candidate RfC for 1,3,5-TMB
33 derived based on the critical effect of decreased maternal body weight gain (corrected for gravid
34 uterine weight] is 15-fold higher than the RfC derived for 1,2,4-TMB, which is based on altered CNS
35 function measured as decreased pain sensitivity. As discussed in Section 1.1.6, the available
36 toxicological database for 1,2,4-TMB and 1,3,5-TMB, across all exposure durations, indicates there
This document is a draft for review purposes only and does not constitute Agency policy.
2-38 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 are important similarities in the two isomers' toxicity that are supportive of an RfC for 1,3,5-TMB
2 that is not substantially different than the RfC derived for 1,2,4-TMB.
3 In acute studies investigating the respiratory irritative effects of the two isomers, the RDso
4 of 1,2,4-TMB and 1,3,5-TMB were observed to be very similar, 2,844 and 2,553 mg/m3,
5 respectively [Korsaketal.. 1997]. This similarity in toxicity for respiratory effects was also
6 observed for neurotoxicity: the ECso for decreased coordination, balance, and neuromuscular
7 function (i.e., performance on the rotarod) was 4,694 mg/m3 for 1,2,4-TMB and 4,738 mg/m3 for
8 1,3,5-TMB. The ECso for decreased pain sensitivity (i.e., latency to paw-lick measured on the hot
9 plate apparatus) was also similar for both isomers: 5,683 mg/m3 for 1,2,4-TMB and 5,963 mg/m3
10 for 1,3,5-TMB (Korsak and Rydzynski, 1996]. Other neurotoxic endpoints similarly affected by
11 either isomer (albeit from oral exposures or i.p. injections) included increased electrocortical
12 arousal and altered EEC function (Tomas etal., 1999a: Tomas etal., 1999c). However, the doses
13 eliciting these effects were LOAELs, and therefore it is unclear whether this represents true
14 similarity in toxic potency or whether testing at lower doses would reveal differences between the
15 two isomers. Additionally, the magnitude of effect differed slightly between the isomers, with
16 1,2,4-TMB inducing greater changes in brain EEGs and 1,3,5-TMB inducing greater changes in
17 electrocortical arousal.
18 In short-term neurotoxicity studies, a similar pattern of effects (inability to learn passive or
19 active avoidance, decreased pain sensitivity, increased spontaneous motor activity) indicating
20 altered neurobehavioral function was observed in rats exposed to either isomer (Wiadernaetal..
21 2002: Gralewicz and Wiaderna. 2001: Gralewicz etal.. 1997a). In these studies, 1,3,5-TMB was
22 shown to be more toxic that 1,2,4-TMB, with neurobehavioral effects occurring at lower exposures
23 (123 vs. 492 mg/m3) in animals exposed to 1,3,5-TMB versus those exposed to 1,2,4-TMB. Also,
24 manifestations of neurotoxicity occurred at earlier time points (3 vs. 7 days) in rats exposed to
25 1,3,5-TMB compared to those exposed to 1,2,4-TMB.
26 Finally, the observed developmental effects observed in Saillenfait et al. (2005) were shown
27 to be similar between the two isomers. Exposure to 1,2,4-TMB and 1,3,5-TMB significantly
28 decreased male fetal body weights to a similar degree (5% and 7%, respectively) at 2,952 mg/m3.
29 1,2,4-TMB and 1,3,5-TMB also decreased female body weights to a similar degree (5% and 6%,
30 respectively) at the same concentration. This body weight decrease was significant in females
31 exposed to 1,2,4-TMB, but was not significant in females exposed to 1,3,5-TMB. 1,3,5-TMB was
32 observed to be more toxic with regard to maternal toxicity, inducing a 75% reduction in maternal
33 weight gain at 2,952 mg/m3 compared to a 50% reduction in females exposed to the same
34 concentration of 1,2,4-TMB.
35 The two isomers are similar to one another in their chemical and toxicokinetic properties,
36 although important differences also exist Both isomers have very similar Log Kow values, and
37 blood:air partition coefficients reported for humans and rats in the literature are similar between
38 the two isomers: 43.0 for 1,2,4-TMB and 59.1 for 1,3,5-TMB. This gives an indication that the two
This document is a draft for review purposes only and does not constitute Agency policy.
2-39 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 isomers would partition into the blood in a similar fashion. Supporting this is the observation that
2 1,2,4-TMB and 1,3,5-TMB absorb equally into the bloodstream of exposed humans (6.5 and 6.2 |iM,
3 respectively) (Jarnbergetal., 1996]. Also, the net respiratory uptake of 1,2,4-TMB and 1,3,5-TMB
4 was similar among humans, and the respiratory uptake for 1,2,4-TMB was similar across humans
5 and rats (Jarnberg et al.. 1996: Dahl et al.. 1988]. Distribution of the two isomers throughout the
6 body is qualitatively similar, although it appears that liver and kidney concentrations for 1,2,4-TMB
7 are greater than those for 1,3,5-TMB after both acute and short-term exposures [Swiercz etal..
8 2006: Swiercz etal.. 2003: Swiercz etal.. 2002].
9 Although 1,2,4-TMB was observed to distribute to the brain [Swiercz etal.. 2003: Eide and
10 Zahlsen. 1996]. distribution of 1,3,5-TMB to the brain was not experimentally measured in any
11 study. However, the predicted brain:air partition coefficient was similar between 1,2,4-TMB and
12 1,3,5-TMB for both humans (206 vs. 199] and rats (552 vs. 535] (Meulenberg and Viiverberg.
13 20001 This strongly suggests that 1,2,4-TMB and 1,3,5-TMB can be expected to distribute similarly
14 to the brain in both humans and rats. Both isomers were observed to primarily metabolize to
15 benzoic and hippuric acids in humans and rats (Tarnbergetal., 1996: Huo etal., 1989: Mikulski and
16 Wiglusz. 1975]. although the amount of inhaled TMB recovered as hippuric acid metabolites
17 following exposure to 1,2,4-TMB or 1,3,5-TMB was somewhat dissimilar in humans (22% vs. 3%,
18 respectively] and rats (24-38% vs. 59%, respectively] (Tarnbergetal., 1996: Mikulski and Wiglusz,
19 1975].
20 Other terminal metabolites included mercapturic acids (~14-19% total dose], phenols
21 (~12% total dose], and glucuronides and sulphuric acid conjugates (4-9% total dose] for
22 1,2,4-TMB, and phenols (~4-8% total dose] and glucuronides and sulphuric acid conjugates (~5-
23 9% total dose] for 1,3,5-TMB (Tsuiimoto etal.. 2005: Tsuiimoto etal.. 2000: Huo etal.. 1989:
24 Wiglusz. 1979: Mikulski and Wiglusz, 1975]. In humans, the half-lives of elimination from blood
25 were observed to be greater for 1,3,5-TMB (1.7 minutes, 29 minutes, 4.9 hours, and 120 hours]
26 than for 1,2,4-TMB (1.3 minutes, 21 minutes, 3.6 hours, and 87 hours] (Tarnbergetal.. 1997a:
27 Tarnbergetal.. 1997b: Tarnbergetal.. 1996]. although this difference may be due to small sample
28 sizes and difficulties in measuring slow elimination phases rather than a true difference in half-
29 lives. At low concentrations in rats, half-lives in elimination from the blood were somewhat similar
30 for 1,2,4-TMB and 1,3,5-TMB (3.6 vs. 2.9 hours], but this difference became much greater with
31 increasing doses (17.3 hours for 1,2,4-TMB and 4 hours for 1,3,5-TMB following exposure to 1,230
32 mg/m3 for 6 hours] (Swiercz etal., 2003: Swiercz etal., 2002].
33 Given the above information regarding the observed toxicity following 1,2,4-TMB and
34 1,3,5-TMB exposures across acute, short-term, and developmental studies, the use of 1,3,5-TMB-
35 specific data for derivation of an RfC was not considered to be scientifically supported. Derivation
36 of an RfC for 1,3,5-TMB based on decreased maternal weight gain, using the only adequate toxicity
37 data available (i.e., Saillenfait et al. (2005]] would result in an RfC 15-fold higher than the RfC
38 derived for 1,2,4-TMB based on altered CNS function (i.e., decreased pain sensitivity]. The available
This document is a draft for review purposes only and does not constitute Agency policy.
2-40 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 toxicity data indicates that 1,2,4-TMB and 1,3,5-TMB are similar in acute respiratory and
2 neurological toxicity and developmental toxicity, but that 1,3,5-TMB appears to be more potent in
3 eliciting neurotoxicity and maternal toxicity following short-term exposures. 1,3,5-TMB is
4 observed to elicit neurotoxic effects in rats in acute and short-term studies, and therefore the
5 selected critical effect for 1,2,4-TMB, altered CNS function (i.e., decreased pain sensitivity), is
6 relevant to observed 1,3,5-TMB-induced toxicity. Similarities in blood:air partition coefficients,
7 respiratory uptake, and absorption into the bloodstream between the two isomers support the
8 conclusion that internal blood dose metrics for 1,3,5-TMB would be similar to those calculated for
9 1,2,4-TMB using the available PBPK model.
10 Thus, the chronic RfC of 2 x 10 2 mg/m3 derived for 1,2,4-TMB was adopted as the RfC
11 for 1,3,5-TMB based on the conclusion that the two isomers were sufficiently similar based
12 on chemical properties, toxicokinetics, and toxicity.
13 2.3.4. Uncertainties in the Derivation of the Reference Concentration for 1,3,5-TMB
14 Uncertainties exist in adopting the RfC derived for 1,2,4-TMB based on altered CNS function
15 (i.e., decreased pain sensitivity) as the RfC for 1,3,5-TMB. While the available database for
16 1,3,5-TMB was considered sufficient to derive an RfC, if the most sensitive endpoint from the only
17 adequate study in the 1,3,5-TMB database (i.e., decreased maternal weight gain; Saillenfait et al.
18 (2005)) was used for the RfC derivation, an RfC 15-fold higher would be derived for 1,3,5-TMB vs.
19 that derived for 1,2,4-TMB (3 x 1Q-1 vs. 2 x 1Q-2 mg/m3, respectively). Although uncertainty exists
20 in adopting the 1,2,4-TMB RfC for 1,3,5-TMB RfC, both isomers share multiple commonalities and
21 similarities regarding their chemical, toxicokinetic, and toxicological properties that support the
22 adoption of the value of one isomer for the other. The majority of uncertainty regarding 1,3,5-
23 TMB's database involves the lack of a chronic, subchronic, or multi-generational reproductive study
24 for this isomer. Given the similarities in toxicity from the developmental toxicity study, and
25 neurotoxicity and respiratory toxicity observed in the available acute and short-term studies, there
26 is strong evidence that the two isomer's toxicity resulting from subchronic exposure can be
27 expected to be similar. Moreover, 1,3,5-TMB may actually be expected to be slightly more toxic
28 than 1,2,4-TMB following subchronic exposures given the observation of earlier onset of effects
29 following 1,3,5-TMB exposures in short-term studies. Therefore, while uncertainty exists in the
30 derivation of 1,3,5-TMB's RfC, the available information regarding sufficient chemical, toxicokinetic,
31 and toxicological similarity between the two isomers supports adopting the RfC for 1,2,4-TMB as
32 the RfC for 1,3,5-TMB.
33 2.3.5. Confidence Statement for 1,3,5-TMB
34 The chronic RfC for 1,2,4-TMB was adopted as the RfC for 1,3,5-TMB; thus, confidence in the
35 study from which the critical effect was identified, Korsak and Rydzynski (1996). is medium
36 (see above). The database for 1,3,5-TMB includes acute, short-term, and developmental toxicity
This document is a draft for review purposes only and does not constitute Agency policy.
2-41 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 studies in rats and mice. However, confidence in the database is low to medium because it lacks
2 chronic, subchronic, multi-generation reproductive/developmental toxicity, and developmental
3 neurotoxicity studies and most of the studies supporting the critical effect come from the same
4 research institute. Reflecting the confidence in the study and the database and the uncertainty
5 surrounding the adoption of the RfC derived for 1,2,4-TMB as the RfC for 1,3,5-TMB, the overall
6 confidence in the RfC for 1,3,5-TMB is low.
7 2.4. Oral Reference Dose for Effects Other Than Cancer for 1,2,4-TMB
8 The RfD (expressed in units of mg/kg-day) is defined as an estimate (with uncertainty
9 spanning perhaps an order of magnitude) of a daily exposure to the human population (including
10 sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a
11 lifetime. It can be derived from a NOAEL, a LOAEL, or a 95% lower bound on the benchmark dose
12 (BMDL), with uncertainty factors (UFs) generally applied to reflect limitations of the data used.
13 2.4.1. Identification of Candidate Principal Studies and Critical Effects for 1,2,4-TMB
14 No chronic or subchronic studies were identified for 1,2,4-TMB that utilized the oral route
15 of exposure. Therefore, the available oral database for 1,2,4-TMB is minimal as defined by EPA
16 guidance (i.e., there is no human data available nor any adequate oral animal data) (U.S. EPA. 2002).
17 and thus this database is inadequate for the derivation of an RfD.
18 2.4.2. Methods of Analysis for 1,2,4-TMB
19 Even though the available oral database for 1,2,4-TMB is inadequate to derive an RfD, a
20 route-to-route extrapolation from inhalation to oral for the purposes of deriving an RfD is possible
21 using the existing inhalation data and the available 1,2,4-TMB PBPK model (Hissinketal.. 2007).
22 The Hissink model was chosen as an appropriate model because it was the only published 1,2,4-
23 TMB model that included parameterization for both rats and humans, the model code was available,
24 and the model adequately predicted experimental data in the dose range of interest. Using route-
25 to-route extrapolation via application of PBPK models is supported by EPA guidance (U.S. EPA.
26 2002.1994b) given enough data and the ability to interpret that data with regard to differential
27 metabolism and toxicity between different routes of exposure. The available database for
28 1,2,4-TMB supports the use of route-to-route extrapolation; sufficient evidence exists that
29 demonstrates similar qualitative profiles of metabolism (i.e., observation of dimethylbenzoic and
30 hippuric acid metabolites) and patterns of parent compound distribution across exposure routes
31 (Section B.2, Appendix B). Further, no evidence exists that would suggest toxicity profiles would
32 differ to a substantial degree between oral and inhalation exposures.
33 Therefore, assuming oral exposure would result in the same systemic effect as inhalation
34 exposure (i.e., altered CNS function, measured as decreased pain sensitivity (Korsak and Rydzynski.
35 1996)), an oral exposure component was added to the Hissink et al. (2007) PBPK model by EPA
This document is a draft for review purposes only and does not constitute Agency policy.
2-42 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 (Section B.3.3.5, Appendix B), assuming continuous oral ingestion and 100% absorption of the
2 ingested 1,2,4-TMB by constant infusion of the oral dose into the liver. This is a common
3 assumption when information about the oral absorption of the compound is unknown. The
4 contribution of the first-pass metabolism in the liver for oral dosing was evaluated by simulating
5 steady-state venous blood levels (at the end of 50 days continuous exposure) for a standard human
6 at rest (70 kg) for a range of concentrations and doses; at low daily doses (0.1-10 mg/kg-day),
7 equivalent inhalation concentrations result in steady state blood concentrations 4-fold higher than
8 those resulting from oral doses, indicating the presence of first-pass metabolism following oral
9 exposure (see Figure B-17, Appendix B). This difference became insignificant for daily doses
10 exceeding 50 mg/kg-day.
11 The human PBPK model inhalation dose metric (weekly average blood concentration,
12 mg/L) for the PODADj (0.086 mg/L) for decreased pain sensitivity was used as the target for the oral
13 dose metric. The human PBPK model was run to determine what oral exposure would yield an
14 equivalent weekly average blood concentration, and then the resulting value of 6.3 mg/kg-day was
15 used as the human equivalent dose POD (PODHED) for the RfD derivation.
16 2.4.3. Derivation of the Reference Dose for 1,2,4-TMB
17 A PODHED of 6.3 mg/kg-day was derived for the oral database using route-to-route
18 extrapolation based on the neurotoxic effects (i.e., decreased pain sensitivity) observed by Korsak
19 and Rydzyhski (1996) following inhalation exposure to 1,2,4-TMB. Thus, the same uncertainty
20 factors applied to derive the RfC (see Section 2.1.3) were also applied to derive the RfD. The
21 uncertainty factors, selected and applied in accordance with the procedures described in EPA's A
22 Review of the Reference Dose and Reference Concentration Processes (U.S. EPA, 2002) (Section 4.4.5
23 of the report), address five areas of uncertainty resulting in a composite UF of 1,000.
24 Application of this 1,000-fold composite UF yields the calculation of the chronic RfD for
25 1,2,4-TMB as follows:
26
27 RfD = PODHED * UF = 6.3 mg/kg-day 4- 1,000 = 0.006 mg/kg-day = 6 x 10 3 mg/kg-day
28 (rounded to one significant digit)
29 2.4.4. Uncertainties in the Derivation of the Reference Dose for 1,2,4-TMB
30 As the oral RfD for 1,2,4-TMB was based on a route-to-route extrapolation in order to
31 determine the oral dose that would result in the same effect (i.e., decreased pain sensitivity) as
32 inhalation exposure in Korsak and Rydzyhski (1996). the uncertainties regarding this derivation
33 are the same as those for the RfC for 1,2,4-TMB (see Section 2.1.4), with the exception of the
34 uncertainty surrounding the route-to-route extrapolation. The model used to perform this route-
35 to-route extrapolation is a well-characterized model considered appropriate for the purposes of
36 this assessment. One source of uncertainty regarding the route-to-route extrapolation is the
This document is a draft for review purposes only and does not constitute Agency policy.
2-43 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 assumption of 100% bioavailability, that is, 100% of the ingested 1,2,4-TMB would be absorbed and
2 pass through the liver. If not all of the compound is bioavailable, a lower blood concentration
3 would be expected compared to the current estimate, and thus, a higher RfD would be calculated.
4 2.4.5. Confidence Statement for 1,2,4-TMB
5 A PBPK model was utilized to perform a route-to-route extrapolation to determine a POD
6 for the derivation of the RfD from the Korsak and Rydzyhski [1996] inhalation study and
7 corresponding critical effect The confidence in the study from which the critical effect was
8 identified, Korsak and Rydzynski (1996). is medium (see above). The database for 1,2,4-TMB
9 includes acute, short-term, subchronic, and developmental toxicity studies in rats and mice.
10 However, confidence in the database for 1,2,4-TMB is low to medium because it lacks chronic,
11 multi-generation reproductive/developmental and developmental neurotoxicity studies, and the
12 studies supporting the critical effect predominantly come from the same research institute.
13 Reflecting the confidence in the study and the database and the uncertainty surrounding the
14 application of the available PBPK model for the purposes of a route-to-route extrapolation, the
15 overall confidence in the RfD for 1,2,4-TMB is low.
16 2.5. Oral Reference Dose for Effects Other Than Cancer for 1,2,3-TMB
17 2.5.1. Identification of Candidate Principal Studies and Critical Effects for 1,2,3-TMB
18 No chronic or subchronic studies were identified for 1,2,3-TMB that utilized the oral route
19 of exposure. Therefore, the available oral database for 1,2,3-TMB is minimal as defined by EPA
20 guidance (i.e., there is no human data available nor any adequate oral animal data) (U.S. EPA. 2002).
21 and thus this database is inadequate for the derivation of an RfD.
22 2.5.2. Methods of Analysis and Derivation of the Reference Dose for 1,2,3-TMB
23 The available oral database is inadequate to derive an RfD for 1,2,3-TMB. No chronic,
24 subchronic, or short-term oral exposure studies were found in the literature. However, as
25 discussed in Sections 1.1.7, there are sufficient similarities between isomers regarding observed
26 toxicological effects that support adopting the RfD for 1,2,4-TMB as the RfD for 1,2,3-TMB.
27 Specifically, the qualitative pattern of neurotoxic effects following short-term and subchronic
28 inhalation exposures is similar between TMB isomers. Particularly important to this determination
29 is that, although 1,2,3-TMB is observed to decrease pain sensitivity at lower concentrations than
30 1,2,4-TMB (LOAEL values of 123 vs. 492 mg/m3, respectively), the magnitude of decreased pain
31 sensitivity is similar for 1,2,4-TMB and 1,2,3-TMB, especially at the low- and mid-concentrations.
32 This similarity of effect in the low-dose region of the dose-response curve is exhibited by equal RfC
33 values derived from isomer-specific data: 2 x 10~2 mg/m3. Although a PBPK model exists for 1,2,4-
34 TMB that allows for route-to-route extrapolation from inhalation to oral exposure, no such model
This document is a draft for review purposes only and does not constitute Agency policy.
2-44 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 exists for 1,2,3-TMB. However, similarities in blood:air and tissue:air partition coefficients and
2 degree of absorption into the bloodstream between 1,2,4-TMB and 1,2,3-TMB support the
3 conclusion that internal blood dose metrics for 1,2,3-TMB would be similar to those calculated for
4 1,2,4-TMB using that isomer's available PBPK model. Also, the qualitative metabolic profiles for the
5 two isomers are similar, with dimethylbenzyl hippuric acids being the major terminal metabolite
6 for both isomers, such that first-pass metabolism through the liver is not expected to differ greatly
7 between 1,2,4-TMB and 1,2,3-TMB. Therefore, given the similarities in chemical properties,
8 toxicokinetics, and toxicity, the RfD derived for 1,2,4-TMB, 6 x 10-3 mg/kg-day was adopted
9 as the RfD for 1,2,3-TMB.
10 2.5.3. Uncertainties in the Derivation of the Reference Dose for 1,2,3-TMB
11 The uncertainties regarding adopting the RfD for 1,2,4-TMB as the RfD for 1,2,3-TMB
12 encompass previous areas of uncertainty involved in the derivation of the RfC for 1,2,3-TMB and
13 the RfD for 1,2,4-TMB (see Sections 2.1.4 and 2.2.4). Additionally, there is uncertainty in this
14 adoption regarding the assumptions made about the similarity in toxicokinetics and toxicity
15 between the two isomers. However, as discussed above in Sections 1.1.7 and in Appendix B
16 (Section B.2,), there is strong evidence that both isomers share multiple commonalities and
17 similarities regarding their toxicokinetic and toxicological properties that support adopting one
18 isomer's value for the other.
19 2.5.4. Confidence Statement for 1,2,3-TMB
20 The chronic RfD for 1,2,4-TMB was adopted as the RfD for 1,2,3-TMB; thus, confidence in
21 the study from which the critical effect was identified, Korsak and Rydzynski (1996). is
22 medium (see above). The database for 1,2,3-TMB includes acute, short-term, and subchronic
23 studies in rats and mice. However, confidence in the database is low to medium because it lacks
24 chronic, multi-generation reproductive/developmental, developmental toxicity, or developmental
25 neurotoxicity studies, and the studies supporting the critical effect predominantly come from the
26 same research institute. Reflecting the confidence in the study and the database and the
27 uncertainty surrounding the adoption of the RfD derived for 1,2,4-TMB as the RfD for 1,2,3-TMB,
28 the overall confidence in the RfD for 1,2,3-TMB is low.
29 2.6. Oral Reference Dose for Effects Other Than Cancer for 1,3,5-TMB
30 2.6.1. Identification of Candidate Principal Studies and Critical Effects for 1,3,5-TMB
31 No chronic or subchronic studies were identified for 1,3,5-TMB that utilized the oral route
32 of exposure. Therefore, the available oral database for 1,3,5-TMB is minimal as defined by EPA
33 guidance (i.e., there is no human data available nor any adequate oral animal data) (U.S. EPA. 2002).
34 and thus this database is inadequate for the derivation of an RfD.
This document is a draft for review purposes only and does not constitute Agency policy.
2-45 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
1 2.6.2. Methods of Analysis and Derivation of the Reference Dose for 1,3,5-TMB
2 The available oral database is inadequate to derive an RfD for 1,3,5-TMB. No chronic,
3 subchronic, or short-term oral exposure studies were found in the literature. However, as outlined
4 in the RfC Derivation for 1,3,5-TMB, the chemical, toxicokinetic, and toxicological similarities
5 between 1,3,5-TMB and 1,2,4-TMB support adopting the RfC for 1,2,4-TMB as the RfC for
6 1,3,5-TMB. These considerations also apply to the oral reference value, thus the RfD for 1,2,4-TMB
7 was adopted for 1,3,5-TMB. 1,3,5-TMB is observed to elicit neurotoxic effects in rats in acute and
8 short-term studies, and therefore the selected critical effect for 1,2,4-TMB, altered CNS function, is
9 relevant to observed 1,3,5-TMB-induced toxicity. Similarities in blood:air and tissue:air partition
10 coefficients and absorption into the bloodstream between the two isomers support the conclusion
11 that internal blood dose metrics for 1,3,5-TMB would be similar to those calculated for 1,2,4-TMB
12 using the available PBPK model. Also, the qualitative metabolic profiles for the two isomers are
13 similar, with dimethylbenzyl hippuric acids being the major terminal metabolite for both isomers,
14 so that first-pass metabolism through the liver is not expected to differ greatly between 1,2,4-TMB
15 and 1,3,5-TMB. Therefore, given the similarities in chemical properties, toxicokinetics, and
16 toxicity, the RfD derived for 1,2,4-TMB of 6 x 10 3 mg/kg-day was adopted as the RfD for
17 1,3,5-TMB.
18 2.6.3. Uncertainties in the Derivation of the Reference Dose for 1,3,5-TMB
19 The uncertainties regarding adopting the RfD for 1,2,4-TMB as the RfD for 1,3,5-TMB
20 encompass previous areas of uncertainty involved in the derivation of the RfC for 1,3,5-TMB and
21 the RfD for 1,2,4-TMB (see Sections 2.3.4 and 2.4.4). There is uncertainty regarding this adoption.
22 However, as discussed above in Section 2.3.3, both isomers share multiple commonalities and
23 similarities regarding their chemical, toxicokinetic, and toxicological properties that support
24 adopting one isomer's value for the other. Additionally, as the RfD derivation for 1,2,4-TMB was
25 based on a route-to-route extrapolation, the uncertainties in that toxicity value's derivation (see
26 Section 2.4.3) apply to the derivation of the RfD for 1,3,5-TMB.
27 2.6.4. Confidence Statement for 1,3,5-TMB
28 The chronic RfD for 1,2,4-TMB was adopted as the RfD for 1,3,5-TMB; thus confidence in
29 the study from which the critical effect was identified, Korsak and Rydzynski (1996). is
30 medium (see above). The database for 1,3,5-TMB includes acute, short-term, and developmental
31 toxicity studies in rats and mice. However, confidence in the database is low to medium because
32 it lacks chronic, multi-generation reproductive/developmental and developmental neurotoxicity
33 studies, and the studies supporting the critical effect predominantly come from the same research
34 institute. Reflecting the confidence in the study and the database and the uncertainty surrounding
35 the adoption of the RfD derived for 1,2,4-TMB as the RfD for 1,3,5-TMB, the overall confidence in
36 the RfD for 1,3,5-TMB is low.
This document is a draft for review purposes only and does not constitute Agency policy.
2-46 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
i 2.7. Cancer Assessment for 1,2,3-TMB, 1,2,4-TMB, and 1,3,5-TMB
2 Under the U.S. EPA Guidelines for Carcinogen Risk Assessment [U.S. EPA. 20051. the database
3 for 1,2,4-TMB, 1,2,3-TMB, and 1,3,5-TMB provides "inadequate information to assess
4 carcinogenic potential". This characterization is based on the limited and equivocal genotoxicity
5 findings, and the lack of data indicating carcinogenicity in experimental animal species via any
6 route of exposure. Information available on which to base a quantitative cancer assessment is
7 lacking, and thus, no cancer risk estimates for either oral or inhalation exposure are derived.
This document is a draft for review purposes only and does not constitute Agency policy.
2-47 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
REFERENCES
ACGIH. (American Conference of Governmental Industrial Hygienists). (2002). Trimethyl benzene isomers. In
Documentation of the threshold limit values and biological exposure indices (7 ed.). Cincinnati, OH.
Andersson. K: Fuxe. K: Toftgard. R: Nilsen. OG: Eneroth. P: Gustafsson. IA. (1980). Toluene-induced activation
of certain hypothalamic and median eminence catecholamine nerve terminal systems of the male rat
and its effects on anterior pituitary hormone secretion. Toxicol Lett 5: 393-398.
http://dx.doi.org/10.1016/0378-4274(80)90021-l.
Andersson. K: Fuxe. K: Nilsen. OG: Toftgard. R: Eneroth. P: Gustafsson. IA. (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 para-xylene, and ethylbenzene. Toxicol Appl
Pharmacol 60: 535-548.
Andersson, K: Nilsen, OG: Toftgard, R: Eneroth, P: Gustafsson, IA: Battistini, N: Agnati, LF. (1983). Increased
amine turnover in several hypothalamic noradrenaline nerve exposure to various concentrations of
toluene. NeuroToxicology 4: 43-55.
Aylward. LL: Becker. RA: Kirman. CR: Hays. SM. (2011). Assessment of margin of exposure based on
biomarkers in blood: an exploratory analysis. Regul Toxicol Pharmacol 61: 44-52.
http://dx.doi.0rg/10.1016/j.yrtph.2011.06.001.
Baettig. K: Grandjean. E: Rossi. L: Rickenbacher. I. (1958). Toxicologische untersuchungen uber
trimethylbenzol. Archiv fuer Gewerbepathologie und Gewerbehygiene 16: 555-566.
Balster. RL. (1998). Neural basis of inhalant abuse. Drug Alcohol Depend 51: 207-214.
Barker. DIP. (2007). The origins of the developmental origins theory. Journal of Internal Medicine 261: 412-
417. http://dx.doi.0rg/10.llll/j.1365-2796.2007.01809.x.
Battig. K: Grandjean. E: Turrian. V. (1956). [Health damage after continuous exposure to trimethyl benzene in
a painting workshop]. Soz Praventivmed 1: 389-403. http://dx.doi.org/10.1007/BF02031676.
Billionnet. C: Gay. E: Kirchner. S: Leynaert. B: Annesi-Maesano. I. (2011). Quantitative assessments of indoor
air pollution and respiratory health in a population-based sample of French dwellings. Environ Res
111: 425-434. http://dx.doi.Org/10.1016/j.envres.2011.02.008.
Bogo. V: Hill. TA: Young. RW. (1981). Comparison of accelerod and rotarod sensitivity in detecting ethanol-
and acrylamide-induced performance decrement in rats: review of experimental considerations of
rotating rod systems. Neurotoxicology 2: 765-787.
Borriston Labs. (Borriston Laboratories). (1983). Four-week oral nephrotoxicity screening study in male
F344 rats. (1706). Temple Hills, MD.
http://www.ntis.gov/search/product.aspx?ABBR=OTS00004600.
Bowen. SE: Batis. 1C: Paez-Martinez. N: Cruz. SL. (2006). The last decade of solvent research in animal models
of abuse: Mechanistic and behavioral studies. Neurotoxicol Teratol 28: 636-647.
http://dx.doi.0rg/10.1016/j.ntt.2006.09.005.
Braes. PU: Gregory. P: lackson. DM. (1984). Passive avoidance in rats: Disruption by dopamine applied to the
nucleus accumbens. Psychopharmacology 83: 70-75.
Brooks. SP: Dunnett. SB. (2009). Tests to assess motor phenotype in mice: a user's guide. Nat Rev Neurosci
10: 519-529. http://dx.doi.org/10.1038/nrn2652.
This document is a draft for review purposes only and does not constitute Agency policy.
R-1 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Chen. R: Dick. F: Seaton. A. (1999). Health effects of solvent exposure among dockyard painters: Mortality and
neuropsychological symptoms. Occup Environ Med 56: 383-387.
http://dx.doi.0rg/10.1136/oem.56.6.383.
Chevron Chemical Company. (1985). One generation reproduction study of FED 5450 in rats with cover
letter. (OTS0206739). Washington, DC: U.S. EPA.
Choi. DW: Moon. KW: Byeon. SH: Lee. El: Sul. DG: Lee. IH: Oh. EH: Kim. YH. (2009). Indoor volatile organic
compounds in atopy patients' houses in South Korea. Indoor Built Environ 18: 144-154.
http://dx.doi.org/10.1177/1420326X08101945.
Chou. CC: Riviere. IE: Monteiro-Riviere. NA. (2003). The cytotoxicity of jet fuel aromatic hydrocarbons and
dose-related interleukin-8 release from human epidermal keratinocytes. Arch Toxicol 77: 384-391.
http://dx.doi.org/10.1007/s00204-003-0461-z.
Cooper. SP: Burau. K: Sweeney. A: Robison. T: Smith. MA: Symanski. E: Colt. IS: Laseter. I: Zahm. SH. (2001).
Prenatal exposure to pesticides: a feasibility study among migrant and seasonal farmworkers. Am J
Ind Med 40: 578-585.
Dahl. AR: Damon. EG: Mauderly. IL: Rothenberg. SI: Seller. FA: Mcclellan. RO. (1988). Uptake of 19
hydrocarbon vapors inhaled by F344 rats. Fundam Appl Toxicol 10: 262-269.
http://dx.doi.org/10.1016/0272-0590(88)90310-7.
Dowty. Bl: Laseter. IL: Storer. I. (1976). The transplacental migration and accumulation in blood of volatile
organic constituents. Pediatr Res 10: 696-701.
Eide. I: Zahlsen. K. (1996). Inhalation experiments with mixtures of hydrocarbons. Experimental design,
statistics and interpretation of kinetics and possible interactions. Arch Toxicol 70: 397-404.
http://dx.doi.org/10.1007/s002040050291.
Gaschen. A: Lang. D: Kalberer. M: Savi. M: Geiser. T: Gazdhar. A: Lehr. CM: Bur. M: Dommen. I: Baltensperger.
U: Geiser. M. (2010). Cellular responses after exposure of lung cell cultures to secondary organic
aerosol particles. Environ Sci Technol 44: 1424-1430. http://dx.doi.org/10.1021/es902261m.
Ginsberg. G: W Ir. S: Bruckner. I: Sonawane. B. (2004). Incorporating children's toxicokinetics into a risk
framework. Environ Health Perspect 112: 272-283. http://dx.doi.org/10.1289/ehp.6013.
Gralewicz. S: Wiaderna. D: Tomas. T: Rydzynski. K. (1997a). Behavioral changes following 4-week inhalation
exposure to pseudocumene (1,2,4-trimethylbenzene) in the rat. Neurotoxicol Teratol 19: 327-333.
http://dx.doi.org/10.1016/S0892-0362(97)00001-9.
Gralewicz. S: Wiaderna. D: Tomas. T. (1997b). Retardation of the age-related increase in spontaneous cortical
spike-wave discharges (SWD) in rats after a 28-day inhalation (SWD) in rats after a 28-day inhalation
exposure to an industrial solvent, pseudocumene (1,2,4-trimethylbenzene). Int J Occup Med Environ
Health 10: 213-222.
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.
http://dx.doi.org/10.1016/S0161-813X(00)00003-6.
Guo. H: Kwok. NH: Cheng. HR: Lee. SC: Hung. WT: Li. YS. (2009). Formaldehyde and volatile organic
compounds in Hong Kong homes: Concentrations and impact factors. Indoor Air 19: 206-217.
http://dx.doi.0rg/10.llll/j.1600-0668.2008.00580.x.
Henderson. RF. (2001). Aromatic hydrocarbons: Benzene and other alkylbenzenes. 4: 231-301.
Hillefors-Berglund. M: Liu. Y: von Euler. G. (1995). Persistent, specific and dose-dependent effects of toluene
exposure on dopamine D2 agonist binding in the rat caudate-putamen. Toxicology 100:185-194.
http://dx.doi.org/10.1016/0300-483X(95)03084-S.
Hissink. AM: Kriise. I: Kulig. BM: Verwei. M: Muijser. H: Salmon. F: Leenheers. LH: Owen. DE: Lammers. IH:
Freidig. AP: McKee. RH. (2007). Model studies for evaluating the neurobehavioral effects of complex
This document is a draft for review purposes only and does not constitute Agency policy.
R-2 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
hydrocarbon solvents III. PBPK modeling of white spirit constituents as a tool for integrating animal
and human test data. Neurotoxicology 28: 751-760. http://dx.doi.Org/10.1016/j.neuro.2007.03.005.
HSDB. (Hazardous Substances Data Bank). (2011a). 1,2,3-trimethylbenzene. Bethesda, MD: National Library
of Medicine.
HSDB. (Hazardous Substances Data Bank). (2011b). 1,2,4-Trimethylbenzene [Database]. Bethesda, MD:
National Library of Medicine. Retrieved from http://toxnet.nlm.nih.gov
HSDB. (Hazardous Substances Data Bank). (2011c). 1,3,5-Trimethylbenzene [Database]. Bethesda, MD:
National Library of Medicine. Retrieved from http://toxnet.nlm.nih.gov
Huo. IZ: Aldous. S: Campbell. K: Davies. N. (1989). Distribution and metabolism of 1,2,4-trimethylbenzene
(pseudocumene) in the rat. Xenobiotica 19: 161-170.
http://dx.doi.org/10.3109/00498258909034688.
Industrial Bio-Test Laboratories. Inc... (1992). Four-week subacute aerosol inhalation toxicity study with
MCS-1809 in albino rats. (88920007305; OTS0545631). St. Louis, MO: Monsanto Company.
lackson. DM: Westlind-Danielsson. A. (1994). Dopamine receptors: Molecular biology, biochemistry and
behavioural aspects. Pharmacol Ther 64: 291-370. http://dx.doi.org/10.1016/0163-7258(94)90041-
8.
lanik-Spiechowicz. E: Wyszynska. K: Dziubaltowska. E. (1998). Genotoxicity evaluation of trimethylbenzenes.
Mutat Res Genet Toxicol Environ Mutagen 412: 299-305. http://dx.doi.org/10.1016/S1383-
5718(97)00202-7.
larnberg. I: lohanson. G. (1995). Liquid/air partition coefficients of the trimethylbenzenes. Toxicol Ind Health
11: 81-88. http://dx.doi.org/10.1177/074823379501100107.
larnberg. I: lohanson. G: Lof. A. (1996). Toxicokinetics of inhaled trimethylbenzenes in man. Toxicol Appl
Pharmacol 140: 281-288. http://dx.doi.org/10.1006/taap.1996.0223.
larnberg. I: lohanson. G: Lof. A: Stahlbom. B. (1997a). Inhalation toxicokinetics of 1,2,4-trimethylbenzene in
volunteers: Comparison between exposure to white spirit and 1,2,4-trimethylbenzene alone. Sci
Total Environ 199: 65-71. http://dx.doi.org/10.1016/S0048-9697(97)05482-X.
larnberg. I: Stahlbon. B: lohanson. G: Lof. A. (1997b). Urinary excretion of dimethylhippuric acids in humans
after exposure to trimethylbenzenes. Int Arch Occup Environ Health 69: 491-497.
http://dx.doi.org/10.1007/s004200050179.
larnberg. I: lohanson. G: Lof. A: Stahlbom. B. (1998). Toxicokinetics of 1,2,4-trimethylbenzene in humans
exposed to vapours of white spirit: Comparison with exposure to 1,2,4-trimethylbenzene alone. Arch
Toxicol 72: 483-491. http://dx.doi.org/10.1007/s002040050532.
liun-Horng. T: Kuo-Hsiung. L: Chih-Yu. C: Nina. L: Sen-Yi. M: Hung-Lung. C. (2008). Volatile organic compound
constituents from an integrated iron and steel facility. J Hazard Mater 157: 569-578.
http://dx.doi.0rg/10.1016/j.jhazmat.2008.01.022.
lones. K: Meldrum. M: Baird. E: Cottrell. S: Kaur. P: Plant. N: Dyne. D: Cocker. I. (2006). Biological monitoring
for trimethylbenzene exposure: A human volunteer study and a practical example in the workplace.
Ann Occup Hyg 50: 593-598. http://dx.doi.org/10.1093/annhyg/mel016.
Kaspar. BK: Llado. I: Sherkat. N: Rothstein. ID: Gage. FH. (2003). Retrograde viral delivery of IGF-1 prolongs
survival in a mouse ALS model. Science 301: 839-842. http://dx.doi.org/10.1126/science.1086137.
Koch Industries. (Koch Industries, Incorporated). (1995a). 14-day oral gavage toxicity study of 1,3,5-
trimethylbenzene in rats with a recovery group, with cover letter dated 2/7/95. (44616). Wichita,
KS. http://www.ntis.gov/search/product.aspx?ABBR=OTS0558836.
Koch Industries. (Koch Industries, Incorporated). (1995b). 90-day oral gavage toxicity study of 1,3,5-
trimethylbenzene in rats with a recovery group. (44618). Wichita, KS: Koch Industries, Inc.
This document is a draft for review purposes only and does not constitute Agency policy.
R-3 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Korsak. Z: Swiercz. R: Rydzynski. K. (1995). Toxic effects of acute inhalation exposure to 1,2,4-
trimethylbenzene (pseudocumene) in experimental animals. Int J Occup Med Environ Health 8: 331-
337.
Korsak. Z: Rydzynski. K. (1996). Neurotoxic effects of acute and subchronic inhalation exposure to
trimethylbenzene isomers (pseudocumene, mesitylene, hemimellitene) in rats. Int J Occup Med
Environ Health 9: 341-349.
Korsak. Z: Rydzynski. K: lajte. I. (1997). Respiratory irritative effects of trimethylbenzenes: An experimental
animal study. Int J Occup Med Environ Health 10: 303-311.
Korsak. Z: Stetkiewicz. I: Majcherek. W: Stetkiewicz. I: lajte. I: Rydzynski. K. (2000a). Sub-chronic inhalation
toxicity of 1,2,4-trimethylbenzene (pseudocumene) in rats. Int J Occup Med Environ Health 13:155-
164.
Korsak. Z: Stetkiewicz. I: Majcherek. W: Stetkiewicz. I: lajte. I: Rydzynski. K. (2000b). Subchronic inhalation
toxicity of 1,2,3-trimethylbenzene (hemimellitene) in rats. Int J Occup Med Environ Health 13: 223-
232.
Kostrewski. P: Wiaderna-Brycht. A. (1995). Kinetics of elimination of mesitylene and 3,5-dimethylbenzoic
acid after experimental human exposure. Toxicol Lett 77: 259-264. http://dx.doi.org/10.1016/0378-
4274(95)03305-X.
Kostrzewski. P: Wiaderna-Brycht. A: Czerski. B. (1997). Biological monitoring of experimental human
exposure to trimethylbenzene. Sci Total Environ 199: 73-81. http://dx.doi.org/10.1016/S0048-
9697(97)05504-6.
Kyrklund. T. (1992). The use of experimental studies to reveal suspected neurotoxic chemicals as
occupational hazards: Acute and chronic exposures to organic solvents. Am J Ind Med 21:15-24.
http://dx.doi.org/10.1002/ajim.4700210105.
Lammers. IH: Emmen. HH: Muijser. H: Hoogendijk. EM: McKee. RH: Owen. DE: Kulig. BM. (2007). Model
studies for evaluating the neurobehavioral effects of complex hydrocarbon solvents II.
Neurobehavioral effects of white spirit in rat and human. Neurotoxicology 28: 736-750.
http://dx.doi.0rg/10.1016/j.neuro.2007.03.003.
Lee. CR: leong. KS: Kim. Y: Yoo. CI: Lee. IH: Choi. YH. (2005). Neurobehavioral changes of shipyard painters
exposed to mixed organic solvents. Ind Health 43: 320-326.
Lutz. P: Gralewicz. S: Wiaderna. D: Swiercz. R: Grzelinska. Z: Majcherek. W. (2010). Contrasting effects of 4-
week inhalation exposure to pseudocumene or hemimellitene on sensitivity to amphetamine and
propensity to amphetamine sensitization in the rat. Int J Occup Med Environ Health 23: 85-94.
http://dx.doi.org/10.2478/vl0001-010-0005-8.
Maltoni. C: Ciliberti. A: Pinto. C: Soffritti. M: Belpoggi. F: Menarini. L. (1997). Results of long-term
experimental carcinogenicity studies of the effects of gasoline, correlated fuels, and major gasoline
aromatics on rats. Ann N Y Acad Sci 837:15-52. http://dx.doi.org/10.llll/).1749-
6632.1997.tb56863.x.
Martins. EM: Arbilla. G: Gatti. LV. (2010). Volatile organic compounds in a residential and commercial urban
area with a diesel, compressed natural gas and oxygenated gasoline vehicular fleet. Bull Environ
Contam Toxicol 84:175-179. http://dx.doi.org/10.1007/s00128-009-9886-2.
McKee. RH: Wong. ZA: Schmitt. S: Beatty. P: Swanson. M: Schreiner. CA: Schardein. IL. (1990). The
reproductive and developmental toxicity of High Flash Aromatic Naphtha. Toxicol Ind Health 6: 441-
460.
McKee. RH: Lammers. IH: Muijser. H: Owen. DE: Kulig. BM. (2010). Neurobehavioral effects of acute exposure
to aromatic hydrocarbons. Int J Toxicol 29: 277-290.
http://dx.doi.org/10.1177/1091581810365089.
This document is a draft for review purposes only and does not constitute Agency policy.
R-4 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Mclanahan. ED: El-Masri. HA: Sweeney. LM: Kopylev. LY: Clewell. HI: Wambaugh. IF: Schlosser. PM. (2012).
Physiologically based pharmacokinetic model use in risk assessment—why being published is not
enough. Toxicol Sci 126: 5-15. http://dx.doi.org/10.1093/toxsci/kfr295.
Meulenberg. C: Vijverberg. H. (2000). Empirical relations predicting human and rat tissue: Air partition
coefficients of volatile organic compounds. Toxicol Appl Pharmacol 165: 206-216.
http://dx.doi.org/10.1006/taap.2000.8929.
Mikulski. PI: Wiglusz. R. (1975). The comparative metabolism of mesitylene, pseudocumene, and
hemimellitene in rats. Toxicol Appl Pharmacol 31: 21-31. http://dx.doi.org/10.1016/0041-
008X(75)90048-4.
MOE. (Ontario Ministry of the Environment). (2006). Rationale for the development of Ontario air standards
for trimethylbenzenes: 1,2,3-Trimethylbenzene. Ontario, Canada.
Mogel. I: Baumann. S: Bohme. A: Kohajda. T: von Bergen. M: Simon. 1C: Lehmann. I. (2011). The aromatic
volatile organic compounds toluene, benzene and styrene induce COX-2 and prostaglandins in human
lung epithelial cells via oxidative stress and p38 MAPK activation. Toxicology 289: 28-37.
http://dx.doi.0rg/10.1016/j.tox.2011.07.006.
Myhre. 0: Vestad. TA: Sagstuen. E: Aarnes. H: Fonnum. F. (2000). The effects of aliphatic (n-nonane),
naphtenic (1,2,4-trimethylcyclohexane), and aromatic (1,2,4-trimethylbenzene) hydrocarbons on
respiratory burst in human neutrophil granulocytes. Toxicol Appl Pharmacol 167: 222-230.
http://dx.doi.org/10.1006/taap.2000.9008.
Myhre. 0: Fonnum. F. (2001). The effect of aliphatic, naphthenic, and aromatic hydrocarbons on production of
reactive oxygen species and reactive nitrogen species in rat brain synaptosome fraction: the
involvement of calcium, nitric oxide synthase, mitochondria, and phospholipase A. Biochem
Pharmacol 62:119-128. http://dx.doi.org/10.1016/S0006-2952(01)00652-9.
Norseth. T: Waage. I: Dale. I. (1991). Acute effects and exposure to organic compounds in road maintenance
workers exposed to asphalt. Am J Ind Med 20: 737-744.
http://dx.doi.org/10.1002/ajim.4700200604.
OSHA. (Occupational Safety & Health Administration). (1996). Occupational safety and health guideline for
trimethylbenzene Retrieved August 1, 2007, from
http://www.osha.gov/SLTC/healthguidelines/trimethylbenzene/recognition.html
Ramaiah. SK. (2007). A toxicologist guide to the diagnostic interpretation of hepatic biochemical parameters.
Food Chem Toxicol 45: 1551-1557. http://dx.doi.Org/10.1016/j.fct.2007.06.007.
Rea. TM: Nash. IF: Zabik. IE: Born. GS: Kessler. WV. (1984). Effects of toluene inhalation on brain biogenic
amines in the rat. Toxicology 31:143-150. http://dx.doi.org/10.1016/0300-483X(84)90006-4.
Reyes. L: Manalich. R. (2005). Long-term consequences of low birth weight. Kidney IntSupplS107-Slll.
http://dx.doi.0rg/10.llll/j.1523-1755.2005.09718.x.
Saillenfait. AM: Gallissot. F: Sabate. IP: Morel. G. (2005). Developmental toxicity of two trimethylbenzene
isomers, mesitylene and pseudocumene, in rats following inhalation exposure. Food Chem Toxicol
43: 1055-1063. http://dx.doi.Org/10.1016/j.fct.2005.02.008.
Snead. OC. III. (1995). Basic mechanisms of generalized absence seizures. Ann Neurol 37:146-157.
http://dx.doi.org/10.1002/ana.410370204.
Sulkowski. Wl: Kowalska. S: Matyja. W: Guzek. W: Wesolowski. W: Szymczak. W: Kostrzewski. P. (2002).
Effects of occupational exposure to a mixture of solvents on the inner ear: A field study. Int J Occup
Med Environ Health 15: 247-256.
Swiercz. R: Rydzyriski. K: Wasowicz. W: Majcherek. W: Wesolowski. W. (2002). Toxicokinetics and
metabolism of pseudocumene (1,2,4-trimethylbenzene) after inhalation exposure in rats. Int J Occup
Med Environ Health 15: 37-42.
This document is a draft for review purposes only and does not constitute Agency policy.
R-5 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Swiercz. R: Wiaderna. D: Wasowicz. W: Rydzyriski. K. (2003). Pseudocumene in brain, liver, lung and blood of
rats after single and repeated inhalation exposure. Int J Occup Med Environ Health 16: 61-66.
Swiercz. R: Wasowicz. W: Majcherek. W. (2006). Mesitylene (1,3,5-trimethylbenzene) in the liver, lung,
kidney, and blood and 3,5-dimethylbenzoic acid in the liver, lung, kidney and urine of rats after single
and repeated inhalation exposure to mesitylene. Pol J Environ Stud 15: 485-492.
Tomas. T: Lutz. P: Wiaderna. D. (1999a). Changes in electrocortical arousal following acute trimethylbenzene
administration in rats. Int J Occup Med Environ Health 12: 67-78.
Tomas. T: Swiercz. R: Wiaderna. D: Gralewicz. S. (1999b). Effects of acute exposure to aromatic hydrocarbons
C 9 on locomotor activity in rats. Trimethylbenzene isomers. Int J Occup Med Environ Health 12: 331-
343.
Tomas. T: Wiaderna. D: Swiercz. R. (1999c). Neurotoxicity assessment of selected organic solvents based on
spontaneous and evoked cortical and hippocampal activity in rats. Int J Occup Med Environ Health
12: 73-84.
TRI. (Toxic Release Inventory). (2008). Toxic Release Inventory [Database]: U.S. Environmental Protection
Agency.
Tsujimoto. Y: Noda. T: Shimizu. M: Moriwaki. H: Tanaka. M. (2000). Identification of the dimethylbenzyl
mercapturic acid in urine of rats administered with 1,2,4-trimethylbenzene. Chemosphere 40: 893-
896. http://dx.doi.org/10.1016/S0045-6535(99)00467-l.
Tsujimoto. Y: Warashina. M: Nam. VD: Noda. T: Shimizu. M: Yamaguchi. Y: Moriwaki. H: Morimoto. T:
Kakiuchi. K: Maeda. Y: Tanaka. M. (2005). Determination of urinary phenolic metabolites from rats
treated with 1,2,3-and 1,3,5-trimethylbenzenes. J Occup Health 47: 337-339.
U.S. EPA. (U.S. Environmental Protection Agency). (1987). Health effects assessment for trimethylbenzenes.
(EPA/600/8-88/060). Cincinnati, OH: U.S. Environmental Protection Agency, Office of Health and
Environmental Assessment. http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=2000T8ZG.txt.
U.S. EPA. (U.S. Environmental Protection Agency). (1991). Guidelines for developmental toxicity risk
assessment [EPA Report]. (EPA/600/FR-91/001). Washington, DC: U.S. Environmental Protection
Agency, Risk Assessment Forum, http://www.epa.gov/iris/backgrd.html.
U.S. EPA. (U.S. Environmental Protection Agency). (1994a). Chemical summary for 1,2,4-trimethylbenzene.
(EPA/749/F-94/022A). Washington, DC: U.S. Environmental Protection Agency, Office of Pollution
Prevention and Toxic Substances, http://www.epa.gov/chemfact/s_trimet.txt.
U.S. EPA. (U.S. Environmental Protection Agency). (1994b). Methods for derivation of inhalation reference
concentrations and application of inhalation dosimetry. (EPA/600/8-90/066F). Research Triangle
Park, NC: U.S. Environmental Protection Agency, Office of Research and Development, Office of Health
and Environmental Assessment, Environmental Criteria and Assessment Office.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=71993.
U.S. EPA. (U.S. Environmental Protection Agency). (1998). Guidelines for neurotoxicity risk assessment.
(EPA/630/R-95/001F). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment
Forum. http://www.epa.gov/raf/publications/pdfs/NEUROTOX.PDF.
U.S. EPA. (U.S. Environmental Protection Agency). (2000). Benchmark dose technical guidance document
[external review draft]. (EPA/630/R-00/001). Washington, DC: U.S. Environmental Protection
Agency, Risk Assessment Forum, http://www.epa.gov/raf/publications/benchmark-dose-doc-
drafLhtm.
U.S. EPA. (U.S. Environmental Protection Agency). (2002). A review of the reference dose and reference
concentration processes. (EPA/630/P-02/002F). Washington, DC.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=51717.
U.S. EPA. (U.S. Environmental Protection Agency). (2005). Guidelines for carcinogen risk assessment.
(EPA/630/P-03/001F). Washington, DC. http://www.epa.gov/cancerguidelines/.
This document is a draft for review purposes only and does not constitute Agency policy.
R-6 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
von Euler. G: Ogren. SO: Li. XM: Fuxe. K: Gustafsson. IA. (1993). Persistent effects of subchronic toluene
exposure on spatial learning and memory, dopamine-mediated locomotor activity and dopamine D2
agonist binding in the rat. Toxicology 77: 223-232. http://dx.doi.org/10.1016/0300-
483X(93)90162-L.
von Euler, G: Ogren, SO: Eneroth, P: Fuxe, K: Gustafsson, IA. (1994). Persistent effects of 80 ppm toluene on
dopamine-regulated locomotor activity and prolactin secretion in the male rat. Neurotoxicology 15:
621-624.
Warter. IM: Vergnes. M: Depaulis. A: Tranchant. C: Rumbach. L: Micheletti. G: Marescaux. C. (1988). Effects of
drugs affecting dopaminergic neurotransmission in rats with spontaneous petit mal-like seizures.
Neuropharmacology 27: 269-274. http://dx.doi.org/10.1016/0028-3908(88)90043-3.
Wiaderna. D: Gralewicz. S: Tomas. T. (1998). Behavioral changes following a four-week inhalation exposure to
hemimellitene (1,2,3-trimethylbenzene) in rats. Int J Occup Med Environ Health 11: 319-334.
Wiaderna. D: Gralewicz. S: Tomas. T. (2002). Assessment of long-term neurotoxic effects of exposure to
mesitylene (1,3,5-trimethylbenzene) based on the analysis of selected behavioral responses. Int J
Occup Med Environ Health 15: 385-392.
Wiglusz. R: Kienitz. M: Delag. G: Galuszko. E: Mikulski. P. (1975a). Peripheral blood of mesitylene vapour
treated rats. Bull Inst MaritTrop Med Gdynia 26: 315-321.
Wiglusz. R: Delag. G: Mikulski. P. (1975b). Serum enzymes activity of mesitylene vapour treated rats. Bull Inst
MaritTrop Med Gdynia 26: 303-313.
Wiglusz. R. (1979). The effect of 1, 3, 5-trimethylbenzene inhalation exposure on the glucuronic acid pathway
and activity of some xenobiotic-metabolizing enzymes. Bull Inst Marit Trop Med Gdynia 30:189-196.
This document is a draft for review purposes only and does not constitute Agency policy.
R-7 DRAFT—DO NOT CITE OR QUOTE
------- |