Toxicological (CASRN 25551-13-7, 95-63-6, Review of Trimethylbenzenes In Support of Summary Information 526-73-8, and on 108-67-8) the Integrated Risk Information System (IRIS) Supplemental Information August 2013 Revised External Review Draft

4>EPA
United States
Environmental Protection
Agency
EPA/63 5/R-13/171b
Revised External Review Draft
www.epa.gov/iris
Toxicological Review of Trimethylbenzenes
(CASRN 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)
Supplemental Information
August 2013
NOTICE
This document is a Revised 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 ofTrimethylbenzene
DISCLAIMER
This document is a preliminary draft for review purposes only. 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.
Mention of trade names or commercial products does not constitute endorsement of
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ii	DRAFT—DO NOT CITE OR QUOTE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 ofTrimethylbenzene
CONTENTS	
TABLES 	iv
FIGURES 	viii
ABBREVIATIONS	x
APPENDIX A. HEALTH ASSESSMENTS AND REGULATORY LIMITS BY OTHER NATIONAL AND
INTERNATIONAL HEALTH AGENCIES	A-l
APPENDIX B. INFORMATION IN SUPPORT OF HAZARD IDENTIFICATION AND DOSE-REPONSE ANALYSIS	B-l
B.l. PHYSICAL AND CHEMICAL PROPERTIES	B-l
B.2. TOXICOKINETICS 	B-2
B.2.1. Absorption	B-2
B.2.2. Distribution	B-4
B.2.3. Metabolism	B-6
B.2.4. Excretion	B-10
B.3. PHYSIOLOGICALLY-BASED PHARMACOKINETIC MODELS	B-ll
B.3.1. Summary of Available PBPK models for 1,2,4-TMB	B-ll
B.3.2. 1,2,4-TMB PBPK Model Selection	B-20
B.3.3. Details of Hissink et al. (2007) Model Analysis	B-21
B.3.4.	Summary of Available PBPK models for 1,3,5-TMB or 1,2,3-TMB	B-55
B.4. HUMAN STUDIES 	B-56
B.5. ANIMAL TOXICOLOGY STUDIES	B-71
B.6. HUMAN TOXICOKINETIC STUDIES	B-162
B.7. ANIMALTOXICOKINETICSTUDIES	B-177
B.8.	ANIMAL AND HUMAN TOXICOKINETIC STUDIES	B-197
APPENDIX C. DOSE-RESPONSE MODELING FOR THE DERIVATION OF REFERENCE VALUES FOR EFFECTS
OTHER THAN CANCER AND THE DERIVATION OF CANCER RISK ESTIMATES	C-l
C.l.	BENCHMARK DOSE MODELING SUMMARY	C-l
C.l.l.	Non-Cancer Endpoints	C-l
C.2. BENCHMARK DOSE MODELING SUMMARY - ALTERNATIVE ANALYSIS WITH HIGH DOSES
INCLUDED	C-33
APPENDIX D. DOCUMENTATION OF IMPLEMENTATION OF THE 2011 NATIONAL RESEARCH COUNCIL
RECOMMENDATIONS	D-l
APPENDIX E. SUMMARY OF AVAILABLE C9 AROMATIC HYDROCARBON FRACTION TOXICITY STUDIES	E-l
APPENDIX F. RESOLUTION OF PUBLIC COMMENTS	F-l
REFERENCES FOR APPENDICES	R-l
This document is a draft for review purposes only and does not constitute Agency policy.
in	DRAFT—DO NOT CITE OR QUOTE
This document is a draft for review purposes only and does not constitute Agency policy.

-------
Toxicological Review ofTrimethylbenzene
TABLES
Table A-l. Other national and international health agency assessments for TMBs A-l
Table B-l. Physical properties and chemical identity of 1,2,4-TMB, 1,3,5-TMB, and 1,2,3-TMB B-l
Table B-2. Measured and calculated partition coefficients for TMB isomers at 37°C B-14
Table B-3. PBPK model parameters for 1,2,4-TMB toxicokinetics in humans using the Jarnberg and
Johanson (1999) model structure B-15
Table B-4. Comparison of rat anatomical and physiological parameters in Hissink et al. (2007) to those of
Brown et al. (1997) B-24
Table B-5. Comparison of human anatomical and physiological parameters in Hissink et al. (2007) to
those of Williams and Leggett (1989) as reported by Brown et al. (1997) B-25
Table B-6. Comparison of chemical-specific parameters in Hissink et al. (2007) to literature data B-26
Table B-7. Parameter values for the rat and human PBPK models for 1,2,4 TMB used by EPA B-32
Table B-8. Rat 1,2,4-TMB kinetic studies used in model development and testing B-33
Table B-9. Model simulated and experimental measured concentrations of 1,2,4 TMB in male Wistar rats
exposed to 1,2,4-TMB, Swiercz et al. (2003) B-38
Table B-10. Model simulated and experimental measured concentrations of 1,2,4-TMB in male Sprague-
Dawley rats exposed to 100 ppm (492 mg/m3) 1,2,4-TMB (12 hr/day, for 3 days) at the end
of exposure or 12 hours after the last exposure B-39
Table B-ll. Model simulated and experimental measured concentrations of 1,2,4-TMB in male Sprague-
Dawley rats exposed to 1,2,4-TMB at the end of 12 hour exposure B-41
Table B-12. Model simulated and experimental measured concentrations of 1,2,4-TMB in male Sprague-
Dawley rats exposed to 1,000 ppm (4,920 mg/m3) 1,2,4-TMB (12 hr/day, for 14 days) at the
end of exposure B-42
Table B-13. Human kinetic studies of 1,2,4-TMB used in model validation B-43
Table B-14. Parameter sensitivity for venous blood 1,2,4-TMB concentration in rats exposed to
1,2,4-TMB via inhalation B-50
Table B-15. Parameter sensitivity for steady-state venous blood 1,2,4-TMB concentration in humans
exposed to 1,2,4-TMB via inhalation B-53
Table B-16. Characteristics and quantitative results for epidemiologic cross-sectional study of exposure
to 1,2,4-TMB. Battig et al. (1956), as reviewed by Baettig et al. (1958) B-56
Table B-17. Characteristics and quantitative results for epidemiologic cross-sectional study of exposure
to 1,2,4-TMB; Billionnet et al. (2011) B-58
Table B-18. Characteristics and quantitative results for epidemiologic cohort study of exposure to
1,2,4-TMB. Chen et al. (1999) B-60
Table B-19. Characteristics and quantitative results for controlled human exposure study of exposure to
1,2,4-TMB in WS. Lammers et al. (2007) B-64
Table B-20. Characteristics and quantitative results for epidemiologic cohort study of exposure to
1,2,4-TMB. Lee et al. (2005) B-66
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
e B-21. Characteristics and quantitative resu
to 1,2,4-TMB; Norseth et al. (1991)..
e B-22. Characteristics and quantitative resu
to 1,2,4-TMB Sulkowski et al. (2002)
e B-23. Characteristics and quantitative resu
e B-24. Characteristics and quantitative resu
e B-25. Characteristics and quantitative resu
e B-26. Characteristics and quantitative resu
e B-27. Characteristics and quantitative resu
e B-28. Characteristics and quantitative resu
e B-29. Characteristics and quantitative resu
e B-30. Characteristics and quantitative resu
e B-31. Characteristics and quantitative resu
e B-32. Characteristics and quantitative resu
e B-33. Characteristics and quantitative resu
e B-34. Characteristics and quantitative resu
e B-35. Characteristics and quantitative resu
e B-36. Characteristics and quantitative resu
e B-37. Characteristics and quantitative resu
e B-38. Characteristics and quantitative resu
e B-39. Characteristics and quantitative resu
e B-40. Characteristics and quantitative resu
e B-41. Characteristics and quantitative resu
e B-42. Characteristics and quantitative resu
e B-43. Characteristics and quantitative resu
e B-44. Characteristics and quantitative resu
e B-45. Characteristics and quantitative resu
e B-46. Characteristics and quantitative resu
e B-47. Characteristics and quantitative resu
e B-48. Characteristics and quantitative resu
e B-49. Characteristics and quantitative resu
e B-50. Characteristics and quantitative resu
e B-51. Characteristics and quantitative resu
e B-52. Characteristics and quantitative resu
e B-53. Characteristics and quantitative resu
e B-54. Characteristics and quantitative resu
s for epidemiologic cross-sectional study of exposure
s for epidemiologic cross-sectional study of exposure
B-68
B-70
s for Baettig et al. (1958)	B-71
s for Gralewicz et al. (1997b)	B-76
s for Gralewicz et al. (1997a)	B-80
s for Gralewicz and Wiaderna (2001)	B-82
s for Janik-Speichowicz et al. (1998)	B-85
sfor Koch Industries (1995b)	B-89
s for Korsak et al. (1995)	B-104
s for Korsak and Rydzynski (1996)	B-107
s for Korsak et al. (1997)	B-lll
s for Korsak et al. (2000a)	B-113
s for Korsak et al. (2000b)	B-118
s for Lammers et al. (2007)	B-123
s for Lutz et al. (2010)	B-126
s for Maltoni et al. (1997)	B-131
s for McKee et al. (2010)	B-133
s for Saillenfait et al. (2005)	B-136
s for Tomas et al. (1999a)	B-141
s for Tomas et al. (1999b)	B-143
s for Tomas et al. (1999c)	B-144
s for Wiaderna et al. (1998)	B-147
s for Wiaderna et al. (2002)	B-151
s for Wiglusz et al. (1975b)	B-153
s for Wiglusz et al. (1975a)	B-157
s for Jarnberg et al. (1996)	B-162
s for Jarnberg et al. (1997a)	B-166
s for Jarnberg et al. (1997b)	B-168
s for Jarnberg et al. (1998)	B-169
s for Jones et al. (2006)	B-171
s for Kostrzewski et al. (1997)	B-173
s for Dahl et al. (1988)	B-177
s for Eide and Zahlsen et al. (1996)	B-178
s for Huo et al. (1989)	B-179
This document is a draft for review purposes only and does not constitute Agency policy.
v	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
Table B-55. Characteristics and quantitative results for Mikulski and Wiglusz (1975)	B-181
Table B-56. Characteristics and quantitative results for Swiercz et al. (2002)	B-182
Table B-57. Characteristics and quantitative results for Swiercz et al. (2003)	B-184
Table B-58. Characteristics and quantitative results for Swiercz et al. (2006)	B-186
Table B-59. Characteristics and quantitative results for Tsujimoto et al. (2000)	B-189
Table B-60. Characteristics and quantitative results for Tsujimoto et al. (2005)	B-190
Table B-61. Characteristics and quantitative results for Tsujino et al. (2002)	B-191
Table B-62. Characteristics and quantitative results for Zahlsen et al. (1990)	B-193
Table B-63. Characteristics and quantitative results for Zahlsen et al. (1992)	B-196
Table B-64. Characteristics and quantitative results for Meulenberg and Vijverberg (2000)	B-197
Table C-l. Non-cancer endpoints selected for dose-response modeling for 1,2,3-TMB, 1,2,4-TMB, and
1,3,5-TMB	C-3
Table C-2. Summary of BMD modeling results for increased latency to paw-lick in male Wistar rats
exposed to 1,2,4-TMB by inhalation for 3 months; BMR = 1SD change from control
mean(constant variance, high dose dropped), (Korsak and Rydzynski, 1996)	C-6
Table C-3. Summary of BMD modeling results for decreased red blood cells in male Wistar rats exposed
to 1,2,4-TMB by inhalation for 3 months; BMR = 1 SD change from control mean (constant
variance, high dose dropped), (Korsak et al., 2000a)	C-8
Table C-4. Summary of BMD modeling results for decreased clotting time in female Wistar rats exposed
to 1,2,4-TMB by inhalation for 3 months; BMR = 1 SD change from control mean (constant
and modeled variance, high dose dropped) (Korsak et al., 2000a)	C-10
Table C-5. Summary of BMD modeling results for decreased fetal weight in male Sprague-Dawley rats
exposed to 1,2,4-TMB by maternal inhalation on GD6-GD20; BMR = 1 SD or 5% change from
control mean (constant variance)(Saillenfait et al., 2005)	C-ll
Table C-6. Summary of BMD modeling results for decreased fetal weight in female Sprague-Dawley rats
exposed to 1,2,4-TMB by maternal inhalation on GD6-GD20; BMR = 1 SD or 5% change from
control mean (constant variance; Saillenfait et al., 2005)	C-14
Table C-l. Summary of BMD modeling results for decreased maternal body weight gain in female
Sprague-Dawley rats exposed to 1,2,4-TMB by inhalation on GD6-GD20; BMR = 1 SD change
from control mean (constant variance) (Saillenfait et al., 2005)	C-17
Table C-8. Summary of BMD modeling results for increased latency to paw-lick in male Wistar rats
exposed to 1,2,3-TMB by inhalation for 3 months; BMR = 1 SD change from control
mean(constant variance and modeled variance), (Korsak and Rydzynski, 1996)	C-19
Table C-9. Summary of BMD modeling results for increased latency to paw-lick in male Wistar rats
exposed to 1,2,3-TMB by inhalation for 3 months; BMR = 1 SD change from control
mean(modeled variance, high dose dropped), (Korsak and Rydzynski, 1996)	C-20
Table C-10. Summary of BMD modeling results for decreased segmented neutrophils in male Wistar rats
exposed to 1,2,3-TMB by inhalation for 3 months; BMR = 1 SD change from control
mean(constant variance), (Korsak et al., 2000b)	C-22
Table C-ll. Summary of BMD modeling results for decreased segmented neutrophils in female Wistar
rats exposed to 1,2,3-TMB by inhalation for 3 months; BMR = 1 SD change from control
mean(constant variance), (Korsak et al., 2000b)	C-24
This document is a draft for review purposes only and does not constitute Agency policy.
vi	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
Table C-12. Summary of BMD modeling results for increased reticulocytes in male Wistar rats exposed
to 1,2,3-TMB by inhalation for 3 months; BMR = 1SD change from control mean(constant
variance), (Korsak et al., 2000b)	C-26
Table C-13. Summary of BMD modeling results for decreased fetal weight in male Sprague-Dawley rats
exposed to 1,3,5-TMB by maternal inhalation on GD6-GD20; BMR = 1 SD change from
control mean (constant and modeled variance)(Saillenfait et al., 2005)	C-28
Table C-14. Summary of BMD modeling results for decreased fetal weight in female Sprague-Dawley rats
exposed to 1,3,5-TMB by maternal inhalation on GD6-GD20; BMR = 1 SD change from
control mean (constant and modeled variance)(Saillenfait et al., 2005)	C-29
Table C-15. Summary of BMD modeling results for decreased maternal body weight gain in female
Sprague-Dawley rats exposed to 1,3,5-TMB by inhalation on GD6-GD20; BMR = 1 SD change
from control mean (constant and modeled variance), (Saillenfait et al., 2005)	C-30
Table C-16. Summary of BMD modeling results for increased latency to paw-lick in male Wistar rats
exposed to 1,2,4-TMB by inhalation for 3 months; BMR = 1 SD change from control
mean(constant and modeled variance), (Korsak and Rydzynski, 1996)	C-33
Table C-17. Summary of BMD modeling results for decreased red blood cells in male Wistar rats exposed
to 1,2,4-TMB by inhalation for 3 months; BMR = 1 SD change from control mean (constant
variance), (Korsak et al., 2000a)	C-34
Table C-18. Summary of BMD modeling results for decreased clotting time in female Wistar rats exposed
to 1,2,4-TMB by inhalation for 3 months; BMR = 1 SD change from control mean (constant
and modeled variance), (Korsak et al., 2000a)	C-36
Table C-19. Summary of BMD modeling results for decreased reticulocytes in female Wistar rats
exposed to 1,2,4-TMB by inhalation for 3 months; BMR = 1 SD change from control mean
(constant and modeled variance), (Korsak et al., 2000a)	C-37
Table D-l. The EPA's implementation of the National Research Council's recommendations in the
trimethylbenzenes assessment	D-3
Table D-2. National Research Council recommendations that the EPA is generally implementing in the
long term	D-9
Table E-l. Composition of the C9 fraction test substance used for toxicity testing in Schreiner et al.
(1989), McKee et al. (1990), and Douglas et al. (1993)	E-l
Table E-2. Composition of the C9 fraction test substance used for toxicity testing in Clark et al. (1989)	E-2
This document is a draft for review purposes only and does not constitute Agency policy.
vii	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
FIGURES
Figure B-l. Metabolic scheme for 1,2,4-TMB B-8
Figure B-2. Metabolic scheme for 1,2,3-TMB B-9
Figure B-3. Metabolic scheme for 1,3,5-TMB B-10
Figure B-4. Physiological based toxicokinetic model for 1,2,4-TMB in humans B-13
Figure B-5. Schematic of human model structure for 1,2,4-TMB using the NLE-based model approach B-17
Figure B-6. Schematic of rat and human PBPK model structure B-19
Figure B-7. Simulated and measured blood concentrations of 1,2,4,-TMB in rats exposed to 600, 2,400, or
4,800 mg/m3 WS for 8 hours B-28
Figure B-8. Simulated and measured brain concentrations of 1,2,4-TMB in rats exposed to 600, 2,400, or
4,800 mg/m3 WS for 8 hours B-29
Figure B-9. Simulated and measured exhaled air concentrations of 1,2,4-TMB in three volunteers exposed
to 600 mg/m3 WS for 4 hours B-30
Figure B-10. Comparisons of model predictions to measured blood concentrations in rats exposed to
1,2,4-TMB in WS B-34
Figure B-ll. Comparisons of model predictions to measured brain concentrations in rats exposed to
1,2,4-TMB in WS B-35
Figure B-12. Comparisons of model predictions to measured venous blood concentrations by Swiercz et al.
(2003) in rats repeatedly exposed to 1,2,4-TMB B-35
Figure B-13. Comparisons of model predictions to measured rat venous blood concentrations by Swiercz
et al. (2002) in acutely exposed rats B-37
Figure B-14. Comparisons of model predictions to measured human venous blood concentrations in
Kostrzewki et al. (1997) in human volunteers exposed to 154 mg 1,2,4-TMB/m3 for 8 hours B-44
Figure B-15. Comparisons of model predictions to measured human venous blood concentrations of
Jarnberg et al. (1998, 1997a; 1996) in volunteers exposed to 2 or 25 ppm (~ 10 or 123
mg/m3) 1,2,4-TMB for 2 hours while riding a bicycle (50 W) B-45
Figure B-16. Comparisons of model predictions to measured (a) human venous blood and (b) end of
exposure exhaled air 1,2,4-TMB in human volunteers exposed to 100 ppm WS with 7.8%
1,2,4-TMB (38.4 mg/m31,2,4-TMB) B-46
Figure B-17. Time course of normalized sensitivity coefficients of moderately sensitive chemical-specific
parameters (response: venous blood concentration) in rats exposed to (a) 25 ppm (123
mg/m3) or (b) 250 ppm (1,230 mg/m3) of 1,2,4-TMB via inhalation for 6 hours B-51
Figure B-18. Effect of route of exposure and dose rate on steady-state venous blood concentration (t =
1,200 hr) for continuous human exposure to 1,2,4-TMB B-55
Figure C-l. Plot of mean response by dose for increased latency to paw-lick in male Wistar rats, with the
fitted curve for Exponential model 4 with constant variance C-6
Figure C-2. Plot of mean response by dose for decreased red blood cells in male Wistar rats, with the
fitted curve for Linear model with constant variance C-8
Figure C-3. Plot of mean response by dose for decreased fetal weight in male Sprague-Dawley rats, with
the fitted curve for Linear model with constant variance C-12
This document is a draft for review purposes only and does not constitute Agency policy.
viii DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
Figure C-4. Plot of mean response by dose for decreased fetal weight in male Sprague-Dawley rats, with
the fitted curve for Linear model with constant variance C-12
Figure C-5. Plot of mean response by dose for decreased fetal weight in female Sprague-Dawley rats, with
the fitted curve for Linear model with constant variance C-15
Figure C-6. Plot of mean response by dose for decreased fetal weight in female Sprague-Dawley rats, with
the fitted curve for Linear model with constant variance C-15
Figure C-7. Plot of mean response by dose for decreased maternal body weight gain in female Sprague-
Dawley rats, with fitted curve for Exponential model 3 with constant variance C-17
Figure C-8. Plot of mean response by dose for increased latency to paw-lick in male Wistar rats, with fitted
curve for Linear model with modeled variance C-20
Figure C-9. Plot of mean response by dose for decreased segmented neutrophils in male Wistar rats, with
fitted curve for Exponential model 2 with constant variance C-22
Figure C-10. Plot of mean response by dose for decreased segmented neutrophils in female Wistar rats,
with fitted curve for Hill model with constant variance C-24
Figure C-ll. Plot of mean response by dose for increased reticulocytes in male Wistar rats, with fitted
curve for Linear model with constant variance C-26
Figure C-12. Plot of mean response by dose for decreased maternal body weight gain in female Sprague-
Dawley rats, with fitted curve for Power model with modeled variance C-31
Figure C-13. Plot of mean response by dose for decreased red blood cells in male Wistar rats, with fitted
curve for Hill model with constant variance C-34
This document is a draft for review purposes only and does not constitute Agency policy.
ix DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
ABBREVIATIONS
1,2,3-TMB
1,2,3-trimethylbenzene
QRTOTC
sum of fractional flows to rapidly
1,2,4-TMB
1,2,4-trimethylbenzene

perfused tissues, liver, and brain
1,3,5-TMB
1,3,5-trimethylbenzene
QSTOTC
sum of fractional flows to slowly
AAQC
Ambient air quality criterion

perfused tissues
ABR
amount of 1,2,4-TMB in the brain
RBC
red blood cell
ACGIH
American Conference of
RD50
50% respiratory rate decrease

Governmental Industrial Hygienists
REL
Recommended exposure limit
ADME
Absorption, distribution, metabolism
RfC
reference concentration

and excretion
RfD
reference dose
AEGL
Acute exposure guideline limit
ROS
reactive oxygen species
AIC
Akaike Information Criterion
SD
standard deviation
BAL
bronchoalveolar lavage
SE
standard error
BMD
benchmark dose
TLV
threshold limit value
BMDL
lower confidence limit on the
TMB
trimethylbenzene

benchmark dose
TSCA
Toxic Substances Control Act
BMDS
benchmark dose software
TWA
time-weighted average
BMR
benchmark response
UV
ultraviolet
BW
body weight
VLC
volume of fat
CAS
Chemical Abstracts Service
Vmax
% maximal enzyme rate
CI
confidence interval
voc
volatile organic compound
CMIX
average of arterial and venous blood
w
watt

concentrations
WBC
white blood cell
CNS
central nervous system
WS
white spirit
cv
concentration in venous blood
X2
chi-squared
CVS
concentration in venous blood exiting
slowly perfused tissues


CXEQ
concentration in exhaled breath


DMBA
dimethylbenzoic acid


DMHA
dimethylhippuric acid


EC50
half maximal effective concentration


EPA
U.S. Environmental Protection
Agency


GD
gestational day


HEC
human equivalent concentration


i.p.
intraperitoneal


IRIS
Integrated Risk Information System


K,„
Michaelis-Menten constant


LOAEL
lowest-observed-adverse-effect level


NCEA
National Center for Environmental
Assessment


NIOSH
National Institute for Occupational
Safety and Health


NOAEL
No-observed-adverse-effect level


OMOE
Ontario Ministry of the Environment


P
probability value


PBPK
physiologically based
pharmacokinetic (model)


POD
point of departure


POI
Point of impingement


QPC
alveolar ventilation rate


This document is a draft for review purposes only and does not constitute Agency policy.
x	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
APPENDIX A. HEALTH ASSESSMENTS AND
REGULATORY LIMITS BY OTHER NATIONAL AND
INTERNATIONAL HEALTH AGENCIES
Table A-l. Other national and international health agency assessments for TMBs
Agency
Toxicity value
National Institue for
Occupational Safety and
Health (NIOSH. 1992. 1988)
Recommended Exposure Limit (REL) for TMBs - 25 ppm (123 mg/m3)
time weighted average for up to a 10 hour work day and a 40 hour work
week, based on the risk of skin irritation, central nervous system
depression, and respiratory failure (Battig et al., 1956)
American Conference of
Governmental Industrial
Hvsienists (ACGIH. 2002)
Threshold Limit Value (TLV) for VOC mixture containing 1,2,4-TMB and
1,3,5-TMB - 25 ppm (123 mg/m3) time weighted average for a normal 8-
hour work day and a 40-hour work week, based on the risk of irritation
and central nervous system effects (Battig et al., 1956)
National Advisory Committee
for Acute Exposure Guideline
Levels for Hazardous
Substances (U.S. EPA, 2007)
Acute Exposure Guideline Level (AEGL)-l (nondisabling) - 180 ppm (890
mg/m3) to 45 ppm (220 mg/m3) (10 min to 8 hrs, respectively) (Korsak
and Rvdzvnski, 1996)
AEGL-2 (disabling) - 460 ppm (2,300 mg/m3) to 150 ppm (740 mg/m3)
(10 min to 8 hrs, respectively) (Gage, 1970)
Ontario Ministry of the
Environment (MOE, 2006)
For TMBs: 24 hr Ambient Air Quality Criterion (AAQC) - 0.3 mg/m3 based
on CNS effects; half-hour Point of Impingement (POI) - 0.9 mg/m3 based
on CNS effects (Wiaderna et al., 2002; Gralewicz and Wiaderna, 2001;
Gralewicz et al., 1997b; Korsak and Rvdzvnski, 1996)
This document is a draft for review purposes only and does not constitute Agency policy.
A-l	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
APPENDIX B. INFORMATION IN SUPPORT OF
HAZARD IDENTIFICATION AND DOSE-REPONSE
ANALYSIS
B.l. PHYSICAL AND CHEMICAL PROPERTIES
Table B-l. Physical properties and chemical identity of 1,2,4-TMB, 1,3,5-TMB, and
1,2,3-TMB
Property
1,2,4 TMB
1,3,5 TMB
1,2,3 TMB
CAS Registry Number
95-63-6
108-67-8
526-73-8
Synonym(s)
1,2,4-Trimethylbenzene,
1,3,5-Trimethylbenzene,
1,2,3-Trimethylbenzene,

pseudocumene,
mesitylene,
hemimellitene,

asymmetrical
symmetrical
hemellitol,

trimethylbenzene
trimethylbenzene
pseudocumol
Molecular formula
CgH12
CgH12
CgH12
Molecular weight
120.19
120.19
120.19
Chemical structure
CH3



CH3
H3 C HS
|CH3
Ch3

ch3
ch3
CH3
Melting point, °C
-43.8
-44.8
-25.4
Boiling point,
°C @ 760 mm Hg
168.9
164.7
176.1
Vapor pressure,
mm Hg @ 25°C
2.10
2.48
1.69
Density, g/mL at 20 °C
0.8758
0.8637
0.8944
Flashpoint, °C
44
50
44
Water solubility, mg/L at
25 °C
57
48.2
75.2
Other solubilities
ethanol, benzene,
alcohol, ether, benzene,
ethanol, acetone, benzene,

ethyl ether, acetone,
acetone, oxygenated and
petroleum ether

petroleum ether
aromatic solvents

Henry's law constant,
atm mB/mol
6.16 x 10"3
8.77 x 10"3
4.36 x 10"3
Log K0w
3.78
3.42
3.66
Log Koc
2.73
2.70-3.13
2.80-3.04
Bioconcentration factor
439
234
133-259
Conversion factors
1 ppm = 4.92 mg/m3
1 ppm = 4.92 mg/m3
1 ppm = 4.92 mg/m3

1 mg/m3 = 0.2 ppm
1 mg/m3 = 0.2 ppm
1 mg/m3 = 0.2 ppm
Source: (HSDB, 2011a, b, c; U.S. EPA, 1987)
This document is a draft for review purposes only and does not constitute Agency policy.
B-l	DRAFT—DO NOT CITE OR QUOTE

-------
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
Toxicological Review ofTrimethylbenzene
B.2. TOXICOKINETICS
There has been a significant amount of research conducted on the toxicokinetics of
1,2,4-TMB, 1,2,3-TMB, and 1,3,5-TMB in experimental animals and humans. In vivo studies have
been conducted to evaluate the adsorption, distribution, metabolism and excretion (ADME) of
all isomers following exposure via multiple routes of exposure in rats fSwiercz etal.. 2006:
Tsuiimoto etal.. 2005: Swiercz etal.. 2003: Swiercz etal.. 2002: Tsuiino etal.. 2002: Tsuiimoto
etal.. 2000: Eide and Zahlsen. 1996: Zahlsen etal.. 1990: Huo etal.. 1989: Dahl etal.. 1988:
Mikulski and Wiglusz. 1975) and human volunteers (Tanasik et al.. 2008: Tones etal.. 2006:
Tarnbergetal.. 1997a: Tarnbergetal.. 1997b: Kostrzewskietal.. 1997: Tarnberg etal.. 1996:
Kostrewski and Wiaderna-Brvcht. 1995: Fukava etal.. 1994: Ichiba etal.. 19921. The following
sections provide a summary of the toxicokinetic properties for all three isomers. For complete
details regarding the toxicokinetics of TMB isomers in humans and animals, see Tables B-46-B-
64 in Appendices B.6-B.8.
B.2.1. Absorption
Both humans and rats readily absorb 1,2,4-TMB, 1,2,3-TMB, and 1,3,5-TMB into the
bloodstream following exposure via inhalation. Humans (n = 9-10, Caucasian males) exposed to
25 ppm (123 mg/m3) 1,2,4-TMB or 1,3,5-TMB for 2 hours exhibited similar maximum capillary
blood concentrations (6.5 ± 0.88 and 6.2 ± 1.6 [J.M, respectively [digitized data]), whereas
absorption for 1,2,3-TMB was observed to be higher (7.3 ± 1.0 [J.M [digitized data]) fTarnberg et
al.. 1998.1997a: Tarnberg et al.. 1996). Kostrewski et al. (1997) observed equivalent maximal
capillary blood concentrations in humans (n = 5) exposed to 30.5 ppm (150 mg/m3) 1,2,4-TMB
or 1,3,5-TMB for 8 hours (8.15 ± 1.4 and 6.3 ± 1.0 [J.M, respectively). In the same study, human
volunteers exposed to 100 mg/m3 (20.3 ppm) 1,2,3-TMB had capillary blood concentrations of
4.3 ± 1.1 [iM. In humans (n = 4, 2 male, 2 female) exposed to 25 ppm (123 mg/m3) 1,3,5-TMB for
4 hours, venous blood concentrations were markedly lower (0.85 [iM [no SD reported]), but this
may be related to measurement of 1,3,5-TMB in the venous blood Hones etal.. 20061. 1,3,5-TMB
has a higher blood:fat partition coefficient (230) than 1,2,4-TMB (173) or 1,2,3-TMB (164)
(Tarnberg and Tohanson. 1999) and therefore much of the 1,3,5-TMB absorbed into capillary
blood may preferentially distribute to adipose tissue before entering into the venous blood
supply. Measurements of respiratory uptake of 1,2,4-TMB, 1,2,3-TMB, or 1,3,5-TMB are similar
in humans (n = 10, Caucasian males) (60 ± 3%, 48 ± 3%, and 55 ± 2%, respectively).
This document is a draft for review purposes only and does not constitute Agency policy.
B-2	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Toxicological Review ofTrimethylbenzene
In rats, rapid absorption into the bloodstream was observed in many studies following
single exposures to 1,2,4-TMB, with maximal blood concentrations of 537 ± 100, 221 (no SD
reported), and 64.6 ± 13.6 [J.M observed after exposures to 1,000 ppm (4,920 mg/m3) for 12
hours, 450 ppm (2,214 mg/m3) for 12 hours, and 250 ppm (1,230 mg/m3) for 6 hours (Swiercz
etal.. 2003: Eide and Zahlsen. 1996: Zahlsen et al.. 19901. Zahlsen et al. (1990) observed a
decrease in blood concentrations of 1,2,4-TMB following repeated exposures, which they
attribute to induction of metabolizing enzymes; a similar decrease in 1,2,4-TMB blood
concentrations following repeated exposures was not observed in Swiercz et al. (2003). Using a
4-comparment toxicokinetic model, Yoshida et al. (2010) estimated that a rat exposed to 50
|ig/m31,2,4-TMB for 2 hours would absorb 6.6 |J.g/kg body weight (no SD reported). Using this
same model, the authors estimated that humans exposed to 24 |ig/m31,2,4-TMB for 2 hours
would absorb 0.45 [ig/kg body weight (no SD reported). 1,2,4-TMB, 1,2,3-TMB, and 1,3,5-TMB
have also been observed to be absorbed and distributed via blood circulation following oral and
dermal exposures in rats (Tsuiino etal.. 2002: Huo etal.. 1989). Lastly, calculated blood:air
partition coefficients for 1,2,4-TMB, 1,2,3-TMB, and 1,3,5-TMB (43.0 [40.8-45.2], 66.5 [63.7-
69.3], and 59.1 [56.9-61.3], respectively) were similar in humans (n = 10, 5 male, 5 female),
indicating that the two isomers would partition similarly into the blood flarnberg and lohanson.
19951. Additionally, the blood:air partition coefficients between humans and rats were very
similar for all three isomers: 1,2,4-TMB (43.0 vs. 55.7), 1,2,3-TMB (66.5 vs. 62.6), and 1,3,5-TMB
(59.1 vs. 57.7) (Meulenberg and Vijverberg. 2000). This further indicates patterns of absorption
would be similar across species.
This document is a draft for review purposes only and does not constitute Agency policy.
B-3	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofTrimethylbenzene
B.2.2. Distribution
No information exists regarding the distribution of any isomer in adult humans. However,
experimentally calculated tissue-specific partition coefficients were similar for all three isomers
across a number of organ systems (fat, brain, liver, muscle, and kidney) (Meulenberg and
Vijverberg. 2000). This strongly indicates that 1,2,4-TMB, 1,2,3-TMB, and 1,3,5-TMB can be
expected to partition similarly into these various organ systems. Trimethylbenzenes
(unspecified isomer) have also been detected in cord blood, and therefore can be expected to
partition into the fetal compartment (Cooper etal.. 2001: Dowty etal.. 1976). In rats, 1,2,4-TMB
was observed to distribute widely to all examined organ systems following oral exposure, with
the highest concentrations found in the stomach (509 ± 313 |ig/g) and adipose tissue (200 ± 64
[ig/g) (Huo etal.. 1989). Following inhalation exposures, 1,2,4-TMB and 1,3,5-TMB were
observed to distribute to all tissues examined, with tissue-specific concentrations dependent on
the external exposure concentration (Swiercz etal.. 2006: Swiercz etal.. 2003: Eide and
Zahlsen. 19961.1,2,4-TMB distributed to the adipose tissue to a much higher degree than to the
brain, liver, or kidneys (Eide and Zahlsen. 1996). Venous blood concentrations of 1,2,4-TMB and
1,3,5-TMB and liver concentrations of 1,2,4-TMB were observed to be significantly lower in
repeatedly exposed animals versus animals exposed only once to higher concentrations
(Swiercz etal.. 2006: Swiercz etal.. 2003: Swiercz etal.. 2002). Kidney concentrations of
1,3,5-TMB were observed to be lower in repeatedly exposed animals versus animals exposed
once, but only at the lowest exposure concentration. The authors suggest that lower tissue
concentrations of TMB isomers observed in repeatedly-exposed animals is mostly likely due to
induction of metabolizing enzymes at higher exposure concentrations. This hypothesis is
supported by the observation of P-450 enzyme induction in the livers, kidneys, and lungs of rats
exposed to 1,200 mg/kg/day 1,3,5-TMB for 3 days fPvykko. 19801.
This document is a draft for review purposes only and does not constitute Agency policy.
B-4 DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Toxicological Review ofTrimethylbenzene
1,2,4-TMB was also observed to distribute to individual brain structures, with the
brainstem and hippocampus having the highest concentrations following exposure fSwiercz et
al.. 20031. Zahlsen et al. (1990) also observed decreasing blood, brain, and adipose tissue
concentrations following repeated exposures versus single day exposures in rats exposed to
1,000 ppm (4,920 mg/m3). In the only study to investigate distribution following dermal
exposure, 1,2,4-TMB preferentially distributed to the kidneys (Tsujino etal.. 2002).
Concentrations in the blood, brain, liver, and adipose tissue were similar to one another, but
1,2,4-TMB concentrations only increased in a dose-dependent manner in adipose tissue, and
continued to accumulate in that tissue following the termination of exposure. Similar results
were reported for 1,2,3-TMB and 1,3,5-TMB, but specific data were not presented. Detailed
information regarding the distribution of 1,2,3-TMB in rats following inhalation or oral
exposures is lacking. However, similar tissue-specific partition coefficients for 1,2,3-TMB
compared to 1,2,4-TMB and 1,3,5-TMB were similar across a number of organ systems
(Meulenberg and Viiverberg. 2000). indicating similar patterns of distribution can reasonably
be anticipated.
This document is a draft for review purposes only and does not constitute Agency policy.
B-5	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofTrimethylbenzene
B.2.3. Metabolism
The metabolic profiles for each isomer were qualitatively similar between humans and rats,
although in some cases, quantitative differences were reported. In humans (n = 10, Caucasian
males), all three isomers are observed to be metabolized to benzoic and hippuric acids.
Approximately 22% of inhaled 1,2,4-TMB was collected as hippuric acid metabolites in urine 24
hours after 2 hour exposures to 25 ppm (123 mg/m3) 1,2,4-TMB (larnberg et al.. 1997b). 3,4-
dimethylhippuric acid (DMHA) comprised 82% of the dimethylhippuric acids collected after
exposure to 1,2,4-TMB, indicating that steric factors are important in the oxidation and/or
glycine conjugation of 1,2,4-TMB in humans. Approximately 11% of inhaled 1,2,3-TMB was
collected as hippuric acid metabolites fTarnbergetal.. 1997bl. As with 1,2,4-TMB, steric
influences seem to play an important role in the preferential selection of which metabolites are
formed: 2,3-DMHA comprised 82% of all hippuric acid metabolites collected. Urinary hippuric
acid metabolites for 1,3,5-TMB following the same exposure protocol accounted for only 3% of
inhaled dose. The lower levels of hippuric acids recovered in urine following exposure to 1,3,5-
TMB may be a result of differing pKa values. The DMHA metabolite of 1,3,5-TMB has the highest
pKa value of any DMHA metabolite, indicating it ionizes to a lesser degree in urine. This may
lead to increased reabsorption in the kidney tubules, consequently lowering the total amount of
DMHA metabolite excreted within 24 hours (Tarnbergetal.. 1997b). Greater amounts of urinary
benzoic and hippuric acid metabolites (73%) were observed in humans (n = 5) following
exposure to higher amounts of 1,3,5-TMB (up to 30.5 ppm) for 8 hours (Kostrzewski et al..
1997: Kostrewski and Wiaderna-Brvcht. 19951. Following occupational exposure to 1,2,4-TMB
or 1,3,5-TMB, urinary benzoic acid and hippuric acid metabolites in workers (n = 6-12) were
highly correlated with TMB isomer air concentrations fTones etal.. 2006: Fukava etal.. 1994:
Ichiba etal.. 19921.
This document is a draft for review purposes only and does not constitute Agency policy.
B-6 DRAFT—DO NOT CITE OR QUOTE

-------
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
30
31
32
33
34
35
36
37
38
Toxicological Review ofTrimethylbenzene
Following oral exposures in animals, the quantitative metabolic profiles of the three
isomers appears to differ. Mikulski and Wiglusz (1975) observed that 73% of the administered
dose of 1,3,5-TMB was recovered as glycine (i.e., hippuric acid, 59.1 ± 5.2%), glucuronide (4.9 ±
1.0), or sulfate (9.2 ± 0.8%) conjugates in the urine of rats within 48 hours after exposure.
However, the total amount of metabolites recovered following exposure to 1,2,3-TMB and 1,2,4-
TMB was much less (33.0% and ~37%, respectively). The major terminal metabolites for
1,2,4-TMB and 1,3,5-TMB are dimethylhippuric acids (23.9 ± 2.3% and 59 .1 ± 5.2 % total dose,
respectively). Dimethylhippuric acid metabolites represent a smaller fraction (10.1 ± 1.2 %) of
the metabolites produced following 1,2,3-TMB exposure. When an estimate of the total amount
of metabolite was calculated, differences between isomers remained but were in closer
agreement: 93.7% (1,3,5-TMB), 62.6% (1,2,4-TMB), 56.6% (1,2,3-TMB) (no SD reported). It is
important to note that Mikulski and Wiglusz (1975) did not measure other TMB metabolites,
such as mercapturic acid conjugates, trimethylphenols, or dimethylbenzoic acids. Huo et al.
(1989) reported that total amount of metabolites (phenols, benzyl alcohols, benzoic acids, and
hippuric acids) recovered with 24 hours following exposure to 1,2,4-TMB was 86.4 ± 23% of
administered dose (~100 mg/kg).
Similar profiles in metabolism were observed in rabbits: DMBAs and DMHAs were observed
following oral exposure of rabbits to either 1,2,4-TMB or 1,3,5-TMB (Laham and Potvin. 1989:
Cerfetal.. 19801. Specifically for 1,3,5-TMB, 68.5% of the administered oral dose was recovered
as the DMHA metabolite, with only 9% recovered as the DMBA metabolite. Additionally, a
minor metabolite not observed in rats, 5-methylisophthalic acid was observed following
exposure of rabbits fLaham and Potvin. 19891. Additional terminal metabolites for the three
isomers include: mercapturic acids (~14-19% total dose), phenols (~12% total dose), and
glucuronides and sulphuric acid conjugates (4-9% total dose) for 1,2,4-TMB; mercapturic acids
(~5% total dose), phenols (<1-8% total dose), and glucuronides and sulphuric acid conjugates
(8-15% total dose) for 1,2,3-TMB; and phenols (~4-8% total dose) and glucuronides and
sulphuric acid conjugates (~5-9% total dose) for 1,3,5-TMB fTsuiimoto etal.. 2005: Tsuiimoto
etal.. 2000.1999: Huo etal.. 1989: Wiglusz. 1979: Mikulski and Wiglusz. 19751.
Phenolic metabolites were also observed in rabbits following oral exposures to 1,2,4-TMB
or 1,3,5-TMB, although the amounts recovered were quite small (0.05-0.4 % of total dose)
fBakke and Scheline. 19701. As observed in humans, the influence of steric factors appeared to
play a dominant role in determining the relative proportion of metabolites arising from
oxidation of benzylic carbons: the less sterically hindered 3,4-DMHA comprised 79.5% of the
collected hippuric acid metabolites (Huo etal.. 1989). Steric factors appear to be minimal
regarding oxidation of the aromatic ring itself: the most hindered phenol metabolites of
1,2,4-TMB and 1,2,3-TMB were either formed in equal or greater proportions compared to less
sterically hinder metabolites fHuo etal.. 19 891 fTsuiimoto etal.. 20051. The proposed metabolic
schemes for 1,2,4-TMB, 1,2,3-TMB, and 1,3,5-TMB are shown in Figures B-l, B-2, and B-3.
This document is a draft for review purposes only and does not constitute Agency policy.
B-7	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
2,4,5 -trimethy lphenol	2,3,5 -trimethy lphenol
2,3,6-trimethylphenol
2,4-dimethyl-
benzyl
alcohol
3,4-dimethyl-
benzyl
alcohol
CH2OH
1,2,4-trimethyl
benzene
CHdOH
Q-hOH
2,5-dimethyl-
benzyl alcohol
CH, O
2,4-dimethyl-
benzyl
mercapturic acid
3,4-dimethyl-
benzyl
mercapturic acid
2,4-dimethyl
benzoic acid
2,5 -dimethy lbenzy 1
,OH mercapturic acid
2,5-dimethyl
benzoic acid
ch3 o

3,4-dimethyl
benzoic acid
H
HO O
O
jJL
CH3 2,5-dimethyl-
hippuric acid
2,4-dimethyl-
hippuric acid
3,4-dimethyl-
hippuric acid
H 5
Figure B-l. Metabolic scheme for 1,2,4-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
B-8	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
3,4,5-trimethyl-
phenol
1,2,3-trimethyl-
benzene
2,3,4-trimethyl-
phenol
2,6-dimethyl-
mercapturic acid
OH
2,6-dimethyl-
benzoic acid
CH
CH
CH
¦CH
•CH
¦CH
CH.
HO
'CH
'CH
OH
CH-
-CH.
¦CH-
CH-
CH-
CH-
2,6-dimethyl-
benzyl alcohol
CH,
2,3-dimethyl-
benzyl alcohol
2,3-dimethyl-
mercapturic acid
CH, O
OH
2,6-dimethyl-
hippuric acid
~ OH
2,3-dimethyl-
benzoic acid
2,3-dimethyl-
CH hippuric acid
Figure B-2. Metabolic scheme for 1,2,3-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
B-9	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Toxicological Review ofTrimethylbenzene
CH,
h3c
CH,
1,3,5-triniethyl-
benzene
CH,
CH,
OH
2.4.6-
trimethylphenol
CH
CH
H,C
HJC
3,5-dimethylbenzyl
alcohol
CH,
1
3,5-demethylbenzyl	i	<
mercapturic acid
OH
3,5-dimethylbenzoic
acid
CH
3,5-dimethylhippuric
acid
HN
O^ OH
Figure B-3. Metabolic scheme for 1,3,5-TMB.
B.2.4. Excretion
In humans (n = 10, Caucasian males) at low doses (25 ppm [123 mg/m3]), half-lives of
elimination from the blood of all TMB isomers were split into four distinct phases, with the half-
lives of the first three phases being similar across isomers: 1,2,4-TMB (1.3 ± 0.8 min, 21 ± 5 min,
3.6 ± 1.1 hr), 1,2,3-TMB (1.5 ± 0.9 min, 24 ± 9 min, 4.7 ± 1.6 hr), and 1,3,5-TMB (1.7 ± 0.8 min,
27 ± 5 min, 4.9 ± 1.4 hr) (Tarnberg etal.. 1996). 1,3,5-TMB had a higher total blood clearance
value compared 1,2,4-TMB or 1,2,3-TMB (0.97 ± 0.06 L/hr/kgvs. 0.68 ± 0.13 or 0.63 ± 0.13
L/hr/kg respectively). The half-life of elimination for 1,3,5-TMB in the last and longest phase is
much greater than those for 1,2,4-TMB or 1,2,3-TMB (120 ± 41 hr vs. 87 ± 27 and 78 ± 22 hr,
respectively). Urinary excretion of unchanged parent compound was extremely low (<0.002%)
in humans (n = 6-10, male) for all three isomers fTanasik etal.. 2008: Tarnberg etal.. 1997bl.
The half-life of elimination of hippuric acid metabolites from the urine was also greater for
1,3,5-TMB, compared to 1,2,4-TMB or 1,2,3-TMB (16 hr vs. 3.8-5.8 and 4.8-8.1 hr, respectively)
(Tarnberg etal.. 1997b).
This document is a draft for review purposes only and does not constitute Agency policy.
B-10	DRAFT—DO NOT CITE OR QUOTE

-------
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
Toxicological Review ofTrimethylbenzene
Differences in the values of terminal half-lives may be related to interindividual variation in
a small sample population (n = 8-10) and difficulty measuring slow elimination phases. All
three isomers were eliminated via exhalation: 20-37% of the absorbed dose of 1,2,4-TMB,
1,2,3-TMB, or 1,3,5-TMB was eliminated via exhalation during exposure to 123 mg/m3 (25
ppm) for 2 hours fTarnbergetal.. 19961 and elimination of 1,3,5-TMB via breath was bisphasic
with an initial half-life of 60 minutes, and a terminal half-life of 600 minutes (Tones etal.. 2006).
Following exposure of rats to 25 ppm (123 mg/m3) 1,2,4-TMB or 1,3,5-TMB for 6 hours, the
terminal half-life of elimination of 1,3,5-TMB from the blood (2.7 hours) was shorter than that
for 1,2,4-TMB (3.6 hours) fSwiercz etal.. 2006: Swiercz etal.. 20021. As dose increased, the half-
lives for elimination from blood following single exposures to 1,2,4-TMB (17.3 hours) became
much longer than those for 1,3,5-TMB (4 hours). This same pattern was observed for 4-week
repeated exposures as well.
B.3. PHYSIOLOGICALLY-BASED PHARMACOKINETIC MODELS
B.3.1. Summary of Available PBPK models for 1,2,4-TMB
B.3.1.1. Jarnbergand Johanson (1999)
Jarnberg and Johanson (19991 describe a PBPK model for inhalation of 1,2,4-TMB in
humans. The model is composed of six compartments (lungs, adipose, working muscles, resting
muscles, liver, and rapidly perfused tissues) for the parent compound and one (volume of
distribution) for the metabolite, 3,4-DMHA (see Figure B-4). The lung compartment includes
lung tissue and arterial blood. Excretion of parent compound is assumed to occur solely by
ventilation. As 1,2,4-TMB has a pronounced affinity to adipose tissue, a separate compartment
for fat is incorporated into the model. Remaining non-metabolizing compartments are rapidly
perfused tissues, comprising the brain, kidneys, muscles, and skin.
Because previous experimental data was gathered during exercise fTarnbergetal.. 1997a:
Tarnberg etal.. 19961. the muscle compartment was divided into two equally large
compartments, resting and working muscles. Two elimination pathways (a saturable Michaelis-
Menten pathway for all metabolites other than 2,4-DMHA [pathway I] and a first order pathway
[pathway II] for formation of 3,4-DMHA) from the hepatic compartment were included.
Metabolism was assumed to occur only in the liver compartment Tissue:blood partition
coefficients of 1,2,4-TMB were calculated from experimentally determined blood:air, water:air,
and olive oil:air partition coefficients (Tarnberg and Tohanson. 19951 (Table B-2).
This document is a draft for review purposes only and does not constitute Agency policy.
B-ll DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
Toxicological Review ofTrimethylbenzene
The model was used to investigate how various factors (work load, exposure level,
fluctuating exposure) influence potential biomarkers of exposure (end-of-shift and prior-to-
shift concentrations of parent compound in blood and 3,4-DMHA in urine). Biomarker levels
estimated at end-of-shift remained fairly constant during the week, whereas biomarker levels
prior-to-shift gradually increase throughout the week. This indicates end-of-shift values
represent the same day's exposures, whereas prior-to-shift values reflect cumulative exposure
during the entire work week. Increased work load increased uptake of 1,2,4-TMB. For example,
a work load of 150 W over an exposure period of 8 hours increased the level of 1,2,4-TMB in the
blood more than 2-fold, compared to levels of 1,2,4-TMB in the blood after an 8 hour exposure
at rest Simulated 8-hour exposures at air levels 0 to 100 ppm (0 to 492 mg/m3) shows that
overall metabolism is saturable, and that the metabolic pathway yielding 3,4-dimethylbenzene
becomes more important as exposure concentrations increase.
This document is a draft for review purposes only and does not constitute Agency policy.
B-12	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
alv
alv
art
rap
cu
/K,
other
metabolites
I 3,4-dimethyl |
I hippuric acid |
Rapidly perfused
tissues
Working muscles
Adipose tissue
Resting muscles
Lungs and arterial
blood
Liver
urine	urine
Legend: C: concentration of 1,2,4-TMB; Cair: concentration in ambient air; Cart: concentration in arterial blood; Cven:
concentration in venous blood; Qaiv: alveolar ventilation; Qc0\ cardiac output; Q,: blood flow to compartment i (where i = rap =
rapidly perfused tissues; f = adipose tissue; w = working muscles, r = resting muscles, h = liver); Vmax: maximum rate of
metabolism, pathway I; Km: Michaelis-Menten constant for metabolic pathway I; CL1: intrinsic hepatic clearance of metabolic
pathway II; ke: excretion rate constant of 3,4-DMHA. Adapted from Jarnberg and Johanson (1999).
Figure B-4. Physiological based toxicokinetic model for 1,2,4-TMB in humans.
This document is a draft for review purposes only and does not constitute Agency policy.
B-13	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
Table B-2. Measured and calculated partition coefficients for TMB isomers at 37°C
Substance
Measured values3
Calculated values
P Saline:Air
n = 42
P Oil:Air
n = 25
Human PBiood:Air
n = 39
Human P Blood:Air"
1,3,5-TMB
1.23 (1.11-1.35)
9,880 (9,620-10,140)
43.0 (40.8-45.2)
60.3
1,2,4-TMB
1.61(1.47-1.75)
10,200 (9,900-10,400)
59.1 (56.9-61.3)
62.2
1,2,3-TMB
2.73 (2.54-2.92)
10,900 (10,500-11,300)
66.5 (63.7-69.3)
67.5
aMean values and 95%CI.
Calculated as (0.79 x P Saiine:Air) + (0.006 x P 0ii:Air); where 0.79 is the relative content of saline in blood and 0.006 is
the relative content of fat in blood (Fiserova-Bergerova, 1983).
Adapted from Jarnberg and Johanson (1995).
1	Previously performed experimental human exposures to 1,2,4-TMB were used to estimate
2	the metabolic parameters and alveolar ventilation flarnberg et al.. 1997a: larnbergetal.. 19961.
3	Individual simulated arterial blood concentrations and exhalation rates of 1,2,4-TMB, as well as
4	the urinary excretion rate of 3,4-DMHA, were simultaneously adjusted to the experimentally
5	obtained values by varying the alveolar ventilation at rest. One individual's compound-specific
6	and physiological parameters were then used for subsequent model predictions (Table B-3).
This document is a draft for review purposes only and does not constitute Agency policy.
B-14	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
Table B-3. PBPK model parameters for 1,2,4-TMB toxicokinetics in humans using the
Jarnberg and Johanson (1999) model structure
Parameters
Rest
Both3
50 W
Body height (m)

1.78

Body weight (kg)

75.5

Vmax(nmol/rnin)

3.49

Km (HM)

4.35

CL1 (L/min)

0.149

Elimination rate constant (min1)

0.0079

Alveolar ventilation (L/min)
9.05

20.2
Compartment volumes (L)
Lungs and arterial blood

1.37

Liver

1.51

Fat

25.0

Brain and kidneys

1.49

Working muscles

16.6

Resting muscles

16.6

Blood flows (L/min)
Cardiac output
5.17

9.16
Liver
1.67


Fat
0.55


Brain and kidneys
1.86

1.78
Working muscles
0.55

4.3
Resting muscles
0.55

0.55
Partition coefficients
Blood:air

59

Fat:blood

125

Liver:blood

5

Rapidly perfused tissues:blood

5

Muscle:blood

5

Parameters used for both working and resting conditions.
Adapted from Jarnberg and Johanson (1999).
This document is a draft for review purposes only and does not constitute Agency policy.
B-15	DRAFT—DO NOT CITE OR QUOTE

-------
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
30
Toxicological Review ofTrimethylbenzene
B.3.1.2. Emond and Krishnan (2006)
The Emond and Krishnan (2006) model was not developed specifically for 1,2,4-TMB, but
rather to test a modeling concept The PBPK model developed was to test the hypothesis that a
model could be developed for highly lipophilic volatile organic chemicals (HLVOCs) using the
neutral lipid-equivalent (NLE) content of tissues and blood as the basis. This NLE-based
modeling approach was tested by simulating uptake and distribution kinetics in humans for
several chemicals including a-pinene, d-limonene, and 1,2,4-TMB. The focus of this model
review is to use of the model for the prediction of 1,2,4-TMB kinetics and distribution.
This model consisted of five compartments (see Figure B-5) with systemic circulation,
where the tissue volumes corresponded to the volumes of the neutral lipids (i.e., their neutral
lipid-equivalents), rather than actual tissue volume as more commonly found. NLE is the sum of
the neutral (nonpolar) lipids and 30% of the tissue phospholipid (fraction of phospholipids
with solubility similar to neutral lipids) content The model describes inhalation of 1,2,4-TMB
using a lumped lung/arterial blood compartment. Clearance of 1,2,4-TMB is described in the
model with exhalation, but more significantly through first order hepatic metabolism. First-
order metabolism is appropriate in the low dose region (<100 ppm [< 492 mg/m3]), where
metabolism is not expected to be saturated.
In the study description, the mixed lung/arterial blood compartment is not a standard
structure for the lung/blood/air interface. The concentration in lung tissue is assumed equal to
alveolar blood, and the exhaled air concentration is equal to the lung/blood concentration
divided by the blood air partition coefficient This approach is appropriate, and appears to be
accurately represented mathematically by the authors.
Physiological parameters appear to be within ranges normally reported. The calculation of
the NLE fraction is clearly explained and values used in the calculations are clear and
transparent Other model parameters (e.g., alveolar ventilation, cardiac output, blood flows, and
volumes of compartments) were taken from Jarnberg and Johanson (1999) and converted to
the approximate NLE. Hepatic clearance rates were taken from literature on in vivo human
clearance calculations and then expressed in terms of NLE. The NLE-based model was able to
adequately predict human blood concentrations of 1,2,4-TMB following inhalation of 2 or 2 5
ppm (9.8 or 123 mg/m3) for 2 hours without alteration to model parameters obtained from
literature.
This document is a draft for review purposes only and does not constitute Agency policy.
B-16 DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene


It




Lungs



1
1









Adipose tissue



V


(Lipid Fraction)


A
I.



R
N
O
L

L
T


Richly perfused tissues


b
u
P

(Lipid Fraction)


R
s
d

d
1
A
B
F

Resting muscles

F
L
L
r

(Lipid Fraction)

r

O
a

a
R
O
D
t

t
L
0
i

Working muscles



n

(Lipid Fraction)

n
U


D



Hepatic tissue






(Lipid Fraction)





1 ~ Metabolism


Note: Arrows represent blood flows, gas exchange, and metabolism as indicated. Source: Emond and Krishnan (2006).
Figure B-5. Schematic of human model structure for 1,2,4-TMB using the NLE-based
model approach.
1	The PBPK model developed by Emond and Krishnan (2006) is used to test the hypothesis
2	that a model could be developed for HLVOCs using the NLE content of tissues and blood as the
3	basis. To test this NLE-based approach, the uptake and distribution kinetics in humans for
4	several chemicals including 1,2,4-TMB were simulated. The model appeared to accurately
5	reflect experimental data; however, a rodent model is needed for this assessment for animal-
6	to-human extrapolation and no known rodent NLE model for 1,2,4-TMB is available.
This document is a draft for review purposes only and does not constitute Agency policy.
B-17	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Toxicological Review ofTrimethylbenzene
B.3.1.3. Hissinketal. f20071
This model was developed to characterize internal exposure following white spirit (WS)
inhalation. Since WS is a complex mixture of hydrocarbons, including straight and branched
paraffins, two marker compounds were used including 1,2,4-TMB and n-decane. The rat models
were developed to predict the levels of 1,2,4-TMB and n-decane in blood and brain, then the rat
model was scaled allometrically to obtain estimates for human blood following inhalation.
Toxicokinetic data on blood and brain concentrations in rats of two marker compounds,
1,2,4-TMB and n-decane, together with in vitro partition coefficients were used to develop the
model. The models were used to estimate an air concentration that would produce human brain
concentrations similar to those in rats at the no-observed-effect-level (NOEL) for central
nervous system (CNS) effects.
This is a conventional five compartment PBPK model for 1,2,4-TMB similar to previously
published models for inhaled solvents. The five compartments were: liver, fat, slowly perfused
tissues, rapidly perfused tissues, and brain (see Figure B-6).
This document is a draft for review purposes only and does not constitute Agency policy.
B-18	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
V
E
N
O
u
s
B
L
o
o
D
Gas Exchange
Slowly perfused
Qr
Richly perfused
Q,
A
R
T
E
R
I
A
L
B
L
O
O
D
Liver
(metabolism)
Vm„ & K
Note: Boxes represent tissue compartments, while solid arrows represent blood flows, gas exchange, and metabolism as
indicated. Source: Hissink et al. (2007).
Figure B-6. Schematic of rat and human PBPK model structure.
This document is a draft for review purposes only and does not constitute Agency policy.
B-19	DRAFT—DO NOT CITE OR QUOTE

-------
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
Toxicological Review ofTrimethylbenzene
All compartments are described as well mixed/perfusion limited. A lung compartment is
used to describe gas exchange. The liver was the primary metabolizing organ where 1,2,4-TMB
metabolism was described as saturable using Michaelis-Menten kinetics. Since the brain is the
target organ for CNS effects due to exposure to hydrocarbon solvents, it was included as a
separate compartment For the rat, the authors reported that Km and Vmax values were obtained
by fitting predicted elimination time courses to observed blood concentration profiles at three
different exposure levels (obtained from the rat exposure portion of the study). For the human
model, rat Vmax data was scaled to human body weight (BW0 74) and Km values were used
unchanged.
The model appears to effectively predict blood concentrations in rats and humans and in
the brains of rats following inhalation of WS. Changes to the rat model parameters to fit the
human data were as expected. The model is simple and includes tissues of interest for potential
dose metrics.
In rats, the model-predicted blood and brain concentrations of 1,2,4-TMB were in
concordance with the experimentally derived concentrations. In humans, experimental blood
concentrations of 1,2,4-TMB were well predicted by the model, but the predicted rate of
decrease in air concentration between 4-12 hours was lower compared to measured values.
The authors did not provide information on how model predictions compared to data from
animals or humans exposed to pure 1,2,4-TMB. Based on good model fits of experimental data,
the model was valid for the purpose of interspecies extrapolation of blood and brain
concentrations of 1,2,4-TMB as a component of WS.
B.3.2. 1,2,4-TMB PBPK Model Selection
All available 1,2,4-TMB PBPK models were evaluated for potential use in this assessment. Of
the three deterministic PBPK models available for 1,2,4-TMB fHissink etal.. 2007: Emond and
Krishnan. 2006: Tarnberg and lohanson. 1999). the Hissink et al. (2007) model was chosen to
utilize in this assessment because it was the only published 1,2,4-TMB model that included
parameterization for both rats and humans, the model code was available, and the model
adequately predicted experimental data in the dose range of concern. The Hissink et al. (2007)
model was thoroughly evaluated, including a detailed computer code analysis (details follow in
Section B.3.3).
This document is a draft for review purposes only and does not constitute Agency policy.
B-20 DRAFT—DO NOT CITE OR QUOTE

-------
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
Toxicological Review ofTrimethylbenzene
B.3.3. Details of Hissink etal. (2007) Model Analysis
B.3.3.1. Review and Verification of the Hissink et al. (2007) 1,2,4-TMB PBPK Model
Verification of accuracy of the model code
In general, the model code and the description of the model in Hissink et al. (2007) were in
agreement. The one significant discrepancy was that the model code contained an element that
changed the metabolism rate (Vmax) during exposure in a manner that was not documented in
the paper. This additional piece of model code, when used in 8 hour rat simulations with a body
weight of 0.2095 kg, resulted in Vmax holding at 1.17 from the beginning of exposure to t = 1 hr,
then increasing linearly to 1.87 by the end of the exposure and to 2.67 by the end of the post
exposure monitoring period (t = 16 hrs, 8 hrs after the end of exposure). The published rat
simulations, however, did not appear to be entirely consistent with the inclusion of these Vmax
adjustments, raising questions as to whether the code that was verified was the code that was
actually used in the final analyses done for the published simulations. The impact of this
deviation from the published Vmax value is described below in regards to the verification of the
Hissink et al. (2007) model.
Other minor issues were identified by examining the code and comparing it to the model
documentation in Hissink et al. (2007). The code contained some elements that were not
necessary (e.g., i.v. dosing, repeated exposure, interruptions in daily exposure), but since these
do not hinder proper functioning of the model, these elements were not removed or modified.
The mass balance equation omitted one term, the amount of 1,2,4-TMB in the brain (ABR); this
term has been added. The coding for the blood flow was not set up so as to ensure flow/mass
balance. That is, values of sum of fractional flows to rapidly perfused tissues, liver, and brain
(QRTOTC) and sum of fractional flows to slowly perfused tissues (QSTOTC) were selected such
that their sum equals one, but if one value were to be changed, the model code would not
automatically compensate by changing the other. Therefore, the code was modified so that
QSTOTC = 1 - QRTOTC, to facilitate future sensitivity analyses.
Human exhaled breath concentrations were compared to CXEQ (= CV/PB based on the
model code and consistent with the description of the experiment), which would be equivalent
to the end-exhaled alveolar air after breath holding, but the method used to calculate CXEQ was
not noted in Hissink et al. (2007). This is important because there can be different definitions of
exhaled breath depending on the measurement technique. For example, mixed exhaled breath
is typically calculated as 70% alveolar air and 30% "inhaled" concentration, due to dead space.
This document is a draft for review purposes only and does not constitute Agency policy.
B-21	DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Toxicological Review ofTrimethylbenzene
Comparisons between the computer .m files and published descriptions fHissink etal..
20071 indicated minor discrepancies and uncertainties in exposure concentrations and body
weight. Exposure concentrations in the simulations were set at the nominal exposure levels,
rather than analytically determined levels. The maximum deviation between the nominal level
and analytically determined levels occurred in the rat high exposure group, with a nominal
exposure of 4,800 mg/m3 WS (7.8% [38.4 mg/m3] 1,2,4-TMB) and mean analytical
concentrations ranging from 4,440 to 4,769 mg/m3—as much as 9.2% lower. Rat body weights
at time of exposure were reported as 242 to 296 g fHissink et al.. 20071. but the .m files use
values of 210.01, 204.88, and 209.88 g in the low-, mid-, and high-exposure groups,
respectively. Human volunteer body weights reportedly ranged from 69 to 82 kg, and the text
states that the fitted Vmax and Km were obtained for a 70 kg male fHissink et al.. 20071. but a
body weight of 74.9 kg was used in the .m file. No changes to these parameters were made in
the model code, based on the assumption that additional data were available to the model
authors.
Measured human blood concentrations were compared to the average of arterial and
venous blood concentrations (CMIX), while the protocol states that blood was taken from the
cubital vein, so a more appropriate measure may have been venous blood exiting the slowly
perfused tissues compartment (CVS). This choice of dose metric is unlikely to have contributed
significantly to any errors in parameterizing the model (i.e., estimating best-fit metabolism
parameters) because the difference between the two values is generally small. Revised model
code and modeling results are provided on EPA's Health Effects Research Online (HERO)
database fU.S. EPA. 2011al.
This document is a draft for review purposes only and does not constitute Agency policy.
B-22 DRAFT—DO NOT CITE OR QUOTE

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Toxicological Review ofTrimethylbenzene
Verification of model parameter plausibility
Anatomical and physiological parameters
The anatomical physiological parameters used by Hissink et al. (2007) were taken from U.S.
EPA (19881. but more current convention is to use the parameters in Brown et al. (19971.
Comparisons of the rat anatomical and physiological parameters in these sources are found in
Table B-4. Many disagreements in values were identified, particularly with respect to the blood
flows. In interpreting the blood flow percentages, it should be noted that the percentages
enumerated by Brown et al. (19971 do not sum to 100%, which is of course a physiological
requirement. Perfusion rates of various depots of fat may differ, so the single value or fractional
blood flow to fat given by Brown et al. (1997) of 7%, may be deemed sufficiently uncertain that
the Hissink et al. (2007) value of 9% is considered acceptable. Brown et al. (1997) report
substantially higher blood flow percentages to slowly perfused tissues (skin: 5.8% and muscle:
27.8%, for a total of 33.6%) than the value of 15% used by Hissink et al. (2007). The difference
cannot be due to a smaller set of tissues being "lumped" into this compartment, because Hissink
et al. (2007) assign a larger volume fraction of tissue to this compartment Hissink et al. (2007)
also assign a higher percentage of blood flow to the liver than indicated by Brown et al. (1997).
Because no sensitivity analyses were conducted by the authors, it is unclear what impact these
discrepancies may have had on the predicted 1,2,4-TMB kinetics and visual optimization of
metabolism parameters.
Comparisons of the human anatomical and physiological parameters in Hissink et al. (2007)
and Brown et al. (1997) are found in Table B-5. In general, the agreement was better for
humans than it was for rats. Brown et al. (1997) propose a higher default body fat percentage
than was used by Hissink et al. (2007), but Hissink et al. (2007) used values derived from
measurements of the volunteers participating in the study. Because these volunteers had
relatively low percentages of body fat, it is appropriate that the volume of slowly perfused
tissue (including muscle) should be increased to compensate.
This document is a draft for review purposes only and does not constitute Agency policy.
B-23 DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
Table B-4. Comparison of rat anatomical and physiological parameters in Hissink
etal. (2007) to those of Brown etal. (1997)
Parameter
Hissink et al. (2007)a
Range from Brown
et al. (1997)
Values in agreement?
Alveolar ventilation rate (L/hr/kg0'7)
20
12-54b
Yes
Total cardiac output (L/hr/kg07)
20
9.6-15
No
Blood flow (% cardiac output)



Liver (total)
25
13.1-22.1
No
Fat
9
7
Acceptable0
Brain
1.2
1.5-2.6
No
Rapidly perfused (total)
49.8
15.3-27.4
No
Adrenals

0.2-0.3

Heart

4.5-5.1

Kidneys

9.5-19

Lung

1.1-3

Slowly perfused (total)
15
33.6
No
Muscle

27.8

Skin

5.8

Total
100
70.5-92.7

Tissue volume (% body weight)



Liver
4
2.14-5.16
Yes
Fat
7
3.3-20.4
Yes
Brain
0.72
0.38-0.83
Yes
Rapidly perfused
4.28
3.702-6.11
Yes
Adrenals

0.01-0.31

Stomach

0.4-0.6

Small intestine

0.99-1.93

Large intestine

0.8-0.89

Heart

0.27-0.4

Kidneys

0.49-0.91

Lungs

0.37-0.61

Pancreas

0.24-0.39

Spleen

0.13-0.34

Thyroid

0.002-0.009

Slowly perfused
75
51.16-69.1
Acceptable0
Muscle

35.36-45.5

Skin

15.8-23.6

Total
91
60.682-101.6

"Values from U.S. EPA (1988).
bAssuming a standard 250 g rat.
cHissink et al. (2007) value outside of literature range, but acceptable (see discussion in text).
Data source: Hissink et al. (2007) and Brown et al. (1997).
This document is a draft for review purposes only and does not constitute Agency policy.
B-24	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
Table B-5. Comparison of human anatomical and physiological parameters in
Hissink et al. (2007) to those of Williams and Leggett (1989) as reported
by Brown et al. (1997)
Parameter
Hissink et al.
(2007)a
Range from Brown
et al. (1997)
Values in
agreement?
Alveolar ventilation rate (L/hr/kg0'7)
20
15
Acceptable
Total cardiac output (L/hr/kg07)
20
16
Acceptable
Blood flow (% cardiac output)
Liver (total)
26
11-34.2
Yes
Fat
5
3.7-11.8
Yes
Brain
14
8.6-20.4
Yes
Rapidly/Richly perfused (total)
30
19.9-35.9
Yes
Adrenals

0.3

Heart

3-8

Kidneys

12.2-22.9

Lung

2.5

Thyroid

1.9-2.2

Slowly perfused (total)
25
9-50.8
Yes
Muscle

5.7-42.2

Skin

3.3-8.6

Total
100
52.2-153.1

Tissue Volume (% body weight)
Liver
2.6
2.57
Yes
Fat
14.6
21.42
Acceptable
(measured)3
Brain
2
2
Yes
Rapidly/Richly perfused
3
3.77
Acceptable
Adrenals

0.02

Stomach

0.21

Small intestine

0.91

Large intestine

0.53

Heart

0.47

Kidneys

0.44

Lungs

0.76

Pancreas

0.14

Spleen

0.26

Thyroid

0.03

Slowly perfused
66.4
43.71
Acceptable
Muscle

40

Skin

3.71

Total
88.6
73.47

aThe Hissink et al. (2007) value differs from Brown et al. (1997), but is acceptable (see discussion in text).
Data source: Hissink et al. (2007); and Williams and Leggett (1989) [as reported by Brown et al. (1997)].
This document is a draft for review purposes only and does not constitute Agency policy.
B-25	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
Chemical-specific parameters
1 The chemical-specific model parameters, the partition coefficients, and the metabolic
2 parameters are summarized in Table B-6.
Table B-6. Comparison of chemical-specific parameters in Hissink et al. (2007) to
literature data
Parameter
Hissink et al. (2007)
Literature
Values in
agreement?

Value
Technique
Value
Technique

Partition coefficients
Saline:Air
3
In vitro
1.47-1.75a
In vitro
Acceptable
Olive oil:Air
13,200
In vitro
9,900-
10,400a
In vitro
Acceptable
Blood:Air - human
85
In vitro
59.6-61.3a
In vitro
Acceptable
Blood:Air - rat
148
In vitro
-

Rapidly perfused:Blood
2.53
Calculated
-

Slowly perfused:Blood
1.21
Calculated
-

Fat:Blood
62.7
Calculated
63b
In vivo
Yes
Brain:Blood
2.53
Calculated
2b
In vivo
Acceptable
Liver: Blood
2.53
Calculated
-

Metabolism
VmaxC - rat (mg/hr/kg0'7)
3.5
Visual
optimization
-

VmaxC- human (mg/hr/kg0 7)
3.5
Assumed
equal to rat
1.2-21°
Optimization
Yes
Km - rat (mg/L)
0.25
Visual
optimization
-

Km - human (mg/L)
0.25
Assumed
equal to rat
0.42-4.0°
Optimization
No
VmaxC/Km - human (L/hr/kg07)
14
Assumed
equal to rat
2.6-15°
Optimization
Yes
aJarnberg and Johanson (1995).
bZahlsen eta I. (1990).
cJarnberg and Johanson (1999).
Source: Hissink et al. (2007)
3 Where data were available, the agreement is generally acceptable. While the rat-derived Km
4 is less than the lower 95% confidence interval value for the human Km, the human VmaxC/Km
5 ratio is in acceptable agreement. When considering sufficiently low exposure concentrations,
6 the performance of the Hissink et al. (2007) human model metabolism parameters would be
7 consistent with the Jarnberg and Johanson (1999) value.
This document is a draft for review purposes only and does not constitute Agency policy.
B-26 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 ofTrimethylbenzene
Verification that the model can reproduce all figures and tables in the publication
by Hissink et al. (2007)
The experimental data in Hissink et al. (2007) were estimated by use of Plot Digitizer
(version 2.4.1) to convert the symbols on the relevant figures into numerical estimates. The
model code provided (adapted for acslX), with a variable value for Vmax, does not appear to
perfectly reproduce the rat simulations in Hissink et al. (2007) (Figures B-7a and b and B-8a
and b) (please note that the Hissink et al. (2007) figures have been "stretched" to produce
approximately the same x-axis scale found in the acslX figures). It appears to yield end-of
exposure blood and brain concentrations that are about the same as in the Hissink et al. (2007)
simulations, but the post-exposure clearance appears faster in EPA's calculations (see, for
example, the 16 hr time points for the high exposures). When the simulations were run with
Vmax constant (Figures B-7c and B-8c), as documented in Hissink et al. (2007). the rat
simulations yield higher blood and tissue concentrations than depicted in Hissink et al. (2007),
most notably at the high exposure concentration. Similar results were obtained for the rat brain
concentrations (Figure B-8). The human simulations of blood and exhaled air appear to be
faithfully reproduced by the model (Figure B-9). The predicted brain concentration for humans
exposed to 600 mg/m3 WS (45 mg/m31,2,4-TMB) for 4 hours was reported as 721 ng/g (0.721
mg/L) in Hissink et al. (2007). whereas the current simulation predicts a concentration of 0.818
mg/L.
This document is a draft for review purposes only and does not constitute Agency policy.
B-27	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review ofTrimethylbenzene
10000
? 1000
¦o
O
o
i 100
10
0	4	8	12	16
(a)	Time (h>
White spirit exposure of rats to 0.05, 0.19, and 0.37 mg/L 1,2,4-TMB (Hissink et al., 2007)
White spirit exposure of rats to 0,05,0.19, and 0.37 mg/L 1,2,4-TMB (Hissink et al., 2007)
(a) Hissink et al. (2007). Figure 2, lower panel (b) variable Vmax, (c) constant Vm3X.
@— linel
[g— line 2
@ — line 3
@ RATT,RATCVLOW
@ • RATT,RATCVMID
@ | RATT, RAT C VHI
@ —
linel
ffl —
line 2
0 —
line 3
0
RATT.RATCVLOW
V t
RATT,RATCVMID
V ¦
RATT, RATC VHI
Figure B-7. Simulated and measured blood concentrations of 1,2,4,-TMB in rats
exposed to 600, 2,400, or 4,800 mg/m3 WS for 8 hours.
This document is a draft for review purposes only and does not constitute Agency policy,
B-28	DRAFT—DO NOT CITE OR QUOTE

-------
Toxicological Review of Trimethylbenzene
0	2	4	6	8	10	12	14	16	18	20	22	24
Time (hr)
(b)
White spirit exposure of rats at 0.05, 0.19 and 0.37 mg/L 1,2,4-TMB (Hissink etaL, 2007)
12
"Time (hr)
(a) Hissink et al. (2007). Figure 3, lower panel, (b) variable Vmax (c) constant Vmax.
100000 F
4	8	12	16	20
Time (h)
White spirit exposure of rats at 0.05,0.19 and 0.37 mg/L 1,2,4-TMB (Hissink etaL, 2007)
10000
1000
0	
linel
SI—
line 2
Si —
t, cbr
g A
RATT, RATC B RLOW
~ •
RATT,RATCBRMID
' ¦
RATT,RATCBRHI
0	
linel
0	
line 2
0	
t, cbr
0 A
RATT, RATC B RLOW
-------
Toxicological Review of Trimethylbenzene
v — t, cmix
V ~ Iine2
0 — t, cxeq
y ~ Ine2
'J-.O.Ol
White spirit exposure of humans at45 mg/L (100 ppm)
8	12	16	20
Time (h)
White spirit exposure of humans at 45 mg/L (100 ppm) (Hissink et al., 2007)
1000
(a) Hissink et al. (2007), Figure 4; (b) model simulation during exposure; and (c) model simulation after exposure.
Figure B-9. Simulated and measured exhaled air concentrations of 1,2,4-TMB in three
volunteers exposed to 600 ing/in3 WS for 4 hours.
This document is a draft for review purposes only and does not constitute Agency policy,
B-30	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Toxicological Review ofTrimethylbenzene
B.3.3.2. PBPK Model Optimization and Validation
Methods and Background
For all optimizations, the Nelder-Mead algorithm was used to maximize the log-likelihood
function (LLF). A constant heteroscedasticity value of 2 (i.e., relative error model) was assumed.
Statistical significance of an increase in the LLF was evaluated for 95% confidence per Collins et
al. (19991. All kinetic studies were conducted with adult animals or adult human volunteers. In
many cases, blood and tissue concentration data in a numerical form were available from the
literature fSwiercz etal.. 2003: Swiercz etal.. 2002: Kostrzewski et al.. 1997: Eide and Zahlsen.
1996: Zahlsen etal.. 1992: Dahl etal.. 1988). The 1,2,4-TMB blood, brain, and exhaled breath
concentration data in Hissink et al. (2007) were published in graphical format and a colleague
of Dr. Hissink also provided these in numerical form to Dr. Lisa Sweeney for use in this analysis.
Average estimates of the blood concentrations of 1,2,4-TMB (average and standard
deviation) in humans exposed only to 1,2,4-TMB as presented in graphs in Jarnberg et al. (1998.
1997a: 1996) were used in this evaluation. Estimates of the blood and tissue 1,2,4-TMB
concentrations in rats presented in graphs in Zahlsen et al. (1990) were also used in this
evaluation. Prior to model optimization, physiological parameters were modified from those in
Hissink et al. (2007) to better reflect a more recent literature compilation fBrown etal.. 19971
than the references cited by Hissink et al. (2007) (Table B-7). Where possible, study specific
body weights and measured concentrations (rather than nominal concentrations) have been
used, as detailed in the .m files (U.S. EPA. 2011a). For the Zahlsen et al. (1990) 14-day study,
body weights for exposures after the first exposure were estimated based on European growth
curves for male Sprague-Dawley rats (linear regression of weights for weeks 6-9) (Harlan
Laboratories. 2012).
This document is a draft for review purposes only and does not constitute Agency policy.
B-31	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-7. Parameter values for the rat and human PBPK models for 1,2,4 TMB used
by EPA
Parameter
RAT
HUMAN (AT REST)
Body weight (kg)
0.230-0.390a
70
Alveolar ventilation rate (L/hr/kg0'70)
14
15
Total cardiac output (L/hr/kg070)
14
16
Blood flow (% of total cardiac output)
Liver
17.6
17.5
Fat
9
8.5
Brain
2.0
11.4
Rapidly perfused
37.8
37.7
Slowly perfused
33.6
24.9
Volume (% of body weight)
Liver
4
2.6
Fat
7
21.42
Brain
0.57
2
Rapidly perfused
4.43
3
Slowly perfused
75
59.58
Partition coefficients (dimensionless)
Blood: air
148
85
Rapidly perfused: blood
2.53
4.4
Slowly perfused: blood
1.21
2.11
Fat: blood
62.7
109
Brain: blood
2.53
4.4
Liver: blood
2.53
4.4
Liver metabolism
VmaxC (mg/hr/kg0'70)
4.17
Km(mg/L)
0.322
aStudy specific.
Source: (U.S. EPA. 2011a).
This document is a draft for review purposes only and does not constitute Agency policy.
B-32	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Rat Model Optimization
1	The rat studies considered in model optimization and model testing (validation) are
2	summarized in Table B-8.
Table B-8. Rat 1,2,4-TMB kinetic studies used in model development and testing
Reference
Strain
Gender
Nominal
concentration
Exposure
regimen
1,2,4-TMB
measurement
Use in model
evaluation
Form of
comparison
Hissink et al.
(2007)
WAG/RijC
R/BR
(Wistar
derived)
Male
102, 410, 820
ppm WS (7.8%
1,2,4-TMB [39.1,
157.3, 314.7
mg/m3])
8 hr
Mixed blood
time course
Optimization
(1,2,4-TMB in
mixture)
Figure B-10
Brain time
course
Testing
Figure B-ll
Swiercz et
al. (2003)
Wistar
Male
25, 100, 250
(123, 492, 1,230
mg/m3)
6 hr/day,
5
days/week
4 weeks
Venous blood
time course
Optimization
(1,2,4-TMB
only)
Figure B-12
Arterial blood,
liver, brain
Testing
Table B-9
6 hr
Arterial blood,
liver, brain
Testing
Table B-9
Swiercz et
al. (2002)
Wistar
Male
25, 100, 250
(123, 492, 1,230
mg/m3)
6 hr
Venous blood
time course
Testing
Figure B-13
Zahlsen et
al. (1990)
Sprague-
Dawley
Male
1,000
(4,920 mg/m3)
12 hr/day
14 days
Blood, brain,
perirenal fat on
days 1, 3, 7,10,
and 14
Testing
Table B-12
Zahlsen et
al. (1992)
Sprague-
Dawley
Male
100
492 mg/m3)
12 hr/day
3 days
Blood, brain,
liver, kidney,
and perirenal
fat at end of
exposures and
after 12 hr
recovery
Testing
Table B-10
Elde and
Zahlsen
(1996)
Sprague-
Dawley
Male
75, 150, 300, 450
369, 738, 1,476,
2,214 mg/m3)
12 hr
Blood, brain,
liver, kidney,
and perirenal
fat
Testing
Table B-ll
Dahl et al.
(1988)
F344/N
Male
100
(492 mg/m3)
80 min
Inhalation
uptake
Testing
Text
This document is a draft for review purposes only and does not constitute Agency policy.
B-33	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
Toxicological Review ofTrimethylbenzene
Values for VmaxC and Km were numerically optimized based on the fit of the model
predictions to the measured blood concentrations of 1,2,4-TMB of Hissink et al. (2007) for rats
exposed once to one of three concentrations of 1,2,4-TMB as a component of WS. The optimized
value of VmaxC was only modestly different from the value determined by Hissink et al. (2007)
(initial: 3.5 vs. optimized: 3.08 mg/hr/kg0-7) from visual optimization (with slightly different
physiological parameters), but the Km value differed by 5-fold (initial: 0.25 vs. optimized: 0.050
mg/L). The increase in the LLF from 42.6 to 58.2, with two adjustable parameters, indicates that
the improvement in fit (Figure B-10) is statistically significant. The percentage of variation
explained increased from 82.3 to 90.4%, and the fit by visual inspection appears to be very good
during exposure (modestly overpredicting) and excellent in the post-exposure period. Using the
optimized kinetic parameters, the rat brain concentrations of 1,2,4-TMB were also well-
predicted (Figure B-ll).
(a)
White spirit exposure of rats to 0.05, 0.19, and 0.37 mg/L 1,2,4-TMB (Hissink et al., 2007)
0.0	1.0	2.0	3.0	4.0	5.0	6.0	7.0	8.0	9.0	10.0 11.0 12.0 13.0 14.0 15.0 16.0
Time (hr)
(b)
White spirit exposure of rats to 0.05, 0.19, and 0.37 mg/L 1,2,4-TMB (Hissink et al., 2007)
0.0	1.0	2.0	3.0	4.0	5.0	6.0	7.0	8.0	9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0
Time (hr)
Note: Rats exposed to 1,2,4-TMB in white spirit (WS) (Hissink et al.. 2007) (a) before and (b) after numerical optimization. See
Legend, Figures B-7 and B-8.
Figure B-10. Comparisons of model predictions to measured blood concentrations in
rats exposed to 1,2,4-TMB in WS.
This document is a draft for review purposes only and does not constitute Agency policy.
B-34	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
White spirit exposure of rats at 0.05, 0.19 and 0.37 mg/L 1,2,4-TMB (Hissink et al., 2007)

























1 1






I






























—
Time (hr)
Note: Rats exposed to 1,2,4-TMB in white spirit (WS) (Hissink et al., 2007). using model parameters optimized for fit to Hissink
et al. {2007) rat blood data. See Legend in Figures B-7 and B-8.
Figure B-ll. Comparisons of model predictions to measured brain concentrations in
rats exposed to 1,2,4-TMB in WS.
Venous blood 1,2,4-TMB in rats repeatedly exposed to 25,100 or 250 ppm 1,2,4-TMB (Swiercz et al., 2003)
Time (hr)
Venous blood 1,2,4-TMB in rats repeatedly exposed to 25, 100 or 250 ppm 1,2,4-TMB (Swiercz etal., 2003)
¦ ¦




! 	
i
i
i
i
1
i
_
—^

" 			
		
i




~~i
			
(b)
609
Time (hr)
Swiercz et al. (2003) in rats repeatedly exposed to 1,2,4-TMB: (a) before and (b) after numerical optimization. See Legend in
Figures B-7 and B-8.
Figure B-12. Comparisons of model predictions to measured venous blood
concentrations by Swiercz et al. (2003) in rats repeatedly exposed to 1,2,4-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
B-35	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Toxicological Review ofTrimethylbenzene
The VmaxC and Km values derived from optimization to the Hissink et al. (2007) rat data were
used as the starting values for optimizing fit to the venous blood data of Swiercz et al. (2003). in
which exposure was to 1,2,4-TMB (only) repeatedly for 4 weeks. Venous blood samples were
collected from the tail vein. The best fit parameters of VmaxC = 4.17 mg/hr/kg0-7 and Km= 0.322
mg/L produced an increase in the LLF from -28.1 to -15.6, a statistically significant
improvement, which increased the variation explained from 47.9 to 68.1% (Figure B-12). The
deviation between the model and experimental data is primarily exhibeted on the high
concentration data set When this set is not considered, the percent variation explained the
remaining two sets is 94.5%. Optimization to the low and middle concentrations alone
(omitting the high concentration) does not substantially change the parameters or increase the
LLF (simulations not shown). Optimization using the high concentration alone yields VmaxC and
Km estimates of 7.91 mg/hr/kg0 7 and 0.11 mg/L, respectively, with 96.7 percent of variation
explained (simulations not shown).
Rat Model Validation
The parameters derived from the Swiercz et al. (2003) venous blood optimizations were
used to simulate other studies in which rats and humans (see below) were exposed to
1,2,4-TMB alone (without co-exposures). The fit to the Swiercz etal. (2002) venous blood data
was very good (Figure B-13). In fact, the fit to the acute, high-exposure blood concentrations
was superior to the fit to the repeated, high-exposure data (Figure B-12b). This may reflect
adaptation (induction of metabolism) resulting from repeated, high concentration exposures.
This document is a draft for review purposes only and does not constitute Agency policy.
B-36	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
(a)
Venous blood 1,2,4-TMB during acute exposure to 25, 100, or 250 ppm 1,2,4-TMB (Swiercz et al., 2002)


i—	1
	
i	
i			1
¦ ¦ ~
; ______
	





¦
i
i 	







0	1	2	3	4	5	6
Time (hr)
Venous blood 1,2,4-TMB after acute exposure to 25, 100, or 250 ppm 1,2,4-TMB (Swiercz et al., 2002)
7	8	9	10	11	12
Time (hrs)
Swiercz et al. (2002) in acutely exposed rats: (a) during and (b) after exposure. See Legend in Figures B-7 and B-8.
Figure B-13. Comparisons of model predictions to measured rat venous blood
concentrations by Swiercz et al. (2002) in acutely exposed rats.
1	The model predictions of arterial blood and tissues in the repeated-exposure Swiercz et al.
2	(2003) study were not very accurate, considering that the venous blood data from the same
3	study were used for optimization (Table B-9). The discrepancies between seemingly
4	contemporaneous venous and arterial blood measurements were noted by the authors of the
5	original study and may be due to collection delays (i.e., tail vein for venous blood, decapitation
6	for arterial samples). The geometric mean error ratio (greater of model/experiment or
7	experiment/model) for these data was 2.8.
This document is a draft for review purposes only and does not constitute Agency policy.
B-37	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-9. Model simulated and experimental measured concentrations of 1,2,4 TMB
in male Wistar rats exposed to 1,2,4-TMB, Swiercz etal. (2003)

Exposure concentration
Model
(mg/L)
Experiment
(mg/L)a
Model:
Experiment ratio
Repeated exposure (Model t = 606 hr)
Arterial blood
25 ppm (123 mg/m3)
0.61
0.33
1.8
100 ppm (492 mg/m3)
5.0
1.54
3.2
250 ppm (1,230 mg/m3)
22.8
7.52
3.0
Brain
25 ppm (123 mg/m3)
1.91
0.45
4.2
100 ppm (492 mg/m3)
14.6
2.82
5.2
250 ppm (1,230 mg/m3)
59.0
18.6
3.2
Liver
25 ppm (123 mg/m3)
0.41
0.45
0.91
100 ppm (492 mg/m3)
10.5
3.00
3.5
250 ppm (1,230 mg/m3)
54.6
22.5
2.4
Acute exposure (Model t = 6 hr)
Arterial blood
25 ppm (123 mg/m3)
0.53
0.31
1.7
100 ppm (492 mg/m3)
7.10
1.24
5.7
250 ppm (1,230 mg/m3)
18.6
7.76
2.4
Brain
25 ppm (123 mg/m3)
2.19
0.49
4.5
100 ppm (492 mg/m3)
20.6
2.92
7.0
250 ppm (1,230 mg/m3)
62.1
18.3
3.4
Liver
25 ppm (123 mg/m3)
0.49
0.44
1.1
100 ppm (492 mg/m3)
16.3
7.13
2.3
250 ppm (1,230 mg/m3)
57.7
28.2
2.0
aData source: Swiercz et al. (2003).
1	Zahlsen and co-workers fEide and Zahlsen. 1996: Zahlsen etal.. 1992: Zahlsen et al.. 19901
2	conducted studies in which male Sprague-Dawley rats were exposed to 1,2,4-TMB by inhalation
3	for 12 hr/day. For the studies conducted at concentrations similar to those in the Swiercz
4	studies (Tables B-ll and B-10), the model error was similar to that of the arterial blood and
5	tissue measurements in the Swiercz studies (geometric mean error of 3.3 for Zahlsen et al.
6	(19901. and 2.9 for Eide and Zahlsen (19961.
This document is a draft for review purposes only and does not constitute Agency policy.
B-38	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-10. Model simulated and experimental measured concentrations of
1,2,4-TMB in male Sprague-Dawley rats exposed to 100 ppm (492 mg/m3)
1,2,4-TMB (12 hr/day, for 3 days) at the end of exposure or 12 hours after
the last exposure

Day
Model
(mg/L)
Experiment
(mg/L)a
Model:
Experiment ratio
Venous blood
1
8.52
1.71
5.0
2
8.71
1.51
5.8
3
8.72
2.06
4.2
Recovery15
1.08
0.024
7.6
Brain
1
22.6
4.58
4.9
2
23.1
4.19
5.5
3
23.1
4.39
5.3
Recovery15
0.46
Nondetect
Not calculated
Liver
1
18.2
4.93
3.7
2
18.7
3.67
5.1
3
18.7
4.25
4.4
Recovery15
0.077
0.072
1.1
Kidney (compared to
rapidly perfused)
1
22.6
13.7
1.7
2
23.1
17.1
1.4
3
23.1
12.5
1.9
Recovery15
0.46
0.24
1.9
Fat
1
491
210
2.3
2
503
165
3.1
3
504
129
3.9
Recovery15
29.1
14.4
2.0
aData from Zahlsen et al. (1992).
bRecovery period is designated as 12 hr after the last exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
B-39	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1	There was essentially no difference in the measured venous blood concentration of
2	1,2,4-TMB in the Zahlsen etal. (1992) study at 100 ppm (492 mg/m3) and at 75 ppm (369
3	mg/m3) in the Eide and Zahlsen (1996) study ((1.70 and 1.69 mg/L, respectively), so there is
4	evidently some inter-study variability or subtle differences in how the studies were conducted,
5	perhaps in the rapidity of sample collection. The Zahlsen et al. (1990) study, which used a
6	higher nominal concentration of 1,000 ppm (4,920 mg/m3), exhibited greater deviation
7	between predicted and measured blood and tissue 1,2,4-TMB concentrations (Table B-12),
8	which generally increased with a greater number of exposure days and then plateaued
9	(geometric mean errors of 2.7, 8.4,12.6,13.9, and 12.1 on exposure days 1, 3, 7,10, and 14,
10	respectively).
This document is a draft for review purposes only and does not constitute Agency policy.
B-40	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-ll. Model simulated and experimental measured concentrations of
1,2,4-TMB in male Sprague-Dawley rats exposed to 1,2,4-TMB at the end of
12 hour exposure


Model
Experiment
Model:

Exposure concentration
(mg/L)
(mg/L)a
Experiment ratio

75 ppm (369 mg/m3)
4.21
1.69
2.5
Venous blood
150 ppm (738 mg/m3)
17.8
6.9
2.6
300 ppm (1,476 mg/m3)
48.3
13.9
3.5

450 ppm (2,252 mg/m3)
78.6
26.6
3.0

75 ppm (369 mg/m3)
11.5
2.83
4.1
Brain
150 ppm (738 mg/m3)
46.6
11.7
4.0
300 ppm (1,476 mg/m3)
125
26.5
4.7

450 ppm (2,252 mg/m3)
203
48.0
4.2

75 ppm (369 mg/m3)
7.39
6.41
1.2
Liver
150 ppm (738 mg/m3)
42.2
14.8
2.9
300 ppm (1,476 mg/m3)
120
30.8
3.9

450 ppm (2,252 mg/m3)
198
56.2
3.5

75 ppm (369 mg/m3)
11.5
6.41
1.8
Kidney (compared
to Rapidly
perfused)
150 ppm (738 mg/m3)
46.6
20.2
2.3
300 ppm (1,476 mg/m3)
125
33.9
3.7

450 ppm (2,252 mg/m3)
203
59.1
3.4

75 ppm (369 mg/m3)
255
61.9
4.1
Fat
150 ppm (738 mg/m3)
987
457
2.2
300 ppm (1,476 mg/m3)
2,636
1,552
1.7

450 ppm (2,252 mg/m3)
4,276
2,312
1.8
aData from Eide and Zahlsen (1996).
1	Dahl et al. (1988) exposed male F344 rats to 1,2,4-TMB at 100 ppm (492 mg/m3) for 80
2	minutes and monitored the total uptake. Under the conditions of the experiment, it was
3	determined that average rat took up 3.28 (trial 1) or 3.89 (trial 2) mg 1,2,4-TMB. In a model
4	simulation, the predicted uptake was 3.61 mg. Geometric mean model error for the two trials
5	was 1.2.
This document is a draft for review purposes only and does not constitute Agency policy.
B-41	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-12. Model simulated and experimental measured concentrations of
1,2,4-TMB in male Sprague-Dawley rats exposed to 1,000 ppm (4,920
mg/m3) 1,2,4-TMB (12 hr/day, for 14 days) at the end of exposure

Day
Model
(mg/L)
Experiment
(mg/L)a
Model:
Experiment ratio
Venous blood
1
181
63.5
2.8
3
293
43.1
6.8
7
372
33.4
11.1
10
395
34.0
11.6
14
399
35.2
11.3
Brain
1
465
120
3.9
3
747
64.9
11.5
7
946
63.5
14.9
10
1,005
62.1
16.2
14
1,014
71.5
14.2
Fat
1
9,919
5,860
1.7
3
17,328
2,282
7.6
7
22,323
1,835
12.2
10
23,763
1,677
14.2
14
23,961
2,169
11.0
aData from Zahlsen et al. (1990).
Human Model Validation
1	Kinetic parameters derived from optimal fit for rat venous blood data (described above)
2	were tested for the applicability to human kinetics by comparison to studies in which humans
3	were exposed to 1,2,4-TMB alone or 1,2,4-TMB in co-exposures with WS (Table B-13). The key
4	data set for validation in humans was deemed to be Kostrzewski et al. (1997) because these
5	volunteers were exposed to 1,2,4-TMB alone (no co-exposure, as in Hissink et al. (2007)) under
6	sedentary conditions (i.e., level of effort was not elevated, as in Jarnberg et al. (1998.1997a:
7	1996)).
This document is a draft for review purposes only and does not constitute Agency policy.
B-42	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
Toxicological Review ofTrimethylbenzene
Using the VmaxC and Km derived from the Swiercz et al. (20031 rat repeated exposure data,
the simulated blood concentration underestimated those measured during exposure of human
volunteers by Kostrzewski et al. (1997), then overpredicted blood concentrations up to 7 hours
post-exposure, and underpredicted subsequent measured blood concentrations (Figure B-14).
Of 21 blood measurements, only two differed from the simulated value by more than a factor of
2 (maximum: 2.6), with a geometric mean deviation of 1.5-fold between the simulated and
measured values. The percent variation explained was 69.74%. When Km was held constant and
VmaxC was optimized (final value: 3.39 mg/hr/kg0 7), the improvement in fit was minimal
(72.14% of variation explained), and not statistically significant, so the rat-derived values were
considered acceptable (see the subsection regarding Rat Model Optimization, in Section
B.3.3.2).
Table B-13. Human kinetic studies of 1,2,4-TMB used in model validation
Reference
Ethnicity
Gender
Nominal
concentration
Exposure
regimen
1,2,4-TMB
measurements
Use in
model
evaluation
Form of
comparison
Kostrzewski
et al. (1997)a
Not stated;
conducted
in Poland
Sex not
stated.
Assumed
male.
30 ppm
(147.6 mg/m3)
8 hr
Venous blood
time course
Testing
Figure B-14
Jarnberg et al.
(1999; 1998,
1997a; 1996)b
Caucasian;
conducted
in Sweden
Male
2 and 25
(~10 and 123
mg/m3)
2 hr at 50
W
(bicycle)
Venous blood
and exhaled air
time course
Testing
(blood data
only)
Figure B-15
Hissink et al.
(2007)°
Not stated;
spoke Dutch
as "native
language"
Male
100 ppm WS
with 7.8%
1,2,4-TMB
(~38.3 mg/m3
1,2,4-TMB)
6 hr
Venous blood
and end exhaled
air time course
Testing
Figure B-16
aFive volunteers, ages 24-37, with no known occupational exposure to 1,2,4-TMB. Height of 1.70 to 1.86 m and BW of 70-97 kg. The
average of the high and low values for age, height, and weight plus assumed gender (male) were used to calculate central tendency
estimate of 22.44% for volume of body fat (VFC), per Deurenberg et al. (1991). QPC estimated from the midpoint of the range for
total ventilation (0.56 to 1 mB/hr), average of high and low body weights, BW0'74 scaling, and an assumption that alveolar ventilation
was 2/3 of total ventilation.
bTen volunteers, average age 35, range 26-48, with no known occupational exposure to solvents; volunteers were instructed to avoid
contact with organic solvent and to refrain from taking drugs or drinking alcoholic beverages for 2 days before exposure. Average BW
76.5 kg. Alveolar ventilation rate (QPC) estimated from the mean value for total ventilation rate during exposure, average body
weights, BW0'74 scaling, and an assumption that alveolar ventilation was 2/3 of total ventilation. Digitized blood data (group averages)
extracted from figures.
cThree volunteers, ages 23-26, BW 69-82 kg, mean body fat of 14.6% (skin caliper measurement); alcohol consumption 10-15
drinks/week (all subjects), one smoker (4 cigarettes per day).
This document is a draft for review purposes only and does not constitute Agency policy.
B-43	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Blood 1,2,4-TMB in human volunteers exposed to 154 mg/m3 1,2,4-TMB (Kostrzewski et al., 1997)
n
¦
S
\














"V


¦

I

¦

i


-










0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92 94 96
Time (hr)
Note: Kostrzewki et al. (1997) in human volunteers exposed to 154 mg 1,2,4-TMB/m3 for 8 hours.
Figure B-14. Comparisons of model predictions to measured human venous blood
concentrations in Kostrzewki et al. (1997) in human volunteers exposed to 154 mg
1,2,4-TMB/m3 for 8 hours.
1	For comparisons between the Jarnberg and Johason (1999) and Jarnberg et al. (1998.
2	1997a: 1996) data and the model, simulations were conducted with QPC (calculated as
3	described in footnote to Table B-13) at the elevated (working) level throughout the simulation,
4	but with no other adjustments made for exercise conditions. The model consistently
5	underpredicted the measured venous blood concentrations of 1,2,4-TMB (Figure B-15). At 25
6	ppm (123 mg/m3), blood concentrations were underpredicted by a factor of 2.1 to 3.5 during
7	exposure and by a factor of 1.04 to 1.5-fold in the post-exposure period, for a geometric mean
8	discrepancy of 1.7 for this concentration. At 2 ppm (~10 mg/m3), blood concentrations were
9	underpredicted by factors of 1.7 to 2.7 during exposure and 1.01 to 1.2 in the post-exposure
10	period, for a geometric mean discrepancy of 1.6 for this concentration.
This document is a draft for review purposes only and does not constitute Agency policy.
B-44	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Blood concentrations of 1,2,4-TMB in volunteers exposed to 2 or 25 ppm 1,2,4-TMB (Jarnberg and coworkers)
1
E
c
o
2 0.1
c
0)
¦o
o
o
CD
0.001
0
1
2
3
4
5
6
Time (hr)
Note: Jarnberg et al. (1998.1997a; 1996) in volunteers exposed to 2 or 25 ppm (~10 or 123 mg/mB) 1,2,4-TMB for 2 hours while
riding a bicycle (50 W).
Figure B-15. Comparisons of model predictions to measured human venous blood
concentrations of Jarnberg etal. (1998,1997a; 1996) in volunteers exposed to 2 or
25 ppm (~10 or 123 mg/m3) 1,2,4-TMB for 2 hours while riding a bicycle (50 W).
1	Comparisons of model predictions and experimental data were also made for the human
2	study described in Hissink et al. (2007) in which volunteers inhaled 100 ppm WS with 7.8%
3	1,2,4-TMB (38.4 mg/m31,2,4-TMB) for 4 hours (Figure B-16). The agreement between
4	simulated and measured concentrations of 1,2,4-TMB in blood during exposure was excellent.
5	The agreement between the modeled and measured 1,2,4-TMB in end-exhaled air during the
6	post-exposure period was very good.
This document is a draft for review purposes only and does not constitute Agency policy.
B-45	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
White spirit exposure of humans at 45 mg/L (100 ppm) (Hissink et al., 2007)



~




























r D
D ~
~
Lo










yi
D


\D










/n



C-h ~










A



~ CJN
k









P
In














0.12
0.1
0.08
0.06
: 0.04
0.02
0
12	13
Time (hr)
White spirit exposure of humans at 45 mg/L (100 ppm)






















e
r——¦-













E
t











8 1.0 xlO E-3
(b)
Note: (a) human venous blood and (b) end of exposure exhaled air 1,2,4-TMB in human volunteers exposed to 100 ppm WS
with 7.8% 1,2,4-TMB (38.4 mg/m 1,2,4-TMB) (Hissink et al„ 2007).
Figure B-16. Comparisons of model predictions to measured (a) human venous blood
and (b) end of exposure exhaled air 1,2,4-TMB in human volunteers exposed to 100
ppm WS with 7.8% 1,2,4-TMB (38.4 mg/m3 1,2,4-TMB).
This document is a draft for review purposes only and does not constitute Agency policy,
B-46	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
Toxicological Review ofTrimethylbenzene
Summary of Optimization and Validation
Numerical optimization of the fit to the rat data in Hissink et al. (2007) produced a similar
VmaxC, but smaller Km than the values determined by Hissink et al. (2007) using visual
optimization. Changes made to values of physiological parameters may have contributed to the
differences in optimized values. Because the rats in the Hissink et al. (2007) study were co-
exposed to other components of WS, the potential for these other components to alter the
kinetics of 1,2,4-TMB was noted as a possible concern for predicting the kinetics of 1,2,4-TMB in
test animals with no co-exposures. Another concern was the potential for kinetic changes with
repeated exposure. As the Swiercz et al. (2003) rat kinetic study involved repeated exposure to
1,2,4-TMB without potentially confounding co-exposures, and provides post-exposure venous
blood time course data, it appears to be the most suitable for describing kinetics relevant to
chronic RfC and RfD development The VmaxC and Km values from the numerical optimization to
the Hissink et al. (2007) rat data were used as starting values for optimization of the fit to the
Swiercz et al. (2003) venous blood data. The improvement in fit for the low and middle
concentrations (25 and 100 ppm [123 and 492 mg/m3]) was apparent from careful visual
inspection and was statistically significant, and these values were used in subsequent validation
simulations.
In general, the model simulations of venous blood concentrations in exposed Wistar rats,
uptake by F344 rats, and venous blood and exhaled breath of human volunteers were
acceptable. The measured Wistar rat arterial blood and tissue concentrations were consistently
overpredicted by the model, suggesting collection delays in the studies. The model also
consistently overpredicted the measured Sprague-Dawley rat tissue and blood concentrations,
including the "recovery" (12 hr post-exposure) samples, which should not be subject to
collection delays. Many of the "validation" comparisons were made at exposure concentrations
(250 ppm [1,230 mg/m3]or greater) for which the optimized model did not provide accurate
venous blood concentrations. It cannot be determined with the available data whether the 2-3-
fold differences between the model and Sprague-Dawley rat blood concentrations at lower
concentrations (75 and 150 ppm [369 and 738 mg/m3]) are due to methodological differences
(e.g., in sample collections and analysis) or true strain differences. Overall, we conclude that the
optimized model produces acceptable simulations of venous blood 1,2,4-TMB for chronic
exposure to < 100 ppm (492 mg/m3) for rats or < 30 ppm (147.6 mg/m3) for humans 1,2,4-TMB
by inhalation. If rat exposures of interest exceed 100 ppm (492 mg/m3), consideration should
be given to reassessing model validation at high concentrations using VmaxC and Km parameters
optimized for repeated, high concentration exposures [e.g., 250 ppm (1,230 mg/m3) from
Swiercz et al.(2003)].
This document is a draft for review purposes only and does not constitute Agency policy.
B-47	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Toxicological Review ofTrimethylbenzene
B.3.3.3. Sensitivity Analysis of Rat Model Predictions
The primary objective of the sensitivity analysis was to evaluate the ability of the available
data to unambiguously determine the values of both VmaxC and Km (i.e., parameter
identifiability). Toward this end, sensitivity analyses were conducted using acslX. Because the
selected key data set was the venous blood concentrations in the Swiercz et al. (2003) study,
simulations were conducted to see how small changes in parameters changed the estimated
venous blood concentrations under the conditions of this study, simulating the first 12 hours
(6 hrs exposure, 6 hrs post-exposure), conditions that are essentially identical to those in
Swiercz et al. (2002). The evaluations were limited to the lowest (25 ppm [123 mg/m3]) and
highest (250 ppm [1,230 mg/m3]) exposure concentrations. It should be noted that after the
optimization (Figure B-13b), the agreement between the model and the experimental data at
the lower exposure concentration was superior to the agreement at the high concentration, so
the low concentration sensitivity analysis results are somewhat more meaningful than the high
concentration results. The results are calculated as normalized sensitivity coefficients (NSC)
(i.e., percent change in output/percent change in input, calculated using the central difference
method).
The interpretation of the sensitivity analysis outputs focused on the times during which
blood concentrations were measured, so the sensitivity analyses for the first 15 minutes of
exposure were not considered relevant. Parameters are grouped (Table B-14) as relatively
insensitive (maximum|NSC| < 0.2 for 0.25 hr < t < 12 hr), moderately sensitive (0.2 <
maximum|NSC| < 1.0), or highly sensitive (maximum|NSC| > 1.0).
VmaxC/Km was identifiable from the data (as opposed to VmaxC and Km each being
identifiable), one would expect that the NSC for these parameters would always be opposite in
sign, and equal in magnitude, which is not the case. We conclude that Km and VmaxC are distinctly
identifiable using the Swiercz et al. (2003: 2002) data.
This document is a draft for review purposes only and does not constitute Agency policy.
B-48	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Toxicological Review ofTrimethylbenzene
While the focus of this sensitivity analysis was to evaluate the identifiability of chemical-
specific parameters from the available data, additional insights can be obtained by considering
the other "sensitive" parameters. Predicted blood concentrations were sensitive to the value of
QPC (ventilation rate). If high concentrations produce a sedative effect, decreases in ventilation
could contribute to the model's greater over-prediction of the experimentally measured values
at high concentrations [e.g., as high as 1,000 ppm (4,920 mg/m3), in Zahlen et al. (1990)]. The
accuracy of the predicted net uptake in the Dahl et al. (1988) study indicates that, at 100 ppm
(492 mg/m3), the model value of QPC is likely appropriate, since net uptake in this relatively
short experiment (80 minutes) is highly sensitive to the breathing rate (simulations not shown).
The fractional volumes of the fat and slowly perfused tissues compartments are also
moderately important parameters (with time courses similar to those of the corresponding
partition coefficients shown in Figure B-15). The volume of the fat compartment in particular is
known to vaiy with age and strain (Brown etal.. 1997). so using the same value for all studies
might have an impact on the predicted kinetics.
This document is a draft for review purposes only and does not constitute Agency policy.
B-49	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-14. Parameter sensitivity for venous blood 1,2,4-TMB concentration in rats
exposed to 1,2,4-TMB via inhalation
Parameter
Insensitive
(maximum |NSC| <0.2)
Moderately sensitive
(0.2 < maximum |NSC| <1.0)
Highly sensitive
(maximum |NSC| > 1.0)
BW

L, H

CONC


L, H
QPC


L, H
VmaxC

L, H

Km
H
L

PB
L
H



L, H

PS

L, H

PR
L, H


PL
L, H


PBR
L, H


VFC

L, H

VSTOTC

L, H

VRTOTC
L, H


VLC
L, H


VBRC
L, H


QCC

H
L
QFC

L, H

QRTOTC

L, H

QLC
H

L
QBRC
L, H


L = low exposure concentration (25 ppm [123 mg/m3]), H = high exposure concentration (250 ppm [,1230 mg/m3]).
Body weight (BW), concentration of 1,2,4-TMB in the air (CONC), alveolar ventilation rate (QPC), Michaelis-Menten maximum
rate of metabolism (VmaxC), Michaelis-Menten constant: concentration where Vm>ax is half-maximal (Vmax), blood:air partition
coefficient (PB), fat:blood partition coefficient (PF), slowly perfused:blood partition coefficient (PS), rapidly perfused:blood
partition coefficient (PR), liver:blood partition coefficient (PL), brain:blood partition coefficient (PBR), volume of fat (VFC),
volume of slowly perfused tissues (VSTOTC), volume of rapidly perfused tissues (VRTOTC), volume of liver (VLC), volume of brain
(VBRC), cardiac output (QCC), blood flow to fat (QFC), blood flow to slowly perfused tissues (QRTOTC), blood flow to liver (QLC),
blood flow to brain (QBRC)
This document is a draft for review purposes only and does not constitute Agency policy.
B-50	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
QJ
O
U
c
01
"O
0.8
(b)
Sensitivity analysis: rat CV, low concentration
exposure
{Swierczet al., 2002, 2003)
cv:km
cv:vmaxc
cv:pb
cv:pf
cv:ps
Time (hr)
Sensitivity analysis: ratCV, high concentration
exposure
(Swierczet al., 2002, 2003)
cv:km
cv:vmaxc
cv:pb
cvrpf
cv:ps
Time (hr)
Note: Rats exposed to (a) 25 ppm (123 mg/m3) or (b) 250 ppm (1,230 mg/m3) of 1,2,4-TMB via inhalation for 6 hours (Swiercz et
ai., 2003; Swiercz et al.. 2002).
Figure B-17. Time course of normalized sensitivity coefficients of moderately
sensitive chemical-specific parameters (response: venous blood concentration) in
rats exposed to (a) 25 ppm (123 mg/m3) or (b) 250 ppm (1,230 mg/m3) of 1,2,4-TMB
via inhalation for 6 hours.
This document is a draft for review purposes only and does not constitute Agency policy,
B-51	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
Toxicological Review ofTrimethylbenzene
B.3.3.4. Sensitivity Analysis of Human Model Predictions
A sensitivity analysis for human model predictions to all parameters was conducted for
continuous inhalation exposures, and results are shown in Table B-15. The results are
presented as normalized sensitivity coefficients (i.e., percent change in output/percent change
in input, calculated using the central difference method; NSC). Similar to analyses performed for
the rat, parameters are noted as relatively insensitive (|NSC| < 0.2), moderately sensitive (0.2 <
| NSC | < 1.0), or highly sensitive (|NSC| > 1.0). To bracketthe range of human equivalent
concentrations (HECs), inhalation sensitivities were evaluated at 10 and 150 ppm (49.2 and
738 mg/m3) concentration. The resulting coefficients (Table B-15) are not surprising. The two
fitted metabolic parameters, VmaxC and Km both influence model predictions. The VmaxC
sensitivity is higher at 150 ppm (738 mg/m3) (|0.88731) than at 10 ppm (49.2 mg/m3) (|0.238|)
due to the slight metabolic saturation.
This document is a draft for review purposes only and does not constitute Agency policy.
B-52	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-15. Parameter sensitivity for steady-state venous blood 1,2,4-TMB
concentration in humans exposed to 1,2,4-TMB via inhalation
•arameter
Insensitive
(maximum|NSC| <0.2)
Moderately sensitive
(0.2< maximum|NSC| < 1.0)
Highly sensitive
(maximum | NSC| > 1.0)
BW
L, H


CONC

L
H
QPC

L, H

VmaxC

L, H

Km
L, H


PB
L, H



L, H


PS
L, H


PR
L, H


PL
L, H


PBR
L, H


VFC
L, H


VSTOTC
L, H


VRTOTC
L, H


VLC
L, H


VBRC

L, H

QCC
L, H


QFC
L, H


QRTOTC

L, H

QLC
L, H


L = low exposure concentration (10 ppm [49.2mg/m3]), H = high exposure concentration (150 ppm [738 mg/m3]).
Body weight (BW), concentration of 1,2,4-TMB in the air (CONC), alveolar ventilation rate (QPC), Michaelis-Menten maximum
rate of metabolism (VmaxC), Michaelis-Menten constant: concentration where Vm>ax is half-maximal (Vmax), blood:air partition
coefficient (PB), fat:blood partition coefficient (PF), slowly perfused:blood partition coefficient (PS), rapidly perfused:blood
partition coefficient (PR), liver:blood partition coefficient (PL), brain:blood partition coefficient (PBR), volume of fat (VFC),
volume of slowly perfused tissues (VSTOTC), volume of rapidly perfused tissues (VRTOTC), volume of liver (VLC), volume of brain
(VBRC), cardiac output (QCC), blood flow to fat (QFC), blood flow to slowly perfused tissues (QRTOTC), blood flow to liver (QLC),
blood flow to brain (QBRC)
This document is a draft for review purposes only and does not constitute Agency policy.
B-53	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Toxicological Review ofTrimethylbenzene
B.3.3.5. Modification of the Hissink et al. (2007) model to include oral route of
exposure
For derivation of an oral RfD, the updated 1,2,4-TMB PBPK model based on Hissink et al.
(20071 was further modified by adding code for continuous oral ingestion. It was assumed that
100% of the ingested 1,2,4-TMB is absorbed by constant infusion of the oral dose into the liver
compartment There were no oral data available to calibrate the model for oral absorption and
no data were available evaluate the model predictions following oral ingestion either. Thus,
although the assumption that 100% of the dose would enter the liver is a common assumption,
it does represent an area of uncertainty in the route-to-route extrapolation used to derive oral
reference values.
The contribution of the first-pass metabolism in the liver for oral dosing was evaluated by
simulating steady state venous blood levels (at the end of 50 days continuous exposure) for a
standard human at rest (70 kg) for a range of concentrations and doses. For ease of visual
comparison (Figure B-18), concentrations were converted to daily doses based on the amount
of 1,2,4-TMB inhaled, as computed by the model. (An inhaled concentration of 0.001 mg/L [0.20
ppm (0.98 mg/m3)] is equivalent to an inhaled dose of 0.12 mg/kg/day.) At both very low and
very high daily doses by inhalation or oral dosing, steady state CV is essentially linear with
respect to the daily dose, but with different CV/dose ratios and a transition zone between 1 and
100 mg/kg/day. At low daily doses, equivalent inhalation doses result in steady state blood
concentrations 4-fold higher than an equivalent oral dose due to the hepatic first-pass effect.
The first-pass effect becomes insignificant with respect to steady-state venous blood
concentrations for daily doses in excess of ~50 mg/kg/day.
This document is a draft for review purposes only and does not constitute Agency policy.
B-54	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Toxicological Review ofTrimethylbenzene
OO
oo
E
w
a)
a.
&o
E
O)
O
TJ
re
-a
0.1
0.01
¦Oral
Inhalation
0.1 1 10 100 1000 10000100000
Daily dose (mg/kg/d)
Figure B-18. Effect of route of exposure and dose rate on steady-state venous blood
concentration (t = 1,200 hr) for continuous human exposure to 1,2,4-TMB.
B.3.3.6. Conclusions
Several changes were made to the model for use in this assessment: (1) Updated
physiological parameters were implemented (Brown et al.. 19971: (2) Hepatic metabolism was
revised to omit variation over time and new VmaxC and Km values were estimated through
numerical optimization; and (3) An oral dosing component was added to the model as constant
infusion into the liver compartment The values were optimized to Hissink et al. (2007) data
and resulted in a VmaxC of 4.17 mg/hr/kg0-7 and Km of 0.322 mg/L. In addition, the model was
tested for its ability to predict published rat data resulting from exposure to 1,2,4-TMB alone
(Swiercz etal.. 2003: Swiercz etal.. 2002: Eide and Zahlsen. 1996: Zahlsen etal.. 1992: Zahlsen
etal.. 1990: Dahl etal.. 19881. Using the optimized values, the model adequately predicted the
data and lower concentrations. Human data (Hissink etal.. 2007: Tarnberg and Tohanson. 1999:
Tarnberg etal.. 1998.1997a: Kostrzewski et al.. 1997: Tarnberg etal.. 19961 were also utilized to
validate model predictions.
B.3.4. Summary of Available PBPK models for 1,3,5-TMB or 1,2,3-TMB
There are currently no available PBPK models for rodents or humans for either 1,3,5-TMB
or 1,2,3-TMB.
This document is a draft for review purposes only and does not constitute Agency policy.
B-55	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
B.4. HUMAN STUDIES
Table B-16. Characteristics and quantitative results for epidemiologic cross-sectional
study of exposure to 1,2,4-TMB. Battig et al. (1956), as reviewed by Baettig
etal. (1958)
Study (location)
Outcome assessment
• Transportation plant in Switzerland
•	Survey was conducted to investigate the CNS,
respiratory, hematological effects of long-term
TMB exposure
•	Additional information on working history,
personal history, and psychiatric health was
collected
POPULATION CHARACTERISTICS
Exposed population
Referent or control description
• 27 TMB-exposed workers that worked primarily in the
painting shop of the transportation plant
• 10 unskilled workers from the same plant that
were not exposed to TMB vapors.
Exposure assessment
Statistical analysis
• Exposure level: 10-60 ppm (49.2-295 mg/m3) in
working rooms
• No statistical analyses were reported.
• Exposure duration: approximately 10 years

• Compounds to which study participants were exposed:
Fleet-X DV-9, a solvent that contained 1,2,4-TMB and
1,3,5-TMB (50% and 30%, respectively) for
approximately 10 years. Fleet-X DV-99 also potentially
contained 1,2,3-TMBand numerous
methylethylbenzenes.

This document is a draft for review purposes only and does not constitute Agency policy.
B-56
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-16 (Continued): Characteristics and quantitative results for epidemiologic
cross-sectional study of exposure to 1,2,4-TMB. Battig et al.
	(1956), as reviewed by Baettig etal. (1958)	
RESULTS
Exposure subgroup
•	Increased self reports of vertigo, headaches, and drowsiness during work.
•	Increased presence of chronic asthmatic bronchitis, anemia, and altered blood clotting characteristics (e.g.,
increased clotting time and tendency to hemorrhage).
•	Increased vitamin C deficiency was observed in controls, but the authors attribute this to nutritional
deficiencies in this population.
Effect estimate ( 95% CI)
Figure 1. Clinical findings obtained from workers exposed to TMB compared to unskilled worker controls not
exposed to TMB.
¦ Painters n=27 ~ Unskilled workers n=10
i	
60	¦	
¦
Subjective Asthmatic Anemia Tendency Vitamin C
complaints bronchitis <4.5x10(6) to deficiency
voiced during RBC/jjL hemorrhage
	work	
Source: Reproduced with permission of Springer-Verlag (Battig et al., 1958)
Data source: Battig et al. (1956), as reviewed by Baettig et al. (1958)
CD


C/3
"^4—»
o
O)
CD
c
Q.
TD

C
CD



CD
Id
-C
o


«4—
"o
o
cn
c
CD
o
-#—»
O
cu
CD
Q_
Q_

This document is a draft for review purposes only and does not constitute Agency policy.
B-57	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-17. Characteristics and quantitative results for epidemiologic cross-sectional
study of exposure to 1,2,4-TMB; Billionnet et al. (2011)
Study (location)
Outcome assessment
• Random selection of dwellings throughout France
•	Standardized, self-administered questionnaire was
completed by participants to determine number and
severity of respiratory effects, particularly asthma
and rhinitis.
•	Additional information on daily habits, smoking
status, and sociodemographic variables was
collected.
•	Diagnosis of rhinitis or asthma was not confirmed by
a physician.
POPULATION CHARACTERISTICS
Exposed population
Referent or control description
•	1,612 individuals living in 567 dwellings, aged 15 or
older.
•	Surveys were conducted and air samples were
collected over a period of one week.
• The study cohort was also used as the control group.
Dwellings with low levels of individual volatile
organic compound (VOCs) were used as controls for
that particular compound.
Exposure assessment
Statistical analysis
•	Exposure level: For 1,2,4-TMB, exposure varied from
undetectable to 111.7 ng/m3, with median
concentration 4.0 ng/m3.
•	Exposure duration: Not reported; reported
measurements represent the means of one week of
monitoring.
•	Pollutant correlations tested by Spearman's rank
correlation coefficient.
•	Generalized estimating equation approach used to
adjust for correlations between individuals within
same dwelling.
•	Global VOC score was created to address exposure
to multiple pollutants.
•	All models were adjusted for age, sex, and smoking
status.
This document is a draft for review purposes only and does not constitute Agency policy.
B-58
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-17 (Continued): Characteristics and quantitative results for epidemiologic
cross-sectional study of exposure to 1,2,4-TMB; Billionnet et
	al. (2011)	
RESULTS
Exposure subgroup
•	Statistically significant increase in odds ratios for asthma following 1,2,4-TMB exposure.
•	No statistically significant increase in odds ratio for rhinitis and 1,2,4-TMB exposure.
Effect estimate ( 95% CI)
Figure 1. Odds ratios for asthma and asthma/rhinitis and exposure to 1,2,4-TMB. For all models, data was
adjusted for confounders.
Odds Ratios for Asthma Associated with 1,2,4-TMB Exposure
Odds ratio for asthma
according to adjusted
marginal model

Odds ratio for asthma
25th vs. 75th
A
percentiles

Odds ratio for asthma
95th vs. 75th
percentiles
	o	
0 1
2 3 4 5 6 7
Odds Ratio
Source: Billionnet et al. (2011)
This document is a draft for review purposes only and does not constitute Agency policy.
B-59	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-18. Characteristics and quantitative results for epidemiologic cohort study of
exposure to 1,2,4-TMB. Chen etal. (1999)
Study (location)
Outcome assessment
• Dockyard in Scotland, United Kingdom
•	Survey was conducted to determine mortality,
symptoms, and risks of paint exposure.
•	Additional information on age, education, smoking,
alcohol consumption, and personality was collected.
POPULATION CHARACTERISTICS
Exposed cohort
Referent or control description
•	1292 TMB-exposed males who worked as painters in
a dockyard for at least 1 yr between 1950 and 1992.
•	Follow up period extended from 1960 through 1994
• 953 individuals matched by age and selected from
lists of patients of local primary care physicians.
Exposure assessment
Statistical analysis
•	Exposure level: Specific concentrations not
discussed
•	Exposure duration: at least 1 yr; range 1-41 years
•	Compounds to which study participants were
exposed: white spirit (1,2,4-TMB), xylene, TMB
(unspecified), n-butanol, trichlorethylene, naptha,
and cumene.
•	Intra-cohort proportional mortality ratios were
calculated, as were standardized mortality ratios for
comparison with all Scottish males. 95% confidence
intervals calculated assuming a Poisson distribution.
•	x2test used to assess differences in
neuropsychological symptoms between painters
and non-painters.
•	Brestow-Cox model used to adjust for covariates
including educational level, smoking, alcohol
consumption, and social conformity.
•	Log-regression model used for case-control study.
This document is a draft for review purposes only and does not constitute Agency policy.
B-60
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-18 (Continued): Characteristics and quantitative results for epidemiologic
cohort study of exposure to 1,2,4-TMB. Chen et al. (1999)
RESULTS
Exposure subgroup
•	Increased prevalence rate ratios for neuropsychological symptoms amongst painters.
•	Rate ratios increased significantly with increasing number of years of exposure, even after adjustment for
possible confounders.
•	Multivariate-adjusted odds ratios within nested case-control analysis showed same relationship.
Effect estimate ( 95% CI)
Figure 1. Unadjusted and adjusted prevalence rate ratios for neuropsychological symptoms in dockyard painters
vs. controls. With increasing years of exposure, rate ratios were found to increase. Symptoms included
difficulty in buttoning and unbuttoning, trembling hands, or unsteadiness in arms or legs. For trend in
unadjusted rate ratios, p<0.00001.
Unadjusted and Adjusted Prevalence Rate Ratios for
Neuropsychological Symptoms in Dockyard Painters
UNADJUSTED

1-4 Years Exposure
	o	
5-9 Years Exposure
	o	
10-14 Years Exposure
	o	
15-41 Years Exposure
	o	
ADJUSTED
1-4 Years Exposure
	o	
5-9 Years Exposure
	o	
10-14 Years Exposure
	o	
15-41 Years Exposure
	o	
0 ]
2 3 4 5 6 7
Prevalence Rate Ratio
Source: Chen et al. (1999).
This document is a draft for review purposes only and does not constitute Agency policy.
B-61	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-18 (Continued): Characteristics and quantitative results for epidemiologic
	cohort study of exposure to 1,2,4-TMB. Chen et al. (1999)
Figure 2. The effect of elapsed time since cessation of painting on all symptoms. Values reported are prevalence
rate ratios for painters vs. non-painters. No significant decrease in risk with increasing post-exposure
time was found.
Prevalence Rate Ratios for Neuropsychological Symptoms in
Dockyard Painters Following Cessation of Exposure
Active Painters
1-10 years post-exposure
11-18 years post-exposure
>19 years post-exposure
2	3
Prevalence Rate Ratio
Source: Chen et al. (1999).
This document is a draft for review purposes only and does not constitute Agency policy.
B-62	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-18 (Continued): Characteristics and quantitative results for epidemiologic
cohort study of exposure to 1,2,4-TMB. Chen et al. (1999)
Figure 3. The effect of exposure duration on odds ratio for neuropsychological symptoms. With increasing years
of exposure, odds ratios were found to increase.
Odds Ratios for Neuropsychological Symptoms in Dockyard Painters
1-4 Years Exposure

V
5-9 Years Exposure
	$	
10-14 Years Exposure
	o	
15-41 Years Exposure
	o	
0
2 4 6 8 10 12 14 16 18 20 22 24 26
Odds Ratio
Source: Chen et al. (1999).
This document is a draft for review purposes only and does not constitute Agency policy.
B-63	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-19. Characteristics and quantitative results for controlled human exposure
study of exposure to 1,2,4-TMB in WS. Lammers et al. (2007)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Humans
M
12
Inhalation
57 or 570 mg/m3
4 hrs
Additional Study details






• Human volunteers were exposed to 57 or 570 mg/m3 during two test sessions separated by 1 week, each lasting 4
hrs.
• Several tests were conducted to evaluate impact of WS on CNS. These included tests of observation, reaction time,
and hand-eye coordination.
• In humans, attention deficit was observed following WS inhalation.


• The study protocol was approved by the TNO's Institutional Review Board




Test scores (mean ± SD) at various time points in humans exposed to
57 or 570 mg/mB WS, for 4 hrs
Observation


57 mg/mB
570 mg/mB
Mood and affect
Fatigue (scale score)
Pre-test
1.11 ±0.04
1.11 ±0.05
1 hr
1.06 ± 0.03
1.17 ±0.09
3 hrs
1.21 ±0.12
1.29 ±0.13
Post-test
1.38 ±0.15
1.51 ±0.23
Vigor (scale score)
Pre-test
3.35 ±0.20
3.53 ±0.09
1 hr
3.58 ±0.16
3.23 ±0.20
3 hrs
3.27 ±0.20
3.32 ±0.22
Post-test
2.98 ±0.23
3.05 ±0.22
Psychomotor skills (hand-eye coordination and finger tapping)
Hand-eye coordination test (pixels in InMAE)
Pre-test
1.69 ±0.05
1.67 ± 0.04
1 hr
1.56 ±0.05
1.64 ± 0.04
3 hrs
1.64 ± 0.05
1.63 ± 0.04
Post-test
1.62 ± 0.04
1.55 ±0.06
Finger tapping test (no. of taps in 30 seconds)
Pre-test
201 ± 7
203 ±6
1 hr
205 ±5
194 ±6
3 hrs
202 ±8
196 ±6
Post-test
198 ± 7
200 ±6
This document is a draft for review purposes only and does not constitute Agency policy.
B-64
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-19 (Continued): Characteristics and quantitative results for controlled
human exposure study of exposure to 1,2,4-TMB in WS.
	Lammers etal. (2007)	
Attention
Reaction time test (latency, ms)
Pre-test
251 ±9
246 ±8
0.25 hrs
248 ± 10
252 ±9
1 hr
248 ±9
254 ±9
2.25 hrs
253 ±9
266 ± 12
3 hrs
253 ± 11
257 ± 10
Post-test
258 ± 11
269 ± 13
Color word vigilance test (latency, ms)
Pre-test
579 ± 28
595 ±22
1 hr
550 ± 20
569 ± 20
3 hrs
537 ± 17
561 ±23
Post-test
532 ± 18
557 ±22
Figure 2. Performance on finger tapping test with the dominant hand at different time points during and after
exposure.
m 2201
CO
I
+*
o 210
c
re
9
<£• 200-
CQ
o
to
c
CO 190-
Q.
re
+¦»
0)
c 180
I4-
Health Effect at LOAEL
NOAEL
LOAEL
n/a
n/a
n/a
Comments: Exposure to 1,2,4-TMB was via WS, which is comprised of additional substances. LOAEL and NOAEL for 1,2,4-TMB alone
cannot be extracted from this study because other constituents of the WS mixture may confound results.
Source: Lammers et al. (2007).
-	placebo
-	White Spirit
pre-test	1-hr	3-hr	post-test
time of testina
This document is a draft for review purposes only and does not constitute Agency policy.
B-65	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-20. Characteristics and quantitative results for epidemiologic cohort study of
exposure to 1,2,4-TMB. Lee etal. (2005)
Study (location)
Outcome assessment
• A shipyard in Ulsan, Korea
•	Various neurobehavioral parameters were
measured with computer-based neurobehavioral
assessments.
•	Measured parameters included simple reaction
time, symbol digit substitution, and finger tapping
speed.
•	Additional information on occupational history,
medical history, age, work duration, education
level, alcohol use, and smoking status.
POPULATION CHARACTERISTICS
Exposed population
Referent or control description
• 180 shipyard workers exposed to mixed organic
solvents.
• 60 Shipyard workers that were not exposed to
mixed organic solvents were used as the referent
• Workers were exposed generally during painting
activities within the shipyard.
group
Exposure assessment
Statistical analysis
• Data on exposure was collected from 61 workers who
wore passive dosimeters on 3 work days.
• A cumulative exposure index was calculated for
each worker.
• Average Exposure duration: 16.5±9 years in exposed
workers.
• Student t-test was used to determine statistical
significance of results in exposed workers
compared to non-exposed workers.
This document is a draft for review purposes only and does not constitute Agency policy.
B-66
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-20 (Continued): Characteristics and quantitative results for epidemiologic
	cohort study of exposure to 1,2,4-TMB. Lee et al. (2005)
RESULTS
Exposure Subgroup
•	Exposed workers showed significant alterations to symbol digit distribution, dominant hand finger tap rate, and
non-dominant hand finger tap rate.
•	Work duration was also found to influence symbol digit substitution
Observation
Results of Neurobehavioral Test of Study Subjects
Unadjusted Mean ±Std Dev
Adjusted3 Mean (S.E.)
Painters
Controls
p-value
Painters
Controls
p-value
Simple Reaction
Time
297.2±70.0
292.2±95.0
0.671
296.0(5.9)
295.8 (10.9)
0.992
Symbol Digit
Substitution
3233.2±998.9
2,693.8±711.8
0.000
3,156.6
(67.7)
2,691.6
(124.3)
0.000
Finger tap speed
DHb
62.6±8.2
66.4±9.7
0.000
63.0 (0.6)
65.5 (1.2)
0.046
Finger tap speed
NDHC
55.9±8.0
60.2±9.7
0.000
56.1 (0.7)
60.3 (1.2)
0.003
Observation
Neurobehavioral Test Results by Duration of Work, Adjusted for Age and Education
<10 Working Years (S.E.)
n = 48
10-20 Working Years (S.E.)
n = 41
>20 Working Years (S.E.)
n = 91
Simple Reation
Time
297.8 (20.4)
297.9 (11.2)
292.3 (11.6)
Symbol Digit
Substitution
2,972.1 (282.5)
3,033.8(155.1)
3,452.4 (160.7)*
Finger Tap
Speed DH
64.8 (2.3)
63.9 (1.3)
61.3 (1.3)**
Finger Tap
Speed NDH
57.6 (2.4)
56.3 (1.3)
55.2 (1.3)
aAdjusted for age and education
bFinger tapping speed of dominant hand
cFinger tapping speed of non-dominant hand
*, **p< 0.05, p = 0.052
Source: Lee et al. (2005).
This document is a draft for review purposes only and does not constitute Agency policy.
B-67	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-21. Characteristics and quantitative results for epidemiologic cross-sectional
study of exposure to 1,2,4-TMB; Norseth et al. (1991)
Study (location)
Outcome assessment
• Norway
•	Symptoms were recorded via a standard
questionnaire on the last day of monitoring.
•	Monitoring of organic compounds was conducted
for 5 days in workers who were divided into subsets
based on their level of exposure.
•	Asphalt, weather, and traffic density data was
recorded daily.
POPULATION CHARACTERISTICS
Exposed population
Referent or control description
•	In the first group, 79 workers were divided into
groups of 5 or 6 based on their exposure level.
•	A second group of 254 (of which the initial group of
79 was representative) workers completed
questionnaires about symptoms.
• A group of 247 maintenance workers who were not
exposed to asphalt. The group was given a
questionnaire similar to the exposed group.
Exposure assessment
Statistical analysis
•	Mean concentration of 1,2,4-TMB was 0.015 ppm
(0.074 mg/m3), with range between 0 and 0.122 (0 -
0.60 mg/m3) ppm.
•	Mean concentration of 1,3,5-TMB was 0.0014 ppm
(0.0069 mg/m3), with range between 0 and 0.011 (0
-0.054 mg/m3) ppm.
•	Exposure duration: Not reported; measurements
represent the means of five days of monitoring.
•	Exact two-sided Fisher-Irving test was used to
analyze differences in symptom frequency.
•	Mean difference between groups calculated via
two-sided Wilcoxon rank-sum test with a
significance level of 5%.
•	Spearman's correlation coefficient used to estimate
correlation between symptoms and possible
confounders.
This document is a draft for review purposes only and does not constitute Agency policy.
B-68
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-21 (Continued): Characteristics and quantitative results for epidemiologic
cross-sectional study of exposure to 1,2,4-TMB; Norseth etal.
	(1991)	
RESULTS
Exposure subgroup
•	An increase in number of several symptoms was associated with asphalt exposure when asphalt-exposed road
workers were compared with workers not exposed to asphalt.
•	1,2,4-TMB was found to increase number of symptoms, while no similar correlation was found for 1,3,5-TMB.
Effect estimates3
Observation
Symptoms associated with asphalt exposure in exposed and
non-exposed groups of workers*
Days with
symptom
Asphalt workers
(n = 79)
Asphalt workers
(n = 254)
Non-asphalt
workers (n = 247)

Symptoms of asphalt exposure
Abnormal fatigue
None
64.6
75.2
84.6
1-2
21.5
14.6
9.7
3-5
13.9
10.2
5.7
Reduced appetite
None
86.1
89.8
95.1
1-2
12.7
7.5
4.1
3-5
1.3
2.8
0.8
Laryngeal/pharyngeal
irritation
None
63.3
74.0
83.0
1-2
21.5
15.4
11.7
3-5
15.2
10.6
5.3
Eye irritation
None
54.4
68.9
85.4
1-2
22.8
22.4
10.5
3-5
22.8
8.7
4.1
Other, unspecified symptom
None
91.1
85.4
92.3
1-5
8.9
14.6
7.7
aFor correlation between symptom sum and 1,2,4-TMB exposure, r = 0.31, p<0.01.
*AII differences between asphalt workers (n = 254) and non-asphalt workers (n = 247) were statistically significant (p<0.05).
Source: Norseth et al. (1991)
This document is a draft for review purposes only and does not constitute Agency policy.
B-69	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-22. Characteristics and quantitative results for epidemiologic cross-sectional
study of exposure to 1,2,4-TMB Sulkowski et al. (2002)
Study (location)
Outcome assessment
• A factory in which paints and varnishes are
produced
•	Hearing examinations were carried out in an
"audiobus," a motor vehicle equipped with
soundproof cabin and diagnostic tools.
•	Several tests were conducted on subjects, including
air and bone pure tone audiometry, impedance
audiometry with tympanometry, acoustic reflex
threshold measurement, and otoacoustic emissions.
•	Electronystagmographic tests were conducted in an
outpatient clinical setting.
POPULATION CHARACTERISTICS
Exposed population
Referent or control description
•	61 factory workers in direct contact with solvent
vapors.
•	Job titles included resin synthesis analyzers, dry
component mixers, mill operators, dispenser
operators, colorists, and product packers.
• 40 non-exposed workers from the same factory.
Exposure assessment
Statistical analysis
•	Data on exposure was collected from 61 workers
who wore passive dosimeters on 3 work days.
•	Average Exposure duration: 15.8±9.1 years.
• Statistical methods utilized included student t-test,
calculation of means, and linear regression analysis.
RESULTS
Exposure Subgroup
•	47.5% of exposed individuals and 5% of the control population exhibited symptoms of vestibular dysfunction,
as indicated by decreased duration, amplitude, and slow-phase angular velocity of induced nystagmus.
•	High frequency hearing loss as indicated by pure tone audiometry was detected in 42% of exposed individuals
versus 5% of the control population.
This document is a draft for review purposes only and does not constitute Agency policy.
B-70	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
B.5. ANIMAL TOXICOLOGY STUDIES
Table B-23. Characteristics and quantitative results for Baettig et al. (1958)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Rats
M
8 rats per
dose
i.p. injection
0, 200, 500, and 1,700 ppm
(0, 984, 2,460, 8,364
mg/m3) TMB mixture.
4 mos; 8 hrs/day, 5/weeks
Additional study details
•	Mixture of 1,2,4-, 1,2,3-, and 1,3,5-TMB were tested for their effects on growth, (as measured by body weight),
behavior, food intake, red blood cell count, and hemoglobin concentration, and various histological parameters.
•	Rat behavior was assessed qualitatively.
•	TMB mixture (i.e., Fleet-X DV-99) was the same as assessed in the occupational exposure study.
•	Study was translated from German to English prior to receipt by EPA.
Figure 2. Effect of long-term exposure to trimethylbenzene (about 1,700 ppm [8,364 mg/m3]) on the growth of rats.
Open circles: Average body weights of the exposed rats. Closed circles: Average weights of the control rats. Hatched
[and dotted] area[s]: Double square deviation from the mean values plotted.
340
330
o> 320
¦- 310
I) 300
o 290
280
270
a>
o
(0
o 260
>
<
250
240
230
220
| Control rats
Exposed rats
Days of Exposure
_L_L
1 Dec 1 Jan 1 Feb
Dates in Treatment
1 Mar
Source: Reproduced with permission of Springer-Verlag (Battig et al., 1958)
This document is a draft for review purposes only and does not constitute Agency policy.
B-71
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-23 (Continued): Characteristics and quantitative results for Baettig et al.
	(1958)	

Cl>
+j
O
o
Q_
E
>*
CD
O
»—
CD
Q-
90
80
70
60
50
40
30
20
10
$ p = 0.05
p = 0.05
Exposed rats
Control rats
Figure 3. Behavior of
the relative number of
lymphocytes in
trimethylbenzene-
exposed rats
(exposure: about
1,700 ppm
[8,364 mg/m3]).
Source: Reproduced with
permission of Springer-
Verlag (Battig et al..
1958)
Days of exposure
!¦¦¦¦ ¦ ¦
Date: 1 Nov 1 Dec 1 Jan 1 Feb 1 Mar
1 April
Month
Number of days exposed
per month
Average daily food intake (g/lOOg bw
per month)
Difference
(absolute)
Difference
(%)
Control Rats
Exposed Rats
November
5
5.32
2.42
-3.10
-56.13
December
14
5.46
5.07
-0.93
-7.16
January
20
5.19
6.16
+0.97
+15.60
February
17
4.80
5.46
+0.66
+12.09
March
15
4.73
4.80
+0.07
+1.46
April
13

4.32


Table 1. Average intake of food by the rats during experimental exposure to TMB mixture
Source: Reproduced with permission of Springer-Verlag (Battig et al.. 1958)
This document is a draft for review purposes only and does not constitute Agency policy.
B-72	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-23 (Continued): Characteristics and quantitative results for Baettig et al.
	(1958)	

a>
o
o
CD
a.
o
Cl>
a>
o
cd
D_
60
50
40
30
20
10
I | Exposed rats
p = 0.05
Days of exposure
Date: 1 Nov 1 Dec 1 Jan 1 Feb 1 Mar 1 April
Figure 4. Behavior of the relative number of neutrophil leukocytes in trimethylbenzene exposed rats (exposure:
about 1,700 ppm [8,364 mg/m3]).
Source: Reproduced with permission of Springer-Verlag (Battig et al., 1958)
Month
Number of days exposed
per month
Average intake of
drinking water (g/lOOg bw
rat/month)
Difference
(absolute)
Difference
(%)
Control rats
Exposed rats
November
5
9.21
10.55
+1.34
+12.70
December
14
9.71
17.18
+7.47
+43.47
January
20
9.38
22.31
+12.93
+57.91
February
17
7.78
15.92
+8.14
+51.13
March
15
7.12
14.16
+7.04
+49.70
April
13

15.66


Table 2. Average intake of drinking water by rats during experimental exposure to TMB.
Source: Reproduced with permission of Springer-Verlag (Battig et al., 1958)
This document is a draft for review purposes only and does not constitute Agency policy.
B-73	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-23 (Continued): Characteristics and quantitative results for Baettig et al.
	(1958)	

1.08

1.07
O)
CD
1.06
O
1.05
>*
(Q
1.04
TO

<_>
1.03
<_>
d>
<—>

CO
o o
-»¦ ro
1	2	3	4	5
Time in hours after dilution test
Figure 5. Specific gravity of spontaneous and dilution urines in TMB-exposed rats (exposure: about 1,700 ppm
[8,364 mg/m3]).
Source: Reproduced with permission of Springer-Verlag (Battig et al., 1958)
This document is a draft for review purposes only and does not constitute Agency policy.
B-74	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-23 (Continued): Characteristics and quantitative results for Baettig et al.
Urinary
phenol
fraction
Intensity of
exposure
(ppm)
Duration of
exposure
(days)
Duration of exposure, in
days to significant increase
of phenol excretion
Time in days to normalization of
phenol excretion after
discontinuation of exposure
Total
1700
15
4
10
Free
1700
15
8
3
Bound
1700
15
4
9
Total
500
21
8
6
Free
500
21
8
1
Bound
500
21
21
1
Total
200
10
10
1
Free
200
10
10
1
Bound
200
10
Not increased
-
Table 3. Effect of TMB inhalation on urinary phenol excretion in the rat.
Source: Reproduced with permission of Springer-Verlag (Battig et al., 1958)
Health Effect at LOAEL
NOAEL
LOAEL
Increased urinary excretion of free
and total phenols
0 ppm
200 ppm (984 mg/m
Comments: Battig et al. (1956) is published in German. However, Baettig et al. (1958) presents an English-translation of the results
originally presented in Battig et al. (1956). As such, a separate study summary table is not provided for Battig et al. (1956). or of the
eight rats in the long-term inhalation experiment died and were subsequently replaced within the first 2 weeks. Behavioral changes
were assessed qualitatively. The substance to which rats were exposed was comprised of a mixture of all three TMB structural
isomers and may have also contained methylethylbenzene structural isomers. Authors make a statement implying that dose was not
consistent throughout experiment.
This document is a draft for review purposes only and does not constitute Agency policy.
B-75	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-24. Characteristics and quantitative results for Gralewicz etal. (1997b)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
15 rats
per dose
Inhalation (6 hr/day,
5 days/week)
0, 25, 100, or 250 ppm
(0, 123, 492, or 1,230
mg/m3) 1,2,4-TMB
4 weeks
Additional study details
•	Animals were exposed to 1,2,4-TMB in 1.3 m3 dynamic inhalation exposure chambers for 6 hrs/day, 5 days/week
for 4 weeks. Food and water was provided ad libitum.
•	Animals were randomized and assigned to the experimental groups.
•	Rats were tested with a variety of behavioral tests, including radial maze performance, open field activity, passive
avoidance, active two-way avoidance, and shock-induced changes in pain sensitivity.
•	Tests were performed on days 14-54 following exposure.
•	Rats displayed decreased performance on several tests at the 100 ppm and 250 ppm (492 and 1,230 mg/m3)
exposure levels.
120
lOO
BO
60
40
20
O
20
i a
> 1 6
1 4
1 2
10
8
6
4
2
O
7
6
f 5
0
2	4
cn
° 3
«?
1	^
3
*= 1
o
TMBO TMB25 TMBIOO TMB25Q
Figure 1. A comparison of spontaneous locomotor
(upper diagram), exploratory (middle diagram, and
grooming (lower diagram) activity of rats in an open
field during a 5-min observation period.
The test was performed 25 days after a 4-week exposure to
TMB. The bars represent group means and SE (n = 15 for each
group). *p<0.05 compared with TMBO group (0 ppm control
group).
Source: Gralewicz et al. (1997b)
TMBO TMBZ5 TMB100 TMB250
TMBO TMB25 TMBIOO TMB250
This document is a draft for review purposes only and does not constitute Agency policy.
B-76
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-24 (Continued): Characteristics and quantitative results for Gralewicz et al.
	("1997b")	
160 r
1 AO
120
100
2 80
-	trial 1
ITU - trial 2
I " \ — trial 3 (shack)
-	trial 4
HH - trial 5
-	trial 6
Figure 2. Diagrams illustrating the effect of a
4-week exposure to 1,2,4-TMB on the step-
down passive avoidance learning in rats.
The test was performed on days 35-45 after
exposure. Trials 1, 2, and 3 were performed at 24-hr
intervals. The step-down response was punished by
a 10-s foot shock only in trial 3. Trials 4, 5, and 6
were performed 24 hr, 3 days, and 7 days after trial
3, respectively. The maximum step-down latency
was 180 s. The bars represent group means and SE
(n = 15 for each group). ***p<0.001 compared with
respective data from group TMBO (0 ppm control
group).
Source: Gralewicz et al. (1997b)
TMBO
TMB25
TMB100
TM8250
This document is a draft for review purposes only and does not constitute Agency policy.
B-77	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-24 (Continued): Characteristics and quantitative results for Gralewicz et al.
	("1997b")	
60
50
>. 40
0
c
*
° 30
1	20
i
J
p 10

X

I
I
~
K3

-	L1
-	L2
TMBO TMB25 TMB100 TMB250
Figure 3. Hot plate behavior tested in rats on day 50
(trials 1 and 2) and day 51 (trial 3) after 4-week
exposure to 1,2,4-TMB. Bars represent group means
and SE (n = 15 for each group).
Upper diagram: a comparison of the latency of the paw-lick
response to a thermal stimulus (54.5°C) on day 50. LI: paw-
lick latency in trial 1 performed before a 2 min intermittent
foot shock. L2: paw-lick latency in trial 2 performed several
seconds after the foot shock. ***p<0.001 compared with LI
in the same group.
o
o
300
250
200
«> 150
3
O
>
100
50
***
_L
JL
Lower diagram: A comparison of the change in the paw-lick
latency noted 24 hrs after foot shock (trial 3). ***p<0.001,
**p<0.01 when compared to TMBO (0 ppm control group).
Source: Gralewicz et al. (1997b)
TMBO TMB25 TMB10D TMB250
This document is a draft for review purposes only and does not constitute Agency policy.
B-78	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-24 (Continued): Characteristics and quantitative results for Gralewicz et al.
	("1997b")	
K
E
o
CL
o
c
o
~o
—	block 3
I I —	block A
k\N -	block 5
—	block 6
Figure 4. A comparison of the
active avoidance performance
increment during a single 30-trial
training session in consecutive
groups of rats.
The testing was performed on day 54
after 4-week exposure to 1,2,4-TMB.
Bars represent the percentage (group
mean and SE, n = 15 for each group) of
avoidance response in successive five-
trial blocks. No avoidance response
was noted in any group during the first
10 trials and therefore blocks 1 and 2
were omitted in the analysis.
Source: Gralewicz et al. (1997b)
TMBO
TMB25
TMB100
TMB250
Health Effect at LOAEL
NOAEL
LOAEL
Open field grooming
significantly increased, lower
than expected step down
latency
25 ppm (123 mg/m
100 ppm (492 mg/m
Comments: CNS disturbances were observed up to 2 months after termination of exposure, indicating the persistence of effects
after the metabolic clearance of 1,2,4-TMB from the test animals. Duration of exposure only 4 weeks. Generally, short-term
exposure studies have limited utility in quantitation of human health reference values.
Source: Gralewicz et al. (1997b)
This document is a draft for review purposes only and does not constitute Agency policy.
B-79	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table B-25. Characteristics and quantitative results for Gralewicz etal. (1997a)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
9 rats per
dose
Inhalation (6 hr/day,
5 days/week)
0, 25, 100, or 250 ppm
(0, 123, 492, or 1,230
mg/m3) 1,2,4-TMB
4 weeks
Additional study details
•	Animals were exposed to 1,2,4-TMB in 1.3 m dynamic inhalation exposure chambers for 6 hrs/day, 5 days/week
for 4 weeks. Food and water was provided ad libitum.
•	Animals were randomized and assigned to the experimental groups.
•	Rats were tested to determine whether exposure to 1,2,4-TMB altered the pattern of occurrence of spike wave
discharges (SWD).
•	Rats exposed to 1,2,4-TMB at 100 or 250 ppm (492 or 1,230 mg/m3) did not show an increase in SWD activity. Rats
exposed to 0 or 25 ppm (0 or 123 mg/m3) 1,2,4-TMB showed progressively decreasing levels of SWD activity.
o
TM BO
TMB100
TMB250
25
~ 20
% 15
10
Y//.A before exposure
3-4 h after exp.
30 days alter exp.
lTTH 120 days after exp,
TMRO TMB25 TMB100 TMB250
Figure 1. Diagrams showing the effect
of a 4-week inhalation exposure to
1,2,4-TMB on the contribution of
transitional (upper diagram, high
arousal (middle diagram), and slow-
wave sleep (lower diagram)) states in
the rat EEG during successive 1-hour
recording periods.
The bars represent group means and SE.
Source: Gralewicz et al. (1997a)
TMBO TMB25 TMB 100 TMB250
This document is a draft for review purposes only and does not constitute Agency policy,
B-80
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table B-25 (Continued): Characteristics and quantitative results for Gralewicz et al.
	("1997a")	
3
O
XL

45
40
<
a: 35
30
25
20
15
10

TMB0 TMB25 TMB100 TMB250
Health Effect at LOAEL
IMOAEL
LOAEL
Decreased spike-wave discharges
25 ppm (123 mg/m
100 ppm (492 mg/m
Comments: CIMS disturbances were observed up to 4 months after termination of exposure, indicating the persistence of effects
after the metabolic clearance of 1,2,4-TMB from the test animals. Duration of exposure only 4 weeks. Generally, short-term
exposure studies have limited utility in quantitation of human health reference values.
Source: Gralewicz et al. (1997a)	
This document is a draft for review purposes only and does not constitute Agency policy,
B-81	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-26. Characteristics and quantitative results for Gralewicz and Wiaderna
(2001)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
10 or 11
rats per
dose
Inhalation (6 hr/day,
5 days/week)
0 or 100 ppm (0 or 492
mg/m3) 1,2,3-, 1,2,4-, or
1,3,5-TMB
4 weeks
Additional study details
•	Animals were exposed to 1,2,3-, 1,2,4- or 1,3,5-TMB in 1.3 m3 dynamic inhalation exposure chambers for 6 hrs/day,
5 days/week for 4 weeks. Food and water was provided ad libitum.
•	Animals were randomized and assigned to the experimental groups.
•	Rats were tested with a variety of behavioral tests, including radial maze performance, open field activity, passive
avoidance, active two-way avoidance, and shock-induced changes in pain sensitivity.
•	Tests were performed starting 2 weeks post-exposure.
•	1,2,3-, 1,2,4-, and 1,3,5-TMB-exposed rats showed alterations in performance in spontaneous locomotor activity,
passive avoidance learning, and paw-lick latencies.
1.8
1.6
1.4
1.3
1.0
0.8
0.6
0.4
0.2
0.0
|— day 5
Control XYL
MES HM
&
B
£
TO- day 1
j==j- day 3 Figure 1. Radial maze performance of rats exposed for 4 weeks to
day 4 m-xylene or a TMB isomer at a concentration of 100 ppm (492
mg/m3).
The test (one trial a day) was performed on days 14-18 after exposure. The
diagrams illustrate the number of perseveration (upper diagram) and
omission (lower diagram) errors in successive daily trials.
Denotation:
Control- sham exposed group (n=10),
XYL- m-xylene exposed group (n=ll),
PS- 1,2,4-TMB exposed group (n=ll),
MES- 1,2,3-TMB exposed group (n=ll),
HM- hemimellitene exposed group (n=ll).
Bars represent group means and SE.
Source: Gralewicz and Wiaderna (2001)
Control XYL
This document is a draft for review purposes only and does not constitute Agency policy.
B-82
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table B-26 (Continued): Characteristics and quantitative results for Gralewicz and
	Wiaderna (2001)	
on
u
<0
s
3
55
120
100
80
60
40
20
p<0.05 compared
to control
p<0.05 compared
to PS and MES
Figure 2. A comparison of open-
field locomotor activity in sham-
exposed and solvent-exposed rats.
The test was performed on day 25
after a 4-week exposure to m-
xylene or a TMB isomer at
concentration of 100 ppm (492
mg/m3).
Bars represent group means and SE.
Source: Gralewicz and Wiaderna (2001)
Control XYL PS MES HM
Figure 3. Diagram illustrating the effect of a 4-week inhalation exposure to m-xylene or a TMB isomer at
concentration of 100 ppm (492 mg/m ) on the step-down response latency in the passive avoidance test.
m
TJ
c
o
o
M
160
140
120
100
80
c
#
0
¦o
1
CL
60
40
20
rrm-
$
trial
trial
trial
trial
trial
(shock)
- trial 6
p<0.05 compared
to control
The test was performed on days 39-48
after exposure. Trials 1, 2, and 3 were
performed at 24 hr intervals. The step-
down response was punished by a 10 s
footshock in trial 3 only. Trials 4, 5, and
6 were performed 24 hr, 3 days, and 7
days after trial 3, respectively. The
maximum time of staying on the
platform was 180 s. Bars represent
means and SE.
Source: Gralewicz and Wiaderna
(2001)
Control XYL
PS
MES
HM
This document is a draft for review purposes only and does not constitute Agency policy,
B-83	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
p<0.05 compared
to control
Table B-26 (Continued): Characteristics and quantitative results for Gralewicz and
Wiaderna (2001)
Figure 4. A comparison of sham-exposed and
solvent-exposed rats with respect to the latency
of the paw-lick response to heat (54.5°C) before
(LI), several seconds after (L2), and 24 hr after a
2 min intermittent footshock.
The test was performed on days 50 and 51 after a 4-
week inhalation exposure to m-xylene or a TMB
isomer at a concentration of 100 ppm (492 mg/mB).
Bars represent group means and SE.
Source: Gralewicz and Wiaderna (2001)
20
i
i
o
0.
Control XYL
45
40
c
O
« 35
-p
"k—
u
-------
Toxicological Review ofTrimethylbenzene
Table B-27. Characteristics and quantitative results for Janik-Speichowicz et al.
(1998)
Study Design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Balb/c Mice
M &
F
4 or 5
mice/
dose
group
i.p. injection
0,1470, 2160, and 2940
mg/kg body weight
Single exposure, or 2 i.p.
injections spaced out over 24
hours
Additional study details



• Animals were given one or two injections of i.p. injections of 1,2,3-TMB.

• Animals were randomized and assigned to the experimental groups.

• Most deaths occurred within the first 2 days following single injections.

• LD50 was determined to be 3,670 mg/kg for males and 2,700 mg/kg for females.

• Micronuclei and chromatid exchange assays were conducted on extracted bone marrow to assess genotoxicity.
• Multiple indicators of genotoxicity were used, giving adequate evidence to assess the genotoxic potential of acute
exposure to 1,2,4-TMB, 1,2,3-TMB, and 1,3,5-TMB .
This document is a draft for review purposes only and does not constitute Agency policy.
B-85
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-27 (Continued): Characteristics and quantitative results for
	Janik-Speichowicz et al. (1998)	
Figure 2, Dose-related increase in the number of His+ revertants for 1,2,3-TMB in S. typhimurium strains
TA102
TA100
TA98
TA97a
dose [>ul/plate]
-S9 +S9
lliii
» : :: :



¦ <5 » '• *» -
Solvent
control

1200	1000 800	600	400	200
-S9	Revertants / plate
200
400
+S9
- mutagenic effect {a 2-fold or greater increase in the number of revertants
per plate, as compared with the solvent control number)
Spontaneous revertants: TA97a 129±10 (-S9); 141±17 (+S9);
TA98	23 + 2 (-S9) ; 35±6 (+S9) ;
TA100 126±4 (-S9) ; 119+5 (+S9);
TA102 282±33 (-S9); 315±32 (+S9)
Source: Janik-Speichowicz et al. (1998)
Observation
Exposure to 1,2,4-TMB (|jg or |iL)
100
(Solvent
control)
10
20
30
TA97a (-S9)
121±7
126±13
148±23
158±10
165±8
141±25
115±3
TA97a (+S9)
145±5
141±12
152±7
168±8
176±21
155±20
106±7
TA98 (-S9)
24±3
23±3
24±3
29±5
41±7
27±8
TOX
TA98 (+S9)
31±3
31±5
35±4
28±1
29±4
30±3
29±6
TA100(-S9)
TA100(+S9)
TA102(-S9)
TA102(+S9)
123±71
25+4
258±6
294±11
125±41
21±10
280±12
315±14
138±15
126±62
290±33
279±24
148±18
125±5
262±16
276±11
143+9
112±4
273±20
276±11
124±7
108+3
214±8
236±32
118±4
110±4
TOX
TOX
This document is a draft for review purposes only and does not constitute Agency policy,
B-86	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-27 (Continued): Characteristics and quantitative results for
	 Janik-Speichowicz etal. (1998)	

Exposure to 1,3,5-TMB (|ig or |iL)
Observation
0
100
(Solvent
control)
1
5
10
20
30
40
TA97a (-S9)
127±15
131±10
141±13
149±29
139±17
129±13
125±8
NTb
TA97a (+S9)
183±6
157±19
180±26
196±16
155±30
137±29
138±20
128±11
TA98 (-S9)
22±4
22±4
27±3
28±5
25±2
37±5
23±5
TOX
TA98 (+S9)
30±3
32±5
31±4
35±5
31±2
39±5
28±2
31±1
TA100(-S9)
138±13
143±15
143±4
152±8
140±26
154±14
130±7
TOX
TA100(+S9)
142±10
138±82
137±3
147±29
139±16
131±10
108±11
115±6
TA102(-S9)
263±23
60±12
268±17
280±19
261±25
238±5
198±2
NT
TA102(+S9)
337±13
336±23
347±34
334±30
353±11
340±37
324±10
NT
Observation
Exposure to 1,2,3-TMB (mg/kg body weight)
0
1470
2160
2940

% of Polychromatic Erythrocytes with Micronuclei (± SD)
Males 30 hr harvest time
--
0.17±0.06
--
0.22±0.07
Males 48 hr harvest time
0.18±009
0.17±0.05
--
0.22±0.10
Males 72 hr harvest time
--
0.17±0.05
--
0.21±0.11
Females 30 hr harvest time
--
--
0.22±0.09
--
Females 48 hr harvest time
0.20±0.08
--
0.20±0.08
--
Females 72 hr harvest time
--
--
0.20±0.14
--

Ratio of polychromatic to normochromatic erythrocytes
Males 30 hr harvest time
--
0.82
--
0.85
Males 48 hr harvest time
0.81
0.45
--
0.72
Males 72 hr harvest time
--
0.50
--
0.62
Females 30 hr harvest time
--
--
0.90
--
Females 48 hr harvest time
0.95
--
0.84
--
Females 72 hr harvest time
--
--
0.78
--
Observation
Exposure to 1,2,4-TMB (mg/kg body weight)
0
2000
3280
4000

% of Polychromatic Erythrocytes with Micronuclei (± SD)
Males 30 hr harvest time
--
0.15±0.10
--
0.23±0.10
Males 48 hr harvest time
0.18±0.07
0.18±0.10
--
0.16±0.8
Males 72 hr harvest time
--
0.20±0.08
--
0.16±0.07
Females 30 hr harvest time
--
--
0.23±0.5
--
Females 48 hr harvest time
0.23±0.05
--
0.18±0.05
--
Females 72 hr harvest time
--
--
0.13±0.05
--
This document is a draft for review purposes only and does not constitute Agency policy.
B-87	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-27 (Continued): Characteristics and quantitative results for
	 Janik-Speichowicz etal. (1998)	

Ratio of polychromatic to normochromatic erythrocytes
Males 30 hr harvest time
--
1.18
--
1.16
Males 48 hr harvest time
0.95
1.02
--
0.74
Males 72 hr harvest time
--
1.02
--
0.68*
Females 30 hr harvest time
--
--
0.98
--
Females 48 hr harvest time
0.95
--
1.01
--
Females 72 hr harvest time
--
--
0.85
--
Observation
Exposure to 1,3,5-TMB (mg/kg body weight)
0
1800
2960
3600

% of Polychromatic Erythrocytes with Micronuclei (± SD)
Males 30 hr harvest time
--
0.20±0.00
--
0.24±0.11
Males 48 hr harvest time
0.21±0.08
0.17±0.09
--
0.17±0.05
Males 72 hr harvest time
--
0.17±0.09
--
0.14±0.05
Females 30 hr harvest time
--
--
0.17±0.09
--
Females 48 hr harvest time
0.20±0.08
--
0.20±0.00
--
Females 72 hr harvest time
--
--
0.22±0.05
--

Ratio of polychromatic to normochromatic erythrocytes
Males 30 hr harvest time
--
0.62
--
0.40*
Males 48 hr harvest time
0.61
0.56
--
0.33
Males 72 hr harvest time
--
0.58
--
0.42*
Females 30 hr harvest time
--
--
0.51
--
Females 48 hr harvest time
0.60
--
0.60
--
Females 72 hr harvest time
--
--
0.58
--
This document is a draft for review purposes only and does not constitute Agency policy.
B-88	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-27 (Continued): Characteristics and quantitative results for
	Janik-Speichowicz etal. (1998)	
CD
U
LLI
O
tn

0
Y
solvent
control
O hemimellitene -
O pseudocumene
V mesitylene
730
-1470 -
900
900
2200
I'
J- x
I
C
t i
t
-- y
2940
T
0
X
j
e
1
t
1800
1800
2700
2700
3600 dose
3600 [mg/kg b.w.]
- significant difference vs. control at p<0.05; Mitomycin C administered
as positive control in dose of 2.0 mg/kg b.w. gave 21.34 ± 1.36 SCE/cell
Figure 3. Sister chromatid exchanges induced in bone marrow cells of lmp:Balb/c mice.
Source: Janik-Speichowicz et al. (1998)
Health Effect at LOAEL
NOAEL
LOAEL
Significant increase in SCE
induction relative to control
0 mg/kg
730 mg/kg
Comments: Multiple indicators of genotoxicity were investigated, giving adequate evidence to assess the genotoxic potential of
acute exposure to 1,2,4-TMB, 1,2,3-TMB, and 1,3,5-TMB. Exposures were acute (occurring within 24 hours) and therefore less
germane to study of health effects resulting from chronic exposure. For 1,2,3-TMB, sister chromatid assays were conducted at
concentrations differing from the other independent variables (1,2,4- and 1,3,5-TMB). It is also difficult to establish a dose-response
relationship for micronucleus formation because there were only two non-control exposure groups in males and only one non-
control exposure group in females.
aTOX = toxic effects (background growth reduced);
bNT = not tested
*Significant difference vs. control at P<0.05
Source: Janik-Speichowicz et al. (1998)
Table B-28. Characteristics and quantitative results for Koch Industries (1995b)
See Next Page (Table B-28 starts on Next Page)
This document is a draft for review purposes only and does not constitute Agency policy.
B-89	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
(1995b)
Study Design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Sprague
Dawley CD
M/F
20
rats/dose
Oral gavage
0, 50, 200, and 600
mg/kg/day 1,3,5-TMB
90 days
Additional study details
•	Rats were treated with 0, 50, 200, or 600 mg/kg/day of 1,3,5-TMB (5 days per week) and were observed daily
for adverse clinical signs
•	Hematology and serum chemistry was analyzed after 30 days, at the end of the exposure period, and after a 28
day recovery period (in an additional 600 mg/kg/day "recovery" group only).
•	No deaths related to 1,3,5-TMB exposure occurred during the study.
•	Cumulative weight gain decreased by approximately 11% in the high-dose male group.
•	High dose females exhibited an increase in absolute and relative liver weight, while males in the same dose
group showed increases in relative liver weight.
•	The NOEL was 200 mg/kg
Mean body weight after 90 day 1,3,5-TMB dosing period
Males
Dose (mg/kg/day)
0
50
200
600
Mean
624
607
602
585
Standard Deviation
48.2
62.0
40.8
66.4
No. of Rats
10
10
9
20
Females
Mean
327
335
334
330
Standard Deviation
24.8
37.6
21.2
29.3
No. of Rats
10
10
10
20
Mean clinical chemistry parameters, terminal and recovery in males
Parameter
Dose (mg/kg/day)
0
50
200
600
600
(recovery)
Na-mean
142.4
142.7
143.0
142.4
141.6
Na-standard
deviation
1.49
0.65
1.40
1.32
1.30
Na-number of rats
10
10
9
10
10
K-mean
4.32
4.51
4.37
4.54
4.33
K-standard
deviation
0.397
0.339
0.328
0.270
0.240
K-number of rats
10
10
9
10
10
Cl-mean
105.3
105.3
106.0
106.2
104.7






This document is a draft for review purposes only and does not constitute Agency policy.
B-90	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
		("1995b")	

Dose (mg/kg/day)

0
50
200
600
600
(recovery)
Cl-standard
deviation
2.59
2.33
1.72
2.18
0.88
Cl-number of rats
10
10
9
10
10
CK-mean
594
962
934
595
884
CK-standard
deviation
340.4
929.8
799.2
389.1
353.4
CK-number of rats
10
10
9
10
10
ALK P-mean
107
112
121
156*
77
ALK P-standard
deviation
28.1
26.5
33.7
56.2
20.5
ALK P-number of
rats
10
10
9
10
10
ALT-mean
29
30
25
33
25
ALT-standard
deviation
6.4
9.8
7.0
9.1
4.4
ALT-number of rats
10
10
9
10
10
AST-mean
72
91
86
85
89
AST-standard
deviation
18.9
31.9
25.5
25.0
16.7
AST-number of rats
10
10
9
10
10
GGT-mean
3
2
2
2
1
GGT-standard
deviation
0.9
0.9
1.0
1.0
1.5
GGT-number of rats
10
10
9
10
10
BUN-mean
11.8
12.3
12.3
11.5
13.5
BUN-standard
deviation
1.45
1.87
1.22
1.30
1.53
BUN-number of
rats
10
10
9
10
10
CREA-mean
0.42
0.43
0.42
0.47
0.48
CREA-standard
deviation
0.092
0.079
0.110
0.065
0.067
CREA-number of
rats
10
10
9
10
10
T PRO-mean
6.0
5.9
6.0
6.1
6.0
T PRO-standard
deviation
0.38
0.24
0.31
0.42
0.25
This document is a draft for review purposes only and does not constitute Agency policy.
B-91	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
		("1995b")				
T PRO-number of
rats
10
10
9
10
10
ALB-mean
3.6
3.6
3.7
3.8
3.7
ALB-standard
deviation
0.23
0.19
0.19
0.22
0.09
ALB-number of rats
10
10
9
10
10
GLOB-mean
2.4
2.3
2.3
2.3
2.3
GLOB-standard
deviation
0.27
0.18
0.16
0.24
0.24
GLOB-number of
rats
10
10
9
10
10
A/G Ratio-mean
1.6
1.6
1.6
1.7
1.7
A/G Ratio-standard
deviation
0.19
0.17
0.11
0.15
0.17
A/G Ratio-number
of rats
10
10
9
10
10
GLU-mean
1.02
134.6
136.9
121.1*
168.4
GLU-standard
deviation
22.80
15.11
15.76
13.14
26.39
GLU-number of rats
10
10
9
10
10
CHOL-mean
38.2
33.1
31.6
45.3
35.3
CHOL-standard
deviation
6.83
9.13
9.93
15.99
10.10
CHOL-number of
rats
10
10
9
10
10
Ca-mean
10.2
10.2
10.2
10.2
9.9
Ca-standard
deviation
0.22
0.29
0.37
0.23
0.24
Ca-number of rats
10
10
9
10
10
PHOS-mean
6.5
6.7
7.0
7.6*
5.8
PHOS-standard
deviation
0.64
0.80
0.68
0.58
0.59
PHOS-numberof
rats
10
10
9
10
10
TBIL-mean
0.4
0.4
0.5
0.5
0.5
TBIL-standard
deviation
0.12
0.10
0.09
0.14
0.09
TBIL-number of rats
10
10
9
10
10


This document is a draft for review purposes only and does not constitute Agency policy.
B-92	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
	("1995b")	
Mean clinical chemistry parameters, terminal and recovery in females
Parameter
Dose (mg/kg/day)
0
50
200
600
600
(recovery)
Na-mean
142.1
141.6
141.7
138.9*
140.9
Na-standard
deviation
1.10
0.96
2.07
2.83
1.47
Na-number of rats
10
10
10
10
10
K-mean
3.94
4.13
4.01
3.86
4.06
K-standard
deviation
0.195
0.200
0.119
0.292
0.259
K-number of rats
10
10
10
10
10
Cl-mean
105.9
106.2
106.1
103.0*
107.0
Cl-standard
deviation
2.32
1.63
1.05
3.81
1.68
Cl-number of rats
10
10
10
10
10
CK-mean
404
574
381
362
532
CK-standard
deviation
172.6
346.4
228.3
242.5
369.7
CK-number of rats
10
10
10
10
10
ALK P-mean
59
57
55
78
38
ALK P-standard
deviation
14.8
10.3
14.9
24.5
10.1
ALK P-number of
rats
10
10
10
10
10
ALT-mean
21
22
23
24
27
ALT-standard
deviation
2.3
4.0
7.3
4.1
7.1
ALT-number of rats
10
10
10
10
10
AST-mean
60
75
62
60
77
AST-standard
deviation
16.5
18.6
15.2
15.0
21.4
AST-number of rats
10
10
10
10
10
GGT-mean
2
3
3
3
2
GGT-standard
deviation
1.1
1.6
1.0
1.4
1.4
GGT-number of rats
10
10
10
10
10
BUN-mean
14.5
14.0
11.9
13.5
16.2
BUN-standard
deviation
1.34
2.57
1.49
4.61
2.31
This document is a draft for review purposes only and does not constitute Agency policy.
B-93	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
		("1995b")				
BUN-number of
rats
10
10
10
10
10
CREA-mean
0.53
0.51
0.53
0.56
0.55
CREA-standard
deviation
0.106
0.085
0.099
0.110
0.099
CREA-number of
rats
10
10
10
10
10
T PRO-mean
6.2
6.3
6.6
6.5
6.3
T PRO-standard
deviation
0.44
0.41
0.69
0.68
0.66
T PRO-number of
rats
10
10
10
10
10
ALB-mean
4.1
4.3
4.5
4.5
4.3
ALB-standard
deviation
0.29
0.36
0.58
0.56
0.51
ALB-number of rats
10
10
10
10
10
GLB-mean
2.1
2.0
2.1
2.1
2.0
GLB-standard
deviation
0.21
0.17
0.19
0.20
0.18
GLB-number of rats
10
10
10
10
10
A/G Ratio-mean
2.0
2.1
2.1
2.1
2.1
A/G Ratio-standard
deviation
0.16
0.22
0.26
0.23
0.18
A/G Ratio-number
of rats
10
10
10
10
10
GLU-mean
131.8
136.4
140.1
132.8
150.7
GLU-standard
deviation
7.65
11.72
14.48
15.91
19.18
GLU-number of rats
10
10
10
10
10
CHOL-mean
36.2
35.2
38.8
51.2*
28.7
CHOL-standard
deviation
8.83
6.64
6.24
17.84
12.93
CHOL-number of
rats
10
10
10
10
10
Ca-mean
10.1
10.2
10.4
10.5
10.0
Ca-standard
deviation
0.35
0.24
0.42
0.63
0.36
Ca-number of rats
10
10
10
10
10
PHOS-mean
6.1
6.1
6.4
7.5*
5.3
PHOS-standard
deviation
1.08
1.27
1.18
1.24
0.80
This document is a draft for review purposes only and does not constitute Agency policy.
B-94	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
		("1995b")				
PHOS-numberof
rats
10
10
10
10
10
TBIL-mean
0.5
0.5
0.4
0.5
0.5
TBIL-standard
deviation
0.08
0.10
0.08
0.07
0.07
TBIL-number of rats
10
10
10
10
10
Mean Male Hematology Parameters Terminal and Recovery
Parameter
Dose (mg/kg/day)
0
50
200
600
600
(recovery)
WBC-mean
9.1
8.1
8.1
7.7
7.8
WBC-standard
deviation
2.70
2.50
1.74
1.76
1.24
WBC-number of
rats
10
10
9
10
10
RBC-mean
8.94
8.50
8.98
8.72
8.51
RBC-standard
deviation
0.375
0.483
0.565
0.275
0.423
RBC-number of rats
10
10
9
10
10
HGB-mean
15.6
15.3
15.8
15.4
15.4
HGB-standard
deviation
0.52
0.76
0.77
0.53
0.58
HGB-number of rats
10
10
9
10
10
HCT-mean
43.9
42.2
44.1
43.3
41.6
HCT-standard
deviation
1.65
2.72
2.12
1.60
1.99
HCT-number of rats
10
10
9
10
10
MCV-mean
49.1
49.7
49.2
49.6
49.0
MCV-standard
deviation
1.17
1.09
1.76
1.66
1.62
MCV-number of
rats
10
10
9
10
10
MCH-mean
17.5
18.0
17.7
17.7
18.2
MCH-standard
deviation
0.45
0.73
0.85
0.68
0.61
MCH- number of
rats
10
10
9
10
10
MCHC-mean
35.6
36.3
35.9
35.6
37.1
MCHC-standard
deviation
0.67
1.07
0.60
0.67
0.60
This document is a draft for review purposes only and does not constitute Agency policy.
B-95	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
		("1995b")				
MCHC-number of
rats
10
10
9
10
10
PLT-mean
1092
1098
1041
1125
1083
PLT-standard
deviation
134.1
120.8
100.9
145.9
112.6
PLT-number of rats
10
10
9
10
10
Mean Female Hematology Parameters Terminal and Recovery
Parameter
Dose (mg/kg/day)
0
50
200
600
600
(recovery)
WBC-mean
5.5
5.6
5.4
5.7
4.6
WBC-standard
deviation
2.05
1.53
1.64
1.99
1.55
WBC-number of
rats
10
10
10
10
10
RBC-mean
7.88
8.01
7.90
8.34
7.70
RBC-standard
deviation
0.729
0.354
0.578
0.548
0.423
RBC-number of rats
10
10
10
10
10
HGB-mean
14.8
15.0
15.2
15.3
15.1
HGB-standard
deviation
0.88
0.48
0.82
0.78
0.57
HGB-number of rats
10
10
10
10
10
HCT-mean
41.0
41.4
41.9
43.3
39.9
HCT-standard
deviation
3.15
1.91
2.93
2.33
1.67
HCT-number of rats
10
10
10
10
10
MCV-mean
52.1
51.7
53.0
52.0
51.9
MCV-standard
deviation
1.65
1.18
1.03
1.24
1.33
MCV-number of
rats
10
10
10
10
10
MCH-mean
18.9
18.7
19.2
18.4
19.6
MCH-standard
deviation
0.89
0.67
0.53
0.68
0.78
MCH- number of
rats
10
10
10
10
10
MCHC-mean
36.2
36.2
36.3
35.4
37.7
MCHC-standard
deviation
0.79
0.86
0.83
0.54
0.64
This document is a draft for review purposes only and does not constitute Agency policy.
B-96	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
		("1995b")				
MCHC-number of
rats
10
10
10
10
10
PLT-mean
1094
1089
1011
1053
1008
PLT-standard
deviation
153.3
132.0
97.2
125.7
105.7
PLT-number of rats
10
10
10
10
10
Mean Male Absolute Differential White Blood Cell Counts (Terminal and Recovery)
Parameter
Dose (mg/kg/day)
0
50
200
600
600
(recovery)
NRBC-mean
0
0
0
0
0
NRBC-standard
deviation
0
0
0.7
0
0
NRBC-number of
rats
10
10
9
10
10
MAT NEU-mean
1.8
1.7
1.4
1.5
1.0
MAT NEU-standard
deviation
1.07
1.10
0.36
0.75
0.29
MAT NEU-number
of rats
10
10
9
10
10
LYM-mean
7.1
6.2
6.4
6.0
6.6
LYM-standard
deviation
2.78
2.16
1.59
2.16
1.23
LYM-number of rats
10
10
9
10
10
MONO-mean
0.1
0.2
0.3*
0.2*
0.2
MONO-standard
deviation
0.09
0.09
0.17
0.18
0.10
MONO-number of
rats
10
10
9
10
10
EOS-mean
0.1
0.1
0.0
0.0
0.1
EOS-standard
deviation
0.06
0.09
0.07
0.05
0.07
EOS-number of rats
10
10
9
10
10
BASO-mean
0
0
0
0
0
BASO-standard
deviation
0
0
0
0
0
BASO-number of
rats
10
10
9
10
10
IMM NEU-mean
0
0
0
0
0
IMM NEU-standard
deviation
0
0
0
0
0
This document is a draft for review purposes only and does not constitute Agency policy.
B-97	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
		("1995b")				
IMM NEU-number
of rats
10
10
9
10
10

Mean Female Absolute Differential White Blood Cell Counts (Terminal and Recovery)
Parameter
Dose (mg/kg/day)
0
50
200
600
600
(recovery)
NRBC-mean
0
0
0
0
0
NRBC-standard
deviation
0
0
0
0
0
NRBC-number of
rats
10
10
10
10
10
MAT NEU-mean
0.8
0.7
0.9
1.0
0.7
MAT NEU-standard
deviation
0.48
0.32
0.69
0.39
0.45
MAT NEU-number
of rats
10
10
10
10
10
LYM-mean
4.6
4.7
4.2
4.4
3.7
LYM-standard
deviation
1.93
1.52
1.52
2.08
1.34
LYM-number of rats
10
10
10
10
10
MONO-mean
0.1
0.1
0.1
0.2
0.2
MONO-standard
deviation
0.14
0.10
0.08
0.17
0.11
MONO-number of
rats
10
10
10
10
10
EOS-mean
0.1
0.1
0.1
0.1
0
EOS-standard
deviation
0.07
0.07
0.09
0.09
0.07
EOS-number of rats
10
10
10
10
10
BASO-mean
0
0
0
0
0
BASO-standard
deviation
0
0
0.03
0
0
BASO-number of
rats
10
10
10
10
10
IMM NEU-mean
0
0
0
0
0
IMM NEU-standard
deviation
0
0
0
0
0
IMM NEU-number
of rats
10
10
10
10
10






This document is a draft for review purposes only and does not constitute Agency policy.
B-98	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
	("1995b")	
Mean Male Absolute Organ Weights (g)
Parameter
Dose (mg/kg/day)
0
50
200
600
600
(recovery)
ADR-mean
0.062
0.059
0.058
0.063
0.060
ADR-standard
deviation
0.010
0.015
0.011
0.010
0.008
ADR-number of rats
10
10
9
10
10
BRN-mean
2.25
2.28
2.23
2.19
2.24
BRN-standard
deviation
0.073
0.090
0.094
0.084
0.112
BRN-number of rats
10
10
9
10
10
KID-mean
3.92
3.95
4.10
4.16
4.05
KID-standard
deviation
0.326
0.262
0.610
0.464
0.491
KID-number of rats
10
10
9
10
10
LIV-mean
19.28
18.91
18.38
20.90
17.38
LIV-standard
deviation
1.843
3.074
2.885
3.313
2.222
LIV-number of rats
10
10
9
10
10
LNG-mean
2.19
2.19
2.20
2.06
2.04
LNG-standard
deviation
0.299
0.292
0.134
0.158
0.229
LNG-number of rats
10
10
9
10
10
TESTES-mean
4.15
3.78
4.04
4.00
3.91
TESTES-standard
deviation
0.290
0.595
0.336
0.250
0.612
TESTES-number of
rats
10
10
9
10
10
Mean Female Absolute Organ Weights (g)
Parameter
Dose (mg/kg/day)
0
50
200
600
600
(recovery)
ADR-mean
0.075
0.078
0085
0.082
0.084
ADR-standard
deviation
0.007
0.012
0.013
0.015
0.015
ADR-number of rats
10
10
10
10
10
BRN-mean
2.06
2.06
2.11
2.06
2.11
BRN-standard
deviation
0.080
0.083
0.094
0.050
0.059
This document is a draft for review purposes only and does not constitute Agency policy.
B-99	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
		("1995b")				
BRN-number of rats
10
10
10
10
10
KID-mean
2.34
2.23
2.38
2.51
2.38
KID-standard
deviation
0.314
0.228
0.116
0.264
0.248
KID-number of rats
10
10
10
10
10
LIV-mean
9.44
9.13
10.05
11.78*
9.71
LIV-standard
deviation
1.601
0.774
0.967
1.444
1.411
LIV-number of rats
10
10
10
10
10
LNG-mean
1.63
1.73
1.66
1.60
1.63
LNG-standard
deviation
0.187
0.140
0.106
0.150
0.140
LNG-number of rats
10
10
10
10
10
OVARIES-mean
0.128
0.123
0.122
0.142
0.142
OVARIES-standard
deviation
0.023
0.039
0.042
0.058
0.036
OVARIES-number of
rats
10
10
10
10
9
Mean Male Relative3 Organ Weights (g)
Parameter
Dose (mg/kg/day)
0
50
200
600
600
(recovery)
FBWb-mean
602
584
576
562
595
FBW-standard
deviation
46.4
60.4
40.1
52.2
81.8
FBW-number of
rats
10
10
9
10
10
ADR-mean
0.011
0.010
0.010
0.011
0.010
ADR-standard
deviation
0.002
0.002
0.002
0.001
0.001
ADR-number of rats
10
10
9
10
10
BRN-mean
0.38
0.39
0.39
0.39
0.38
BRN-standard
deviation
0.033
0.032
0.035
0.035
0.044
BRN-number of rats
10
10
9
10
10
KID-mean
0.65
0.68
0.71
0.74*
0.68
KID-standard
deviation
0.052
0.052
0.082
0.045
0.039
KID-number of rats
10
10
9
10
10
LIV-mean
3.20
3.23
3.19
3.71*
2.93
This document is a draft for review purposes only and does not constitute Agency policy.
B-100 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
		("1995b")				
LIV-standard
deviation
0.158
0.336
0.402
0.288
0.274
LIV-number of rats
10
10
9
10
10
LNG-mean
0.37
0.38
0.38
0.37
0.34
LNG-standard
deviation
0.045
0.052
0.027
0.038
0.042
LNG-number of rats
10
10
9
10
10
TESTES-mean
0.69
0.65
0.71
0.72
0.67
TESTES-standard
deviation
0.060
0.101
0.092
0.089
0.136
TESTES-number of
rats
10
10
9
10
10
Mean Female Relative3 Organ Weights (g)
Parameter
Dose (mg/kg/day)
0
50
200
600
600
(recovery)
FBWb-mean
309
317
316
308
336
FBW-standard
deviation
23.4
34.8
20.0
28.2
33.9
FBW-number of
rats
10
10
10
10
10
ADR-mean
0.025
0.025
0.027
0.027
0.025
ADR-standard
deviation
0.003
0.005
0.005
0.004
0.005
ADR-number of rats
10
10
10
10
10
BRN-mean
0.67
0.66
0.67
0.68
0.63
BRN-standard
deviation
0.067
0.075
0.047
0.065
0.059
BRN-number of rats
10
10
10
10
10
KID-mean
0.76
0.71
0.76
0.82
0.71
KID-standard
deviation
0.059
0.088
0.051
0.059
0.040
KID-number of rats
10
10
10
10
10
LIV-mean
3.04
2.90
3.19
3.82*
2.88
LIV-standard
deviation
0.365
0.330
0.357
0.223
0.207
LIV-number of rats
10
10
10
10
10
LNG-mean
0.53
0.55
0.53
0.52
0.49
LNG-standard
deviation
0.071
0.059
0.052
0.047
0.079
This document is a draft for review purposes only and does not constitute Agency policy.
B-101 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
		("1995b")				
LNG-number of rats
10
10
10
10
10
OVARIES-mean
0.041
0.040
0.039
0.046
0.043
OVARIES-standard
deviation
0.006
0.015
0.014
0.018
0.011
OVARIES-number of
rats
10
10
10
10
9
Summary of gross necropsy observations (count)
Tissue and
Observation
Dose (mg/kg/day)
0
50
200
600
600
(recovery)
M
F
M
F
M
F
M
F
M
F
No gross lesions
observed
9
8
8
8
7
9
8
10
8
10
Mandibular lymph
nodes;
enlarged/red
—c
1
--
--
1
--
--
--
1
--
Mandibular lymph
nodes; enlarged
1
--
--
--
1
--
--
--
1
--
Tibia; lesion
(fracture)
--
1
--
--
--
--
--
--
--
--
Adrenals; small,
unilateral
--
--
1
--
--
--
--
--
--
--
Testes; small, white
(right)
--
--
1
--
--
--
--
--
--
--
Testes; absent (left)
--
--
--
--
--
--
--
--
1
--
Eye; opaque (left)
--
--
--
1
--
1
--
--
--
--
Thymus; focus, red
--
--
--
1
--
--
--
--
--
--
Thymus; mottled
--
--
--
--
--
--
1
--
--
--
Lung enlarged
--
--
--
--
Id
--
--
--
--
--
Large intestine,
cecum; focus, red
--
--
--
--
1
--
--
--
--
--
Liver; pale
--
--
--
--
--
--
1
--
--
--
Comments; 1,3,5- TMB was the only isomer tested in this study. Effects reported in study appeared reversible in the recovery
group, which was observed for 28 days following cessation of exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
B-102 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-28 (Continued): Characteristics and quantitative results for Koch Industries
	("1995b")	
^Significantly different from vehicle control, p<0.05
a Relative organ weight=[Absolute organ weight (g)/Fasted body weight (g)]xlOO
b fasted body weight
c zero incidence
d animal died due to gavage error (accidental death)
Na = sodium (mE/litter serum); K = potassium (mE/litter serum); CI = chloride (mE/litter serum); CK = creatine kinase (lU/liter
serum); ALK P = alkaline phosphatase (lU/liter serum); ALT = alanine aminotransferase (lU/liter serum); AST = aspartate
aminotransferase (lU/liter serum); GGT = gamma glutamyl transpeptidase (lU/liter serum); BUN = urea nitrogen (mg N/dL
serum); CREA = creatinine (mg/dL serum); T PRO = total protein (g protein/dL serum); ALB = albumin (g/dL serum); GLOB =
globulin (g/dL serum); A/G Ratio = albumin/globulin ratio; GLU = glucose (mg/dL serum); CHOL = cholesterol (mg/dL serum); T
BIL = total bilirubin (mg/dL serum); WBC = white blood cell (103/mm3); RBC = red blood cell (106/mm3); HGB = hemoglobin
(g/dL blood); HCT = hematocrit (%); MCV = mean corpuscular volume (femoliter); MCH = mean corpuscular hemoglobin
(picogram); MCHC = mean corpuscular hemoglobin concentration (%); PLT = platelet (103/mm3); NRBC = nucleated red blood
cells (number/100 white blood cells); MAT NEU = mature neutrophils (103/mm3); LYM = lymphocytes (103/mm3); MONO =
monocytes (103/mm3); EOS = eosinophils (103/mm3); BASO = basophils (103/mm3); IMM NEU = immature neutrophils
(103/mm3); ADR = adrenal glands; BRN = brain; KID = kidneys; LIV = liver; LNG = lung.
Source: Koch Industries (1995b)
This document is a draft for review purposes only and does not constitute Agency policy.
B-103 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-29. Characteristics and quantitative results for Korsak et al. (1995)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
IMP:DAK Wistar
rats and Balb/c
mice
M
8-
10/dose
Inhalation
250-2000 ppm (1,230-
9840 mg/m3) 1,2,4-TMB
4 hrs - neurotoxicity tests
6 minutes - respiratory tests
Additional study details
•	Animals were exposed to 1,2,4-TMB in a dynamic inhalation chamber (1.3 m3 volume) with 12-15 air changes/hr.
•	Mean initial body weights were 250-300 g for rats and 23-30 g for mice; animals were housed in wire mesh
stainless steel cages, with food and water provided ad libitum.
•	Animals were randomized and assigned to the experimental groups. Before rotarod experiment, rats were trained,
and only rats that balanced for 2 minutes on 10 consecutive days were used.
•	Rotarod, hot plate, and respiratory tests were conducted to measure effects on neuromuscular activity, pain
sensitivity, and respiratory rate respectively.
Figure 1. Rotarod performance of rats exposed to 1,2,4-TMB (i.e., pseudocumene). Rats were exposed to vapors of
solvent for 4 hrs.
Rotarod performance was tested immediately after termination of exposure. Each point represents probit of failures on rotarod in a
group of 10 rats.
8 8
iw
J2
£ 7H
s 6H
o
* 5
CD
CO
o 4-
CL
10
IT 3
a
a
D
£0^= 4693 mg/m3
(954 ppm)
1000
i	1	r-—i	1—i—i—rr
~i	1	1	r—f—i—r
100000
10000
Concentration of pseudocumene, mg/m3
Source: Korsak et al. (1995)
This document is a draft for review purposes only and does not constitute Agency policy.
B-104
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-29 (Continued): Characteristics and quantitative results for Korsak et al.
	(1995)	
Figure 2. Hot-plate behavior in rats exposed to 1,2,4-TMB (i.e., pseudocumene). Rats were exposed to vapors of
solvent for 4 hrs.
Hot-plate behavior was tested immediately after termination of exposure. Each point represents the mean value of
separate measurements of latency over the control in 10 rats.
65-
sp 60-
2 55-
•*-*
8 5CH
k_
CD
o 45H
5*
§ 40-
35h
30-
EQ0 = 5682 mg/m3
(1155 ppm)
a
1000
-i	1	1	r
Concentration of pseudocumene, mg/m 3
10000
Source: Korsak et al. (1995)
Figure 3. Time-response relationship for the effect of 1,2,4-TMB (i.e., pseudocumene) on respiratory rate in mice.
Each point represents the mean value in 8-10 mice. After termination of 6 min exposure recovery of respiratory rate was observed.
100'
G
O
U
o
a*
-------
Toxicological Review ofTrimethylbenzene
Table B-29 (Continued): Characteristics and quantitative results for Korsak et al.
	(1995)	
Figure 4. Respiratory rate of mice exposed to 1,2,4-TMB (i.e., pseudocumene) in 8-10 mice.
The decrease of respiratory rate observed in the 1st minute of exposure was taken for consideration. The regression line was
determined by the least squares procedure.
100
c
o
'8
CD
Q_
CD
"O
*5
CD"
iH
&
o
ta
a
CD
CC
50
20
10


a /
/
D/
*'

.#*

6

RC^, = 2844 mg/m3
(578 ppm)
1000
10000
Concentration of pseudocumene, mg/m3
Source: Korsak et al. (1995)
Health Effect at LOAEL
NOAEL
LOAEL
Decreased respiration rate,
impaired rotarod test
performance, decreased pain-
response time
n/a
n/a
Comments: No values are provided for dose-specific responses, and NOAEL and LOAEL cannot be determined. Exposures were of an
acute duration, and therefore not suitable for reference value derivation. However, qualitatively, this study provided evidence of
CNS disturbances that, when considered together with short-term and subchronic neurotoxicity studies, demonstrate that TMB
isomers perturb the CNS of exposed animals. The respiratory effects in mice also qualitatively support respiratory effects observed
in rats exposed subchronically to 1,2,4-TMB and 1,2,3-TMB.
Source: Korsak et al. (1995)
This document is a draft for review purposes only and does not constitute Agency policy.
B-106 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-30. Characteristics and quantitative results for Korsak and Rydzynski (1996)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
IMP: Wistar
rats
M
9-10/
dose
(1,2,4-TM
B)
10-30/
dose
(1,2,3-TM
B)
Inhalation (4 hrs or
6hr/day, 5
days/week, for 3
mos)
Acute exposure: 250-2,000
ppm 1,230 - 9840 mg/m3)
1,2,3-, 1,2,4-, or 1,3,5-TMB
Subchronic exposure: 0,
123,492, or 1,230 mg/m3
4 hrs or 3 mos
Additional study details
•	Animals were exposed to either 1,2,3-, 1,2,4-, or 1,3,5-TMB in a dynamic inhalation chamber (1.3 m3 volume) with
12 to 15 air changes/hour.
•	Mean initial body weights were 250-300 g; rats were housed in wire mesh stainless steel cages, with food and
water provided ad libitum.
•	Animals were randomized and assigned to the experimental groups.
•	Rotarod and hot plate tests were conducted to measure effects on neuromuscular function and pain sensitivity
respectively.
•	Rotarod performance was tested before, and immediately after, termination of exposure.
•	Normal neuromuscular function was indicated by the rats' ability to remain on a rod rotating at 12rpm for 2
minutes.
•	Hot-plate behavior was tested immediately after termination of exposure.
•	Latency of 60 seconds was considered as 100% inhibition of pain sensitivity.
•	Authors investigated the effects of exposure to 1,2,3-, 1,2,4- and 1,3,5- TMB on rotarod test performance and pain-
sensing response two weeks after the termination of exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
B-107
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table B-30 (Continued): Characteristics and quantitative results for Korsak and
	Rydzynski (1996)	
Figure 1. Rotarod performance of rats exposed to 1,2,3-TMB (hemimellitene), 1,2,4-TMB (pseudocumene), or
1,3,5-TMB (mesitylene). Rats were exposed to solvent vapors for 4 hrs.
Rotarod performance was tested immediately after termination of exposure. Each point represents probit of failures on rotarod in
a group of 10 rats. Normal neuromuscular function was indicated by the rats' ability to remain on a rod rotating at 12 rpm for 2
mins. The rotating rod was suspended 20 cm above metal bars connected to a 80 V/2 mA power source.
(/>
CD
3
'(5
-Q
o
0>

c
o
Q.
tn
CD
(/)
<15
3
(Z
-Q
O
PSEUDOCUMENE
/ O
ECW=4693 mg/m3
(954 ppm)
1000
9-
n	1—i—i i i i i	
10000
~1—I—I—I I I I
100000
Q- 5-
4
-------
Toxicological Review of Trimethylbenzene
Table B-30 (Continued): Characteristics and quantitative results for Korsak and
	Rydzynski (1996)	
Figure 2. Hot-plate behaviors in rats exposed to 1,2,3-TMB (hemimeliitene), 1,2,4-TMB (pseudocumene), or
1,3,5-TMB (mesitylene). Hot-plate behavior was tested immediately after termination of exposure.
Each point represents the mean value of separate measurements of latency in 10 rats. Latency of 60 sec was considered as 100%
inhibition of pain sensitivity.
Effects of exposure to trimethythenzene isomers
c
o
CO
£
Q.
fiJ
"O
&
>
V)
c
<1)
U)
e
're
Q.
c
0

T3
>
(O
c
0)
U!
re
Cl
70
65-
60-
55-
50-
45


10000
90-

70
60
50
40-
30-
20-
10000
90
80-
70-
60-
50-
40-
20-
10-
10	
1 000
100-1—
30	
1 000
100-T—
MESITYLENE
ECm = 5663 mg/m
(1212 ppm)
HEMIMELLITENE
EC50 = 4172 mg/m1
(848 ppm)
PSEUDOCUMENE
ECm = 5682 mg/m
(1155 ppm)
& 50
>
S 40
(/)
S 30
in
•I 20-I
Q.
1000
10000
Concentration, mg/m1
Source: Reproduced from Korsak and Rydzynski (1996)
This document is a draft for review purposes only and does not constitute Agency policy.
B-109 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-30 (Continued): Characteristics and quantitative results for Korsak and
	Rydzynski (1996) 	

Figure 3. Rotarod performance of rats exposed to
1,2,3-TMB (hemimellitene) or 1,2,4-TMB
(pseudocumene) at concentrations of 25,100,
and 250 ppm (123, 492,1,230 mg/mB).
Rats were exposed to vapors of solvents for 6 hr/day,
5 days/week, 3 mos. Statistical significance marked by
asterisks, p<0.005.
Source: Reproduced from Korsak and Rydzynski (1996)
40
35
2
= 20
•2
o
35 10
0
80
60
S?
<3 40
<*—
«*-
o
5?
20
0
PseudocurncTO^^-^^"
/ 	X
/ control
/ + + 4*25 ppm
/"3IH00 ppm
	x	K	X 250 PPm
j	exposure-	J
0 4 8 13 15 weeks
Hemimellitene
/ -X" control
/ yA I +25 ppm

'	exposure	'
0 4 8 13 15 weeks

Observation
Latency of the paw-lick response, sec
1,2,4-TMB
1,2,3-TMB
Control
15.4 ±5.8
9.7 ±2.1
25 ppm (100 mg/m3)
18.2 ±5.7
11.8 ±3.8*
100 ppm (492 mg/m3)
27.6 ±3.2**
16.3 ± 6.3***
250 ppm (1,230 mg/m3)
30.1 ±7.9**
17.3 ±3.4**
250 ppm (1,230 mg/m3) 2 weeks after termination of
exposure
17.3 ±3.9
11.0 ±2.4
Health Effect at LOAEL
NOAEL
LOAEL
Decreased pain sensitivity
n/a for 1,2,3-TMB
25 ppm (123 mg/m3) for
1,2,4-TMB
25 ppm (123 mg/m3) for
1.2.3-TMB
100 ppm (492 mg/m3) for
1.2.4-TMB
Comments: Although rotarod data are useful in providing a qualitative description of neuromuscular impairment following
1,2,4-TMB or 1,2,3-TMB exposure, in comparison to effects on pain sensitivity, the data are not considered as robust regarding
suitability for derivation of reference values. Namely, data are presented as dichotomized values instead of a continuous
measurement of latency. The acute exposures were not suitable for reference value derivation. However, qualitatively, effects
observed following acute exposures provided evidence of CNS disturbances that, when considered together with subchronic
neurotoxicity tests, demonstrate that TMB isomers perturb the CNS of exposed animals. It is unclear whether the latency to pawlick
and rotarod tests were performed sequentially in the same cohort of animals.
*, ** statistically significant from controls at p < 0.05 and p < 0.01, respectively.
*** Level of significance not reported in Table 1 from Korsak and Rydzynski (1996). however the results of an ad-hoc t-test
(performed by EPA) indicated significance at p < 0.01
This document is a draft for review purposes only and does not constitute Agency policy.
B-110 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-31. Characteristics and quantitative results for Korsak et al. (1997)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration

M

Acute -Inhalation, 6
minutes
Subchronic 0
Inhalation,6 hr/day,
5 days/week
Acute - 250-2000 ppm
(1,230-9840 mg/m3)
1,2,4-TMB, 1,2,3-TMB, or
1,3,5-TMB
Subchronic - 0,123, 492,
1,230 mg/m31,2,4-TMB
Acute - 6 minutes
Subchronic - 90 days
MP:DAK
i/Vistar rats
and Balb/C
nice
Acute -
8/dose
Subchronic
- 6-7/dose
Additional study details
•	Animals were exposed to 1,2,4-TMB in a dynamic inhalation chamber (1.3 m3 volume) with 12-15 air changes/hr.
•	Rats weighed 250-300 g and were housed in stainless steel wire mesh cages, with food and water provided ad
libitum.
•	Rats were anesthetized 24 hrs after termination of exposure, and bronchoalveolar (BAL) fluid was collected from
lung lavage.
•	All rats exposed to 1,2,4-TMB survived until the end of exposure and no clinical observations of toxicological
significance were reported.
Observation
Exposure concentration (mg/mB)
0
123
492
1,230

Body weight (mean ± SD)
Body weight (g)
411 ±28
383 ± 25
409 ± 56
416 ±27

BAL cell counts (mean ± SD)
Total cells (106/cm3)
1.93 ±0.79
5.82 ± 1.32***
5.96 ±2.80**
4.45 ± 1.58*
Macrophages (106/cm3)
1.83 ± 0.03
3.78 ±0.8
4.95 ±0.2**
3.96 ±0.3**
Polymorphonuclear leucocytes
(106/cm3)
0.04 ± 0.02
1.54 ±0.7
0.52 ±0.6
0.21 ±0.3
Lymphocytes (106/cm3)
0.06 ±0.01
0.5 ±0.2
0.5 ±0.4
0.2 ±0.1
Cell viability (%)
98.0 ± 1.7
95.5 ± 1.6
95.3 ±3.5
95.3 ±3.1

BAL protein levels and enzyme activities (mean ±SD)
Total protein (mg/mL)a
0.19 ±0.04
0.26 ±0.07*
0.26 ±0.06*
0.24 ± 0.08
Mucoproteins (mg/mL)a
0.16 ±0.03
0.14 ±0.02*
0.13 ±0.02
0.12 ±0.02
Lactate dehydrogenase (mU/mL)a
34.2 ± 8.52
92.5 ±37.2***
61.3 ±22.9*
53.8 ±28.6
Acid phosphatase mU/mL)a
0.87 ±0.20
1.28 ±0.37*
1.52 ±0.42*
1.26 ±0.22*
This document is a draft for review purposes only and does not constitute Agency policy.
B-lll
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-31 (Continued): Characteristics and quantitative results for Korsak et al.
	(1997)	
120
100
Z 80
-------
Toxicological Review ofTrimethylbenzene
Table B-32. Characteristics and quantitative results for Korsak et al. (2000a)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
IMP: Wistar
rats
M
and F
10/dose
Inhalation
(6 hr/day,
5 days/week)
0, 123, 492, 1,230 mg/m3
90 days
Additional study details
•	Animals were exposed to 1,2,4-TMB in a dynamic inhalation chamber (1.3 m3 volume) with 16 air changes/hr.
•	Mean initial body weights were 213 ± 20 for males and 160 ± 11 for females; rats were housed in polypropylene
cages with wire-mesh covers (5 animals/cage), with food and water provided ad libitum.
•	Animals were randomized and assigned to the experimental groups.
•	Hematological parameters were evaluated prior to exposure and 1 week prior to termination of exposure, and for
the 1230 mg/m3 exposure group, also evaluated two weeks after termination of exposure; blood clinical chemistry
parameters were evaluated 18 hrs after termination of exposure (animals were deprived of food for 24 hrs).
•	Necropsy was performed on all animals. Pulmonary lesions were graded using an arbitrary scale: 1 = minimal, 2 =
mild, 3 = moderate, 4 = marked.
Observation
Exposure concentration (mg/mB)
0
123
492
1,230

Body and Organ weights (mean ± SD)

Males
Terminal body weight (g)
368 ± 22
390 ± 26
399 ± 22
389 ± 29
Absolute organ weight (g)




Lungs
1.78 ±0.28
1.83 ±0.25
2.93 ±0.26*
1.78 ±0.36
Liver
10.27 ± 1.82
11.43 ± 1.05
10.78 ± 1.33
10.86 ± 2.04
Spleen
0.68 ±0.08
0.85 ±0.19*
0.79 ±0.09
0.72 ±0.08
Kidney
2.06 ±0.13
2.24 ±0.15
2.14 ±0.15
2.18 ±0.16
Adrenals
0.048 ± 0.007
0.046 ± 0.005
054 ±0.011
0.047 ± 0.005
Testes
3.72 ±0.35
3.90 ±0.38
4.03 ±0.27
3.87 ±0.24
Heart
0.90 ± 0.04
0.94 ±0.06
0.94 ±0.08
0.96 ±0.07
Relative organ weight (g)




Lungs
0.496 ± 0.056
0.475 ±0.056
0.586 ±0.115
0.477 ± 0.080
Liver
2.896 ±0.456
2.894 ±0.427
2.990 ±0.465
2.901 ±0.479
Spleen
0.189 ±0.011
0.220 ±0.041
0.210 ±0.018
0.200 ± 0.018
Kidney
0.588 ± 0.029
0.585 ±0.022
0.587 ± 0.065
0.586 ± 0.040
Adrenals
0.011 ± 0.003
0.010 ± 0.000
0.022 ± 0.024
0.011 ±0.003
Testes
1.041 ± 0.076
1.020 ±0.079
1.067 ±0.102
1.039 ± 0.077
Heart
0.252 ±0.013
0.239 ±0.020
0.249 ± 0.014
0.258 ±0.020
This document is a draft for review purposes only and does not constitute Agency policy.
B-113
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-32 (Continued): Characteristics and quantitative results for Korsak et al.
	("2000a")	

Females
Terminal body weight (g)
243 ± 16
243 ± 19
230 ± 14
229 ±21
Absolute organ weight (g)




Lungs
1.29 ±0.18
1.32 ±0.12
1.25 ±0.13
1.23 ±0.11
Liver
6.48 ± 1.02
6.54 ±0.69
5.81 ±0.83
6.72 ± 1.34
Spleen
0.59 ±0.08
0.61 ±0.11
0.49 ± 0.06*
0.52 ±0.08
Kidney
1.55 ±0.12
1.50 ±0.14
1.38 ±0.11*
1.44 ±0.19
Adrenals
0.065 ± 0.007
0.070 ± 0.008
0.066 ± 0.010
0.061 ±0.013
Ovaries
0.09 ± 0.02
0.09 ±0.01
0.09 ±0.27
0.09 ± 0.02
Heart
0.66 ±0.07
0.64 ±0.05
0.61 ±0.07
0.63 ±0.06
Relative organ weight (g)




Lungs
0.555 ±0.058
0.581 ±0.040
0.596 ±0.051
0.569 ±0.053
Liver
2.770 ±0.222
2.881 ±0.309
2.758 ±0.223
3.078 ± 0.434
Spleen
0.255 ±0.025
0.266 ±0.031
0.237 ±0.036
0.24 ±0.033
Kidney
0.667 ± 0.030
0.661 ±0.047
0.660 ± 0.042
0.662 ±0.036
Adrenals
0.028 ± 0.006
0.031 ±0.006
0.032 ± 0.006
0.029 ± 0.006
Ovaries
0.043 ± 0.008
0.041 ±0.006
0.045 ± 0.013
0.047 ± 0.009
Heart
0.284 ± 0.023
0.283 ±0.025
0.291 ±0.025
0.289 ±0.015
Observation
Exposure concentration (mg/mB)
0
123
492
1,230
l,230a
Trend
testb

Hematological parameters (mean ± SD)

Males
Hematocrit (%)
49.9 ± 1.9
50.4 ± 2.0
50.0 ± 1.9
50.6 ± 1.5
50.1 ± 1.1
0.2993
Hemoglobin (g/dL)
15.1 ± 1.1
15.6 ±0.9
15.4 ±0.9
15.4 ±0.6
16.0 ± 1.0
0.2112
RBCs (x 106/mm3)c
9.98 ± 1.68
9.84 ± 1.82
8.50 ± 1.11
7.70 ± 1.38**
7.61 ± 1.6
0.0004
WBCs (x 103/mm3)d
8.68 ±2.89
8.92 ± 3.44
8.30 ± 1.84
15.89 ±5.74**
7.11 ± 2.1
0.0019
Rod neutrophil (%)
0.0 ± 0.0
0.4 ±0.5
0.2 ±0.4
0.9 ± 1.5
0.7 ±0.8
0.0586
Segmented neutrophil (%)
24.1 ±9.2
19.7 ±6.5
20.7 ±7.7
18.9 ± 10.8
29.4 ±6.4
0.0730
Eosinophil (%)
1.2 ± 1.7
1.2 ± 1.0
0.4 ±0.6
1.7 ± 1.4
1.5 ± 1.5
0.2950
Lymphocyte (%)
73.5 ± 10.3
76.2 ±7.1
76.3 ±8.5
75.8 ± 16.0
65.4 ±8.9
0.1297
Monocyte (%)
1.1 ± 1.3
2.5 ±2.1
2.3 ±2.2
1.8 ±2.5
2.7 ±2.5
0.3818
Lymphoblast (%)
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.8 ± 1.3
0.3 ±0.9
0.1387
Myelocyte (%)
0.0 ±0.0
0.0 ±0.0
0.2 ±0.4
0.0 ±0.0
0.0 ±0.0
0.4046
Erythroblase (%)
0.0 ± 0.0
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.5000
Reticulocyte (%)
3.1 ±2.3
2.3 ± 1.4
2.8 ±2.1
3.1 ±2.5
6.4 ±3.2
0.4900
Platelet (x 103/mm3)
294 ± 46
293 ± 73
359 ± 46
335 ± 80
386 ± 70
0.0741
Clotting time (sec)
43 ± 19
41 ± 17
37 ± 13
33 ±7
56 ±21
0.1457
This document is a draft for review purposes only and does not constitute Agency policy.
B-114 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-32 (Continued): Characteristics and quantitative results for Korsak et al.
	("2000a")	

Females
Hematocrit (%)
46.0 ± 1.6
46.6 ±2.7
47.0 ±2.7
46.5 ±4.1
45.8 ± 1.3
0.2336
Hemoglobin (g/dL)
14.5 ±0.9
13.8 ± 1.3
14.4 ±0.9
14.2 ±0.9
14.9 ±0.9
0.3461
RBCs (x 106/mm3)c
8.22 ± 1.16
7.93 ± 2.04
8.51 ± 1.13
7.71 ± 1.58
6.99 ± 1.8
0.1891
WBCs (x 103/mm3)d
7.50 ± 1.31
6.76 ±2.95
9.55 ±4.48
9.83 ±3.74
7.11 ±2.4
0.0307
Rod neutrophil (%)
1.4 ± 1.6
0.5 ±0.7
0.4 ±0.5
0.4 ±0.9
0.5 ±0.7
0.3270
Segmented neutrophil (%)
22.8 ±6.5
15.5 ±7.9
20.7 ±7.5
17.4 ±9.3
20.5 ±9.5
0.1868
Eosinophil (%)
1.2 ±0.6
1.6 ± 1.6
1.1 ± 1.7
1.2 ±2.1
2.0 ± 1.7
0.1051
Lymphocyte (%)
73.2 ±7.9
79.4 ±8.4
75.5 ±7.4
78.8 ± 11.6
74.1 ±9.5
0.2140
Monocyte (%)
1.2 ± 1.3
2.6 ±2.8
1.3 ± 1.7
1.5 ±0.8
1.5 ± 1.4
0.4156
Lymphoblast (%)
0.0 ±0.0
0.1 ±0.3
0.5 ± 1.5
0.7 ± 1.1
0.8 ± 1.3
0.1361
Myelocyte (%)
0.0 ±0.0
0.0 ±0.0
0.5 ± 1.5
0.1 ±0.3
0.1 ±0.3
0.3189
Erythroblase (%)
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.5000
Reticulocyte (%)
3.5 ±2.6
1.7 ±2.0
1.8 ±0.9
1.0 ±0.6*
5.8 ±3.6
0.0137
Platelet (x 103/mm3)
306 ± 34
234 ± 50*
303 ± 48
325 ± 57
349 ± 77
0.1542
Clotting time (sec)
30 ± 10
23 ±4
19 ± 5**
22 ±7*
48 ± 19
0.0034
Observation
Exposure concentration (mg/mB)
0
123
492
1,230
Trend
testb

Clinical chemistry parameters (mean ± SD)

Males
AST (U/dL)e
138.7 ± 20.6
141.3 ±21.0
134.5 ± 27.0
138.4 ± 35.0
0.2223
ALT (U/d L)f
51.7 ±5.9
48.3 ±7.8
49.7 ±9.1
46.8 ±5.1
0.0637
ALP (U/d L)8
80.4 ± 12.0
86.2 ±22.0
84.9 ±21.0
90.5 ± 19.0
0.1518
SDH (U/dL)h
6.6 ± 1.4
8.1 ±0.8**
7.8 ± 1.0*
8.0 ± 1.1**
0.0083
GGT (nU/ml)'
0.22 ± 0.44
0.20 ±0.42
0.20 ±0.42
0.20 ± 0.42
0.4700
Bilirubin (mg/dL)
1.027 ±0.193
0.974 ±0.338
1.106 ±0.289
0.932 ±0.175
0.2594
Total cholesterol (mg/dL)
63.6 ± 13.0
69.1 ± 12.0
72.4 ± 14.9
70.6 ± 19.5
0.0920
Glucose (mg/dL)
141.9 ±23.9
163.8 ±29.7
157.9 ±23.2
162.2 ±28.9
0.0876
Total protein (g)
5.43 ± 1.00
5.47 ± 1.39
5.34 ± 1.29
5.82 ± 1.49
0.3242
Albumin (g)
3.25 ±0.60
3.45 ±0.56
3.41 ±0.83
3.53 ±0.66
0.2279
Creatinine (mg/dL)
0.506 ± 0.099
0.437 ±0.138
0.510 ±0.150
0.490 ±0.178
0.3982
Urea (mg/dL)
54.2 ±8.6
48.8 ±8.3
47.6 ±3.4
49.0 ±8.7
0.1145
Calcium (mg/dL)
10.4 ±0.5
10.8 ±0.5
10.7 ±0.8
10.8 ±0.7
0.2449
Phosphorus (mg/dL)
6.27 ±0.49
6.50 ±0.57
6.49 ±0.61
6.46 ±0.78
0.1580
Sodium (mmol/L)
139.0 ± 1.4
1393 ± 1.3
139.6 ± 1.4
139.0 ± 1.4
0.4950
Potassium (mmol/L)
4.87 ±0.36
4.97 ±0.34
4.97 ±0.25
4.83 ± 0.40
0.2907
Chloride (mmol/L)
106.6 ± 1.2
106.1 ± 1.7
106.3 ± 1.5
106.7 ± 1.2
0.4353
This document is a draft for review purposes only and does not constitute Agency policy.
B-115 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-32 (Continued): Characteristics and quantitative results for Korsak et al.
	("2000a")	

Females
AST (U/dL)e
139.4 ± 16.6
136.7 ±27.1
145.5 ±22.7
141.4 ± 15.6
0.2118
ALT (U/d L)f
49.8 ±6.3
51.4 ±8.2
50.4 ±9.0
55.1 ±9.5
0.1844
ALP (U/d L)8
41.2 ±7.8
37.2 ±6.8
39.8 ± 11.0
49.8 ± 15.5
0.1740
SDH (U/dL)h
5.9 ± 1.5
7.3 ± 1.7
7.1 ± 1.8
7.0 ± 1.6
0.0637
GGT (nU/ml)'
0.20 ± 0.42
0.30 ±0.48
0.10 ±0.32
0.44 ± 0.53
0.2821
Bilirubin (mg/dL)
0.745 ± 0.342
0.690 ±0.396
0.743 ± 0.248
0.642 ±0.257
0.3092
Total cholesterol (mg/dL)
64.5 ± 11.9
65.7 ± 12.8
64.1 ± 10.8
62.5 ±7.6
0.4775
Glucose (mg/dL)
118.2 ±28.8
138.8 ±38.5
104.5 ± 23.8
129.9 ±39.7
0.4838
Total protein (g)
6.91 ±0.53
7.44 ±0.89
7.08 ±0.35
6.94 ±0.64
0.4036
Albumin (g)
3.42 ±0.24
3.46 ±0.27
3.61 ±0.26
3.42 ±0.15
0.2408
Creatinine (mg/dL)
0.655 ±0.135
0.553 ±0.104
0.629 ±0.153
0.577 ±0.133
0.1641
Urea (mg/dL)
52.7 ±7.8
49.6 ±6.7
52.8 ± 10.5
52.2 ± 11.8
0.4718
Calcium (mg/dL)
10.5 ±0.6
10.8 ±0.8
10.6 ±0.5
10.8 ±0.6
0.3011
Phosphorus (mg/dL)
4.75 ±0.54
5.05 ±0.70
5.34 ±0.74
4.90 ± 1.01
0.4050
Sodium (mmol/L)
137.9 ± 1.7
138.0 ± 1.8
137.8 ±2.5
138.2 ±2.2
0.3628
Potassium (mmol/L)
4.54 ±0.22
4.39 ±0.61
4.51 ±0.26
4.46 ±0.25
0.4108
Chloride (mmol/L)
104.9 ± 2.0
105.5 ± 1.3
105.9 ± 1.6
106.4 ± 1.8
0.0601
Observation
Exposure concentration (mg/mB)
[Dose Group ID]
0
[1]
123
[2]
492
[3]
1,230
[4]
Comparison to
controls'
Trend
testb

Males
Proliferation of peribronchial
lymphatic tissue (0—4)k
16.01
15.6
30.6
17.4
1-3*
0.13
Formation of
lymphoepithelium in bronchii
(0-4)
18.1
15.6
27.9
18.2

22
Bronchitis and
bronchopneumonia (0-4)
19.0
18.3
26.1
16.5

0.49
Interstitial lymphocytic
infiltration (0-3)
14.8
18.4
26.9
19.4
1-3*
0.12
Alveolar macrophages (0-3)
14.1
14.8
24.1
26.4
1-4*
0.002
Cumulative score of all
individuals
13.9
15.1
29.1
21.3
1-3*
0.02
This document is a draft for review purposes only and does not constitute Agency policy.
B-116 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-32 (Continued): Characteristics and quantitative results for Korsak et al.
	("2000a")	

Females
Proliferation of peribronchial
lymphatic tissue (0—4)k
19.4
21.7
21.2
17.5

0.36
Formation of
lymphoepithelium in bronchii
(0-4)
18.3
20.1
25.1
16.1

0.48
Bronchitis and
bronchopneumonia (0-4)
19.0
22.9
19.0
19.0

0.48
Interstitial lymphocytic
infiltration (0-3)
15.8
14.5
21.5
29.2
1-4*
0.0017
Alveolar macrophages (0-3)
19.7
14.9
16.6
29.8
ns
0.03
Cumulative score of all
individuals
16.8
15.3
21.3
27.3
ns
0.01

Health Effect at LOAEL
NOAEL
LOAEL
Increased pulmonary lesions,
decreased RBCs, and
increased WBCs in males
123 mg/m3
492 mg/m3
Comments: The observed inflammatory lesions are coherent with observations of increased inflammatory cell populations in
bronchoalveolar lavage fluid in Korsak et al. (1997). The authors did not report the incidences of pulmonary lesions, but rather the
results of the Kruskall-Wallis test. This makes it difficult to interpret the dose-response relationship and limits analysis of these
endpoints to the NOAEL/LOAEL method for determining a POD, rather than using BMD modeling.
aEffects measured in rats exposed to 1,230 mg/mB 2 weeks after termination of exposure.
bp-value reported from Jonckheere's trend test
cred blood cells,
dwhite blood cells,
Aspartate aminotransferase,
'alanine aminotransferase,
galkaline phosphatase,
h sorbitol dehydrogenase,
'y-glutamyltransferase,
J Reports the results of pair-wise statistical significance of exposure groups compared to controls (i.e., 1-3 would indicate that the
492 mg/mB was statistically significantly different from controls)
kgrading system (0-4, 0-3; see Additional study details above)
1 results presented as ranges of the Kruskal-Willis test.
*, ** Statistically significant from controls at p < 0.05 and 0.01, respectively.
Source: Korsak et al. (2000a)
This document is a draft for review purposes only and does not constitute Agency policy.
B-117 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-33. Characteristics and quantitative results for Korsak et al. (2000b)
Study design
Species
Sex
N
Exposure route
Concentration range
Exposure duration
IMP: Wistar
rats
M & F
10/dose,
20 in the
1,230
mg/m3
group
Inhalation
(6 hr/day,
5 days/week)
0, 123, 492, 1,230 mg/m3
1,2,3-TMB
90 days
Additional study details
•	Animals were exposed to 1,2,3-TMB in a dynamic inhalation chamber (1.3 m3 volume) with 16 air changes/hr.
•	Mean initial body weights were 290 ± 25 g for males and 215 ± 13 g for females; rats were housed in polypropylene
cages with wire-mesh covers (5 animals/cage), with food and water provided ad libitum.
•	Animals were randomized and assigned to the experimental groups.
•	Hematological parameters were evaluated prior to exposure and 1 week prior to termination of exposure, and for
the 1230 mg/m3 exposure group, also evaluated two weeks after termination of exposure; blood clinical chemistry
parameters were evaluated 18 hrs after termination of exposure (animals were deprived of food for 24 hrs).
•	Necropsy was performed on all animals.
•	Pulmonary effects were graded using an arbitrary scale: 0 = normal status, 1 = minimal, 2 = mild, 3 = moderate,
4 = marked.
Observation
Exposure concentration (mg/mB)
0
123
492
1,230

Body and organ weights (mean ± SD)

Males
Terminal Body weight (g)
390 ± 35
408 ± 50
404 ± 33
413 ± 46
Absolute organ weight (g)




Lungs
1.90 ±0.22
1.86 ±0.26
1.99 ±0.37
1.88 ±0.34
Liver
8.28 ±0.97
8.83 ± 1.40
9.05 ±0.99
9.54 ± 1.50
Spleen
0.71 ±0.06
0.12 ±0.10
0.82 ±0.11
0.79 ±0.20
Kidney
2.34 ±0.27
2.29 ±0.23
2.48 ±0.25
2.50 ±0.25
Adrenals
0.059 ±0.012
0.061 ±0.016
0.061 ±0.013
0.061 ±0.012
Testes
3.78 ±0.44
3.69 ±0.24
3.71 ±0.36
3.91 ±0.12
Heart
1.04 ±0.13
0.98 ±0.11
1.08 ±0.13
1.15 ±0.19
Relative organ weight (g)




Lungs
0.510 ±0.071
0.479 ± 0.026
0.504 ± 0.082
0.468 ± 0.073
Liver
2.208 ±0.163
2.271 ±0.129
2.287 ±0.115
2.414 ±0.214*
Spleen
0.190 ±0.019
0.187 ±0.015
0.207 ±0.021
0.203 ± 0.058
Kidney
0.623 ± 0.049
0.594 ± 0.029
0.629 ±0.033
0.637 ± 0.060
Adrenals
0.016 ± 0.003
0.016 ± 0.003
0.015 ± 0.003
0.016 ± 0.003
Testes
1.014 ± 0.087
0.961 ±0.091
0.941 ±0.063
1.002 ±0.106
Heart
0.277 ±0.027
0.252 ± 0.018
0.274 ± 0.032
0.284 ± 0.026
This document is a draft for review purposes only and does not constitute Agency policy.
B-118
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-33 (Continued): Characteristics and quantitative results for Korsak et al.
	("2000b")	

Females
Terminal Body weight (g)
268 ± 18
262 ±21
263 ± 14
259 ± 23
Absolute organ weight (g)




Lungs
1.62 ±0.15
1.55 ±0.33
1.47 ±0.18
1.51 ±0.16
Liver
6.05 ±0.42
5.85 ± 0.47
5.94 ±0.51
6.05 ± 0.44
Spleen
0.63 ±0.05
0.61 ±0.10
0.57 ±0.05*
0.56 ±0.06*
Kidney
1.58 ±0.16
1.53 ±0.12
1.54 ±0.10
1.62 ±0.16
Adrenals
0.080 ± 0.014
0.082 ± 0.010
0.083 ±0.011
0.075 ±0.015
Ovaries
0.12 ±0.03
0.12 ±0.03
0.13 ±0.02
0.14 ±0.04
Heart
0.74 ±0.05
0.71 ±0.50
0.75 ±0.06
0.73 ±0.08
Relative organ weight (g)




Lungs
0.651 ±0.053
0.637 ±0.122
0.604 ± 0.049
0.639 ±0.076
Liver
2.434 ±0.143
2.400 ± 0.088
2.448 ±0.190
2.555 ±0.214
Spleen
0.257 ±0.027
0.249 ± 0.032
0.234 ±0.19
0.237 ±0.022
Kidney
0.639 ±0.076
0.628 ± 0.024
0.638 ±0.032
0.686 ±0.058
Adrenals
0.032 ± 0.005
0.034 ± 0.004
0.034 ± 0.005
0.032 ± 0.008
Ovaries
0.051 ±0.014
0.050 ± 0.014
0.056 ± 0.006
0.060 ± 0.018
Heart
0.298 ±0.016
0.291 ±0.012
0.309 ± 0.024
0.307 ± 0.026
Observation
Exposure concentration (mg/mB)
0
123
492
1,230
1230a
Trend
testb

Hematological parameters (mean ± SD)
Hematocrit (%) Males
46.4 ± 1.6
45.8 ±2.6
45.7 ± 1.3
45.5 ±2.1
43.5 ±26
0.1615
Hematocrit (%) Females
42.7 ±2.2
45.0 ±2.4
41.8 ± 1.6
41.5 ±24
41.7 ±20
0.0198
Hemoglobin (g/dL) Males
16.4 ± 1.0
17.6 ± 1.6
17.6 ±0.8
15.0 ± 1.2
ND
0.0688
Hemoglobin (g/dL) Females
13.9 ±0.7
15.1 ± 1.0*
14.6 ±0.6
14.7 ±0.9
ND
0.0748
RBCs (x 106/mm3)c Males
9.49 ± 2.03
10.25 ± 1.29
10.11 ± 1.27
8.05 ± 1.38*
8.6 ± 1.5
0.0011
RBCs (x 106/mm3)c Females
8.03 ± 1.11
8.73 ± 1.24
7.79 ± 1.57
7.27 ± 1.32
6.6 ± 1.8
0.0185
WBCs (x 103/mm3)d Males
10.09 ± 2.23
9.38 ±3.29
7.71 ±3.45
9.03 ± 275
6.3 ±4.6
0.1661
WBCs (x 103/mm3)d Females
10.71 ±4.28
9.54 ±2.37
13.02 ± 3.07
13.01 ±4.53
62 ±2.5
0.0189
Rod neutrophil (%) Males
0.8 ± 1.0
1.0 ± 1.1
0.4 ±0.5
0.5 ±0.6
5.2 ±3.0
0.1878
Rod neutrophil (%) Females
0.4 ±0.8
0.6 ±0.6
1.1 ± 1.4
0.4 ±0.8
1.8 ±2.2
0.4711
Segmented neutrophil (%)
Males
24.8 ±4.5
25.4 ±5.8
20.7 ±5.8
17.7 ±8.3*
27.5 ±9.2
0.0032
Segmented neutrophil (%)
Females
23.1 ± 6.1
19.7 ±3.4
16.4 ±4.2*
11.9 ±7.1**
19.6 ±8.3
0.0000
Eosinophil (%) Males
1.3 ± 1.4
0.8 ± 1.0
0.8 ± 1.1
0.6 ±0.8
0.6 ±0.6
0.1439
Eosinophil (%) Females
1.4 ± 1.0
0.6 ±0.6
0.7 ±0.8
0.8 ±0.9
0.7 ±0.8
0.2778
Lymphocyte (%) Males
71.2 ±5.0
71.6 ±6.8
75.4 ±4.7
79.3 ±
78.0**
63.7 ± 11.3
0.0015
Lymphocyte (%) Females
73.2 ±7.9
77.5 ±4.9
80.4 ±5.1
84.0 ±
78.0**
75.7 ±9.9
0.0003
This document is a draft for review purposes only and does not constitute Agency policy.
B-119 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-33 (Continued): Characteristics and quantitative results for Korsak et al.
	("2000b") 				
Monocyte (%) Males
1.9 ± 1.6
1.3 ± 1.4
2.3 ±20
1.6 ± 22
3.1 ±3.7
0.3014
Monocyte (%) Females
2.0 ±2.0
1.6 ± 1.6
1.1 ± 1.3
2.1 ± 1.7
1.3 ± 1.8
0.2426
Lymphoblast (%) Males
0.0 ±0.0
0.0 ±0.0
0.2 ±0.6
0.2 ±0.6
0.0 ±0.0
0.2911
Lymphoblast (%) Females
0.0 ±0.0
0.0 ±0.0
0.1 ±0.3
0.3 ±0.7
0.0 ±0.0
0.1403
Myelocyte (%) Males
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.5000
Myelocyte (%) Females
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.5 ±0.2
0.0 ±0.0
0.3963
Erythroblast (%) Males
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.5000
Erythroblast (%) Females
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.1 ±0.3
0.0 ±0.0
0.2995
Reticulocyte (%) Males
2.8 ± 1.3
2.1 ± 1.7
3.8 ±2.1
4.5 ± 1.8*
6.9 ±3.1**
0.0017
Reticulocyte (%) Females
2.6 ±0.9
4.6 ±2.5*
5.2 ± .50*
4.4 ±3.0
6.8 ±3.5**
0.0459
Platelet (x 103/mm3) Males
262 ±51
266 ± 70
257 ± 81
242 ± 76
277 ± 80
0.1708
Platelet (x 103/mm3) Females
224 ± 68
290 ± 70
249 ± 53
204 ± 44
258 ±45
0.0329
Clotting time (sec) Males
29.7 ±8.6
23.0 ± 10.0
37.9 ±9.9
29.2 ± 15.6
21.7 ±5.4
0.4650
Clotting time (sec) Females
27.2 ±2.8
25.0 ±9.4
23.8 ±9.5
25.1 ± 12.1
25.9 ±8.0
0.3479
Observation
Exposure concentration (mg/mB)
0
123
492 1,230
Trend
testb

Clinical chemistry parameters (mean ± SD)
AST (U/dL)e Males
107.8 ± 14.2
102.9 ± 15.1
103.6 ± 14.5
119.6 ±27.3
0.2223
AST (U/dL)e Females
96.1 ±9.4
96.9 ±9.9
117.1 ±23.9
104.6 ± 15.7
0.2118
ALT (U/dL)f Males
41.3 ±2.0
40.7 ±3.1
41.5 ±5.5
45.5 ±5.6
0.0637
ALT (U/dL)f Females
39.7 ±3.5
39.5 ±6.4
36.2 ±3.3
30.5 ± 9.9**
0.1844
ALP (U/dL)s Males
70.5 ± 15.2
70.6 ± 11.7
66.5 ± 10.8
63.7 ± 15.7
0.1518
ALP (U/dL)s Females
21.5 ±2.7
25.8 ±8.4
31.1 ±8.6*
30.5 ±9.9*
0.1740
SDH (U/dL)h Males
1.6 ±0.7
2.3 ± 1.3
2.5 ±0.9
2.7 ±0.7*
0.0083
SDH (U/dL)h Females
1.7 ±0.7
1.9 ±0.9
1.5 ±0.7
1.8 ± 1.0
0.0637
GGT (nU/ml)' Males
0.77 ±0.66
0.77 ±0.97
0.40 ±0.51
0.50 ±0.75
0.4700
GGT (nU/ml)1 Females
0.55 ±0.72
0.44 ± 1.01
0.66 ± 1.11
0.30 ± 0.48
0.2821
Bilirubin (mg/dL) Males
0.600 ±0.516
0.600 ±0.516
0.800 ± 0.422
0.625 ±0.518
0.2594
Bilirubin (mg/dL) Females
0.911 ±0.348
1.161 ±0.469
0.930 ±0.463
0.976 ±0.421
0.3092
Total cholesterol (mg/dL)
Males
63.1 ± 10.1
62.2 ± 11.6
64.5 ± 16.2
65.0 ±9.1
0.0920
Total cholesterol (mg/dL)
Females
60.1 ± 12.2
62.4 ± 15.3
62.3 ±7.7
64.4 ± 14.1
0.4775
Glucose (mg/dL) Males
95.5 ± 13.1
110.8 ± 14.7
100.2 ± 15.2
114.5 ± 20.6
0.0876
Glucose (mg/dL) Females
115.9 ±8.5
121.0 ± 17.5
109.2 ± 5.8
109.8 ± 10.8
0.4838
Total protein (g) Males
7.84 ±0.13
8.02 ±0.50
7.76 ±0.27
8.04 ± 0.59
0.3242
Total protein (g) Females
8.24 ± 1.24
8.36 ± 1.14
8.65 ±0.84
8.62 ±0.96
0.4036
Albumin (g) Males
3.15 ±0.73
3.15 ± 1.33
3.08 ± 1.30
2.95 ± 1.12
0.2279
Albumin (g) Females
3.22 ± 1.28
3.17 ± 1.03
2.58 ± 1.28
3.60 ± 1.17
0.2408
Creatinine (mg/dL) Males
41.24 ±8.94
41.35 ± 11.28
40.79 ± 9.30
43.61 ± 13.10
0.3982
Creatinine (mg/dL) Females
62.54 ± 10.66
61.60 ±7.07
67.11 ± 10.86
59.71 ±7.51
0.1641
This document is a draft for review purposes only and does not constitute Agency policy.
B-120 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-33 (Continued): Characteristics and quantitative results for Korsak et al.
	("2000b") 			
Urea (mg/dL) Males
38.7 ±4.5
38.1 ±9.1
36.9 ±4.1
41.7 ±7.5
0.1145
Urea (mg/dL) Females
42.0 ±5.5
43.5 ±4.4
40.0 ± 4.3
39.0 ±29
0.4718
Calcium (mg/dL) Males
10.6 ±0.6
10.7 ±0.8
10.8 ±0.7
10.9 ±0.5
0.2449
Calcium (mg/dL) Females
11.1 ±0.8
11.7 ±0.3
11.8 ±0.2
11.8 ±0.7
0.3011
Phosphorus (mg/dL) Males
8.60 ±0.95
8.26 ±0.60
9.19 ±0.88
9.41 ±0.55
0.1580
Phosphorus (mg/dL) Females
6.56 ±0.70
6.25 ± 1.17
6.41 ± 1.02
7.18 ± 1.09
0.4050
Sodium (mmol/L) Males
143.9 ±2.1
144.1 ± 1.5
143.9 ± 25
144.8 ± 24
0.4950
Sodium (mmol/L) Females
144.0 ± 1.5
143.8 ± 1.3
142.7 ± 1.3
143.8 ± 1.4
0.3628
Potassium (mmol/L) Males
4.70 ±0.35
4.45 ±0.28
4.75 ±0.37
4.97 ±0.56
0.2907
Potassium (mmol/L) Females
4.52 ±0.41
4.51 ±0.43
4.28 ±0.41
4.37 ±0.34
0.4108
Chloride (mmol/L) Males
107.3 ± 2.3
107.7 ±4.3
106.8 ± 1.8
106.5 ± 1.9
0.4353
Chloride (mmol/L) Females
108.1 ± 3.2
108.1 ± 1.5
107.1 ± 1.3
107.2 ± 23
0.0601
Observation
Exposure concentration (mg/mB)
[Dose group ID]
0
[1]
123
[2]
492
[3]
1230
[4]
Comparison to
controls'
Trend
testb
Proliferation of peribronchial
lymphatic tissue (0-3)kMales
2.01 (23.4)m
1.2 (11.5)
1.8(22.0)
2.0(23.5)
1-2*
p = 0.2
Proliferation of peribronchial
lymphatic tissue (0-3)Females
2.4(22.8)
1.3 (12.1)
1.5 (16.4)
1.3 (22.3)
1-2**; 1-3
p = 0.2
Formation of
lymphoepithelium in bronchii
(0-3) Males
1.5 (23.9)
0.9 (14.9)
1.0(16.0)
1.5 (25.7)
1-3*; 1-4**
p = 0.3
Formation of
lymphoepithelium in bronchii
(0-3) Females
1.8(27.9)
0.7 (11.1)
1.1(16.9)
1.5 (23.8)

p = 0.3
Goblet cells (0-3) Males
1.8(18.6)
1.5 (14.5)
2.5 (28.5)
1.8(18.2)

p = 0.18
Goblet cells (0-3) Females
1.3 (11.9)
1.6 (16.9)
2.0(23.1)
2.4(28.4)
1-3*; 1-4**
p = 0.001
Interstitial lymphocytic
infiltration (0-3) Males
0.4(18.0)
0.1 (14.1)
0.4(18.0)
1.5 (31.0)
1-4*
p = 0.006
Interstitial lymphocytic
infiltration (0-3) Females
1.2 (23.7)
0.6 (15.3)
0.8(17.9)
1.1(22.9)

T3
II
O
Alveolar macrophages (0-3)
Males
0.9(17.9)
0.9 (17.9)
1.2 (22.6)
1.2 (21.7)

p = 0.15
Alveolar macrophages (0-3)
Females
1.5 (26.1)
1.1 (21.1)
0.5 (17.8)
0.7 (14.8)

p = 0.01
This document is a draft for review purposes only and does not constitute Agency policy.
B-121 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-33 (Continued): Characteristics and quantitative results for Korsak et al.
	("2000b") 				
Bronchitis and broncho-
pneumonia (0-4) Males
0.5 (20.1)
0.2 (16.6)
0.8(23.8)
0.7(19.5)

p = 0.3
Bronchitis and broncho-
pneumonia (0-4) Females
0.2(17.6)
0.4 (22.5)
0.2 (17.5)
0.6 (21.8)

p = 0.3
Cumulative score of all
individual Males
7.1(19.8)
4.8 (11.2)
7.7 (24.2)
8.7 (25.8)

p = 0.01
Cumulative score of all
individual Females
8.4(24.9)
5.7 (13.5)
6.5 (16.8)
8.2 (24.6)
1-2*
T3
II
O

Health Effect at LOAEL
NOAEL
LOAEL
Pulmonary lesions
492 mg/m3
1230 mg/m3
Comments: The observed inflammatory lesions are coherent with observations of increased inflammatory cell populations in
bronchoalveolar lavage fluid due to 1,2,4-TMB exposure in Korsak et al. (1997). The authors did not report the incidences of
pulmonary lesions, but rather the results of the Kruskall-Wallis test. This makes it difficult to interpret the dose-response
relationship and limits analysis of these endpoints to the NOAEL/LOAEL method for determining a POD, rather than using BMD
modeling.
aEffects measured in rats exposed to 1,230 mg/mB 2 weeks after termination of exposure,
p-value reported from Jonckheere's trend test
cred blood cells,
white blood cells,
Aspartate aminotransferase,
alanine aminotransferase,
galkaline phosphatase,
sorbitol dehydrogenase,
'y-glutamyltransferase,
J Reports the results of pair-wise statistical significance of exposure groups compared to controls (i.e., 1-3 would indicate that the
492 mg/mB was statistically significantly different from controls)
kgrading system (0-4, 0-3; see Additional study details above)
1 mean
m results presented as ranges of the Kruskal-Willis test.
*, ** Statistically significant from controls at p < 0.05 and 0.01, respectively.
Source: Korsak et al. (2000b).
This document is a draft for review purposes only and does not constitute Agency policy.
B-122 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-34. Characteristics and quantitative results for Lammers et al. (2007)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
WAG/RijCR/BR
Wistar rats
M
8/group
Inhalation (8 hr/day
for 3 consecutive
days)
0, 600, 2,400, or 4,800
mg/m3l,2,4-TMB (as a
constituent of WS)
3 days
Additional study details
•	Rats were exposed to 1,2,4-TMB as a constituent of WS at concentrations of 0, 600, 2,400, or 4,800 mg/m3 for 3
days. Several tests were conducted to evaluate impact of WS on CNS. These included tests of observation,
spontaneous motor activity and learned visual discrimination.
•	White spirit was found to affect performance and learned behavior in rats.
Observation
Functional observations and physiological parameters in rats following exposure to
WS (exposure concentration mg/mB)
0
600
2,400
4,800
Functional observation battery (mean ± SD)
Gait score3
Before first 8 hr
exposure
1.00 ± 0.00
1.00 ± 0.00
1.00 ± 0.00
1.00 ± 0.00
After first 8 hr
exposure
1.00 ± 0.00
1.00 ± 0.00
1.13 ±0.13
1.25 ±0.16
After third 8 hr
exposure
1.00 ± 0.00
1.00 ± 0.00
1.00 ± 0.00
1.00 ± 0.00
Click response"
Before first 8 hr
exposure
2.13 ±0.13
2.63 ±0.18
2.38 ±0.18
2.50 ±0.19
After first 8 hr
exposure
2.88 ±0.13
2.50 ±0.19
2.75 ±0.37
2.63 ±0.18
After third 8 hr
exposure
2.13 ±0.13
3.25 ±0.31*
2.88 ±0.23
2.75 ±0.25
Physiological parameters (mean ± SD)
Body weight (g)
Before first 8 hr
exposure
270.0 ±2.61
269.2 ±2.48
273.3 ±3.52
272.8 ±2.20
After first 8 hr
exposure
279.7 ±2.53
277.7 ±3.11
278.0 ±3.21**
273.8 ±2.51***
After third 8 hr
exposure
280.9 ± 2.68
278.4 ± 2.44
275.9 ±2.83***
268.5 ±2.67***
Body temperature (°C)
Before first 8 hr
exposure
37.60 ±0.34
37.33 ±0.39
37.49 ±0.39
37.29 ±0.37
After first 8 hr
exposure
36.41 ±0.05
36.25 ±0.12
36.16 ±0.11
35.95 ±0.21
After third 8 hr
exposure
36.60 ±0.10
36.44 ±0.17
36.25 ±0.05
36.11 ±0.09**
This document is a draft for review purposes only and does not constitute Agency policy.
B-123
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-34 (Continued): Characteristics and quantitative results for Lammers etal.
	(2007)	
3000
to
I 2500
c
re
-------
Toxicological Review ofTrimethylbenzene
Table B-34 (Continued): Characteristics and quantitative results for Lammers etal.
	(2007)	
Number of lever response latencies >6 sec
Before first 8 hr
exposure
3.88 ±0.90
5.25 ±0.84
3.25 ±0.45*
5.62 ±0.92**
After first 8 hr
exposure
5.00 ± 1.10
7.62 ± 1.83
11.12 ±0.85*
25.75 ±5.05**
After second 8 hr
exposure
4.38 ±0.96
5.62 ±0.78
5.00 ±0.65*
12.25 ±3.80**
After third 8 hr
exposure
7.38 ±2.07
6.88 ± 1.16
10.88 ± 1.96*
17.50 ±2.76**
One day after third 8
hr exposure
4.62 ± 1.31
4.38 ± 1.07
3.75 ±0.70*
6.50 ± 1.86**
Drink response latency (sec)
Before first 8 hr
exposure
0.35 ±0.04
0.29 ±0.03
0.36 ±0.03
0.32 ±0.02*
After first 8 hr
exposure
0.37 ±0.04
0.31 ±0.03
0.39 ±0.02
0.52 ±0.04*
After second 8 hr
exposure
0.36 ±0.04
0.28 ±0.03
0.33 ±0.02
0.39 ±0.04*
After third 8 hr
exposure
0.38 ±0.05
0.32 ± 0.04
0.39 ±0.02
0.43 ±0.07*
One day after third 8
hr exposure
0.36 ±0.03
0.31 ±0.02
0.34 ± 0.02
0.33 ±0.04*

Health Effect at LOAEL
NOAEL
LOAEL
n/a
n/a
n/a
Comments: Exposure to 1,2,4-TMB was via WS, which is comprised of additional substances. LOAEL and NOAEL cannot be extracted
from this study because other constituents of the WS mixture may confound results.
aGait score indicates the severity of gait changes and is scored as 1 (normal) to 4 (severely abnormal).
bClick response was scored as 0 (no reaction) to 5 (exaggerated reaction).
cData for parameters that did not show statistically significant group differences are not shown; statistical analysis: repeated
measures ANCOVA + pairwise group comparisons.
*,**,*** Statistically significant from controls at p < 0.05, p < 0.01, and p < 0.001 respectively.
Source: Lammers et al. (2007)
This document is a draft for review purposes only and does not constitute Agency policy.
B-125 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-35. Characteristics and quantitative results for Lutz et al. (2010)
Study design
Species Sex N
Exposure route	Dose range
Exposure duration
Wistar rats M
6-8 rats
per dose

4 weeks
Additional study details
•	Animals were exposed to 1,2,3- or 1,2,4-TMB in 1.3 m3 dynamic inhalation exposure chambers for 6 hrs/day,
5 days/week for 4 weeks. Food and water was provided ad libitum.
•	Animals were randomized and assigned to the experimental groups.
•	Behavioral sensitivity to amphetamine was measured via test of open-field locomotor activity.
•	Differences were observed between 1,2,3- and 1,2,4-TMB exposed rats, with 1,2,3-TMB-exposed rats
displaying greater amphetamine sensitization than 1,2,4-TMB exposed rats.
This document is a draft for review purposes only and does not constitute Agency policy.
B-126
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table B-35 (Continued): Characteristics and quantitative results for Lutz et al.
	(2010)	
; 120 n
80
Session 1
(before AMPH sensitization)
ft/fh. iv ifii
a
5x2.5 mg/kg AMPH
ft
u
Session 2
(after AMPH sensitization)
u

u
I
Control	HEM	HEM	HEM
25ppm 100 ppm 250 ppm
Control
HEM	HEM	HEM
25 ppm	100 ppm 250 ppm
~ Block 1	Block 2 ~ Block 3 ~ Block 4 ~ Block 5
Block 6
Block 1 — control (preinjection) activity, block 2 — activity after the SAL injection, blocks 3,4, 5 and 6 — activity during successive 30 min sections
after AMPH (0.5 mg/kg) injection.
ANOVA: group effects: F (3.24) =9.80; P = 0.0002; session effects: F (1.24) =34.22; P = 0.0000; interaction: F (3.24) =20.64; P = 0.0000.
* P < 0.05 — compared to post SAL measurement.
** P < 0.05 — compared to control 0 in the same session.
*** P < 0.05 — compared to corresponding measure before sensitization.
The bars represent mean values and SEM of the ambulatory activity (distance in metres) in successive 30 min blocks in the rats exposed
to hemimellitene on the locomotor response to AMPH challenge before (session 1) and 14 days after (session 2) a repeated
(2.5 mg'kg, l/dayx5 days) AMPH treatment.
Figure 1. Diagram illustrating the effect of prior exposure to 1,2,3-TMB on the locomotor response (all
measurements) to the amphetamine challenge before (session 1) and 14 days after (session 2) a repeated
(2.5 mg/kg, 1/day x 5 day) amphetamine treatment.
Source: Lutz et al. (2010)
This document is a draft for review purposes only and does not constitute Agency policy,
B-127 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table B-35 (Continued): Characteristics and quantitative results for Lutz et al.
	(2010)	
Session 1
(before AMPH sensitization)
Session 2
(after AMPH sensitization)
CO P
32 ^
£ R
: 280
>
J 240
200 -
160 -
120 -
80 -
40
0
X
5x2.5 mg/kgAMPH
JL
_L
Q
Control
HEM	HEM	HEM
25ppm 100 ppm 250 ppm
Control HEM	HEM
25 ppm 100 ppm 250 ppm
* P < 0.05 — compared to control. " P < 0.05 — compared to corresponding measure before sensitization.
Bars represent mean values and SEM of the cumulated locomotor activity (distance in metres) during the 2-hour measurement
following AMPH (0.5 mg/kg) challenge.
Figure 2. Diagram illustrating the effect of prior exposure to 1,2,3-TMB on the locomotor response (pooled
measurements) to the amphetamine challenge before (session 1) and 14 days after (session 2) a repeated
amphetamine treatment (2.5 mg/kg, 1/day x 5 days).
Source: Lutz et al. (2010)
This document is a draft for review purposes only and does not constitute Agency policy,
B-128 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-35 (Continued): Characteristics and quantitative results for Lutz etal.
	(2010)	
Figure 3. Diagram illustrating the effect of prior exposure to 1,2,4-TMB on the locomotor response (all
measurements) to the amphetamine challenge before (session 1) and 14 days after (session 2) a repeated
(2.5 mg/kg, 1/day x 5 days) amphetamine treatment. Remaining notations are the same as in Figure 1.
:• 120
40
Session 1
(before AMPH sensitization)
5x2.5 mg/kg AMPH

Session 2
(after AMPH sensitization)


Control
PS	PS	PS
25ppm 100 ppm 250 ppm
Control
PS
25 ppm
PS
100 ppm
~ Block 1 m Block 2 ~ Block 3 ~ Block 4 ~ Block 5
Block 6
PS
250 ppm
ANOVA: group effects: F (3.25) =8.90; P = 0.004. Session effects: F (1.25) =30.91; P = 0.0000. Interaction: F (3.25) =29.48; P = 0.0000.
* P < 0.05 — compared to post SAL measurement.
** P < 0.05 — compared to control 0 in the same session.
*** P < 0.05 — compared to corresponding measure before sensitization.
The bars represent mean values and SEM of the ambulatory activity (distance in metres) in successive 30 min blocks.
Source: Lutz et al. (2010)
This document is a draft for review purposes only and does not constitute Agency policy.
B-129 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-35 (Continued): Characteristics and quantitative results for Lutz etal.
	(2010)	
Figure 4. Diagram illustrating the effect of prior exposure to 1,2,4-TMB on the locomotor response (pooled
measurements) to amphetamine challenge before (session 1) and 14 days after (session 2) a repeated
amphetamine treatment (2.5 mg/kg, 1/day x 5 days).
-= =280
S E
8 =>
C CNJ
1 g240
S O
,2
200
160
120
80
40
Session 1
(before AMPH sensitization)
Session 2
(after AMPH sensitization)
5x2.5 mg/kg AMPH
_L
x

x
X
Control
PS
25ppm
PS
100 ppm
PS
250 ppm
Control
PS
25 ppm
PS
PS
100 ppm 250 ppm
* P < 0.05 — compared to control. ** P < 0.05 — compared to corresponding measure before sensitization.
Bars represent mean values and SEM of the cumulated locomotor activity (distance in metres) during the 2-hour measurement
following AMPH (0.5 mg/kg) challenge.
Source: Lutz et al. (2010)
Health Effect at LOAEL
NOAEL
LOAEL
Increased sensitivity to
amphetamine as measured by
open-field locomotion	
0 ppm
25 ppm (123 mg/m )
1,2,4-TMB or 1,2,3-TMB
Comment: This study observed increased amphetamine sensitization, particularly in rats exposed to 100 ppm (492 mg/m )
1,2,3-TMB, and provided evidence for differences in toxicity between different TMB isomers. Control group for 1,2,4-TMB also
showed statistically significant increase in locomotor activity after receiving amphetamine treatment.
Source: Lutz et al. (2010)
This document is a draft for review purposes only and does not constitute Agency policy.
B-130 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-36. Characteristics and quantitative results for Maltoni et al. (1997)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Sprague-
Dawley rats:
CRC/BT
M
50 males,
50 females
per group
Stomach tube (in
olive oil)
0 or 800 mg/kg BW
1,2,4-TMB
4 days/week for 104 weeks
Additional study details
•	Rats were exposed to 1,2,4-TMB for 2 years via stomach tube administration 4 days/week.
•	Animals were 7 weeks old at start of experiments.
•	Systematic necropsy was conducted upon animal death.
•	A slight increase in total number of tumors was detected amongst males and females, and an increase in the
number of head cancers in males was also observed.
Observation
Long-term carcinogenicity of 1,2,4-TMB
0 mg/kg
800 mg/kg
Total number of tumors
Males
Total benign and
malignant tumors
54.0
62.0
Malignant tumors
24.0
26.0
No. malignant
tumors/100 rats
26.0
34.0
Females
Total benign and
malignant tumors
70.0
66.0
Malignant tumors
22.0
24.0
No. malignant
tumors/100 rats
22.0
32.0
Both sexes
Total benign and
malignant tumors
62.0
64.0
Malignant tumors
23.0
25.0
No. malignant
tumors/100 rats
24.0
33.0
This document is a draft for review purposes only and does not constitute Agency policy.
B-131
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-36 (Continued): Characteristics and quantitative results for Maltoni et al.
(1997)
Head cancers
Males
Zymbal gland cancer
2.0
4.0
Ear duct cancer
-
2.0
Neuroesthesio-
epitheliomas
-
2.0
Oral cavity cancers
-
2.0
Total head cancers
2.0
10.0
Females
Zymbal gland cancer
2.0
2.0
Ear duct cancer
2.0
-
Neuroesthesioepi-
theliomas
-
4.0
Oral cavity cancers
2.0
-
Total head cancers
6.0
6.0
Both sexes
Zymbal gland cancer
2.0
3.0
Ear duct cancer
1.0
1.0
Neuroesthesio-
epitheliomas
-
3.0
Oral cavity cancers
1.0
1.0
Total head cancers
4.0
8.0
Health Effect at LOAEL
NOAEL
LOAEL
Various malignant and non-
malignant cancers
n/a
800 mg/kg
Comments: Neuroesthesioepithelioma is uncommon in Sprague-Dawley rats, although there were increases in the number of
neuroesthesioepithelioma in both males and females. Only one dose level was tested (800 mg/kg), making any determination of
dose-response impossible. Statistical significance of data not provided, although post-hoc statistical tests performed by EPA failed to
observe any statistical increase in tumors.
Source: Maltoni et al. (1997)


This document is a draft for review purposes only and does not constitute Agency policy.
B-132 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-37. Characteristics and quantitative results for McKee et al. (2010)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
8 rats per
group
Inhalation
0, 125,1,250, or 5,000
mg/m31,2,4-TMB
8 hrs/day for
3 consecutive days
Additional study details







• Animals were exposed to 1,2,4-TMB for 8 hrs/day for 3 days in modified H1000 inhalation chambers.
• Animals were randomized and assigned to the experimental groups.



• Test on neurobehavioral effects were conducted prior to, during, and after exposure period.

• Motor activity was affected on the third day of exposure in the highest exposure group, although brain
concentrations of 1,2,4-TMB were lower than on previous days.




Exposure concentration 1,2,4-TMB (mg/mB)
Observation

0
125
1,250
5,000
Results of functional and motor activity observations
Forelimb grip strength (g)

One-day pre-exposure
1,107 ± 41.2
1,065 ± 52.3
1,223 ± 25.9
1,090 ± 47.0
First 8 hr exposure
1,064 ± 39.9
814 ±91.7*
1,059 ± 59.8
1,023 ± 55.7
Third 8 hr exposure
908 ±56.1
847 ± 64.3
956 ±67.7
1,156 ± 68.7*
Total distance traveled (cm)

One-day pre-exposure
3,773 ± 120
3,598 ± 301
3,543 ± 167
3,575 ± 119
First 8 hr exposure
2,479 ± 110
3,048 ± 257
2,125 ± 171
1,897 ± 200
Third 8 hr exposure
2,459 ± 118
2,740 ± 226
1,967 ±316
1,172 ± 226*
Number of movements

One-day pre-exposure
1,054 ±31
999 ± 80
990 ± 44
998 ± 32
First 8 hr exposure
697 ± 29
848 ± 66
600 ± 48
529 ± 53
Third 8 hr exposure
687 ± 31
744 ± 56
541 ± 82
329 ±61*




Exposure concentration 1,2,4-TMB (mg/mB)
Observation



0
125
1,250
5,000
Visual discrimination performance testing (means ± SD)
Trials3
One-day pre-exposure
100 ± 0.0
100 ± 0.0
100 ± 0.0
100 ± 0.0
First 8 hr exposure
100 ± 0.0
100 ± 0.0
100 ± 0.0
99.13 ±0.88
Third 8 hr exposure
100 ± 0.0
100 ± 0.0
100 ± 0.0
100 ± 0.0
One-day post-exposure
100 ± 0.0
100 ± 0.0
100 ± 0.0
100 ± 0.0
Percentage reinforcements obtained15
One-day pre-exposure
99.88 ±0.13
99.88 ±0.13
99.88 ±0.13
100 ± 0.0
First 8 hr exposure
100 ± 0.0
100 ± 0.0
99.38 ±0.63
99.74 ±0.17
Third 8 hr exposure
99.63 ±0.26
99.63 ±0.26
99.63 ±0.38
100 ± 0.0
One-day post-exposure
99.63 ±0.26
99.88 ±0.13
99.88 ±0.13
100 ± 0.0
This document is a draft for review purposes only and does not constitute Agency policy.
B-133
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-37 (Continued): Characteristics and quantitative results for McKee etal.
(2010)
Discrimination ratio0
One-day pre-exposure
0.81 ±0.84
0.84 ± 0.03
0.83 ± 0.02
0.83 ± 0.03
First 8 hr exposure
0.86 ± 0.02
0.91 ±0.03
0.91 ±0.01
0.95 ±0.01*
Third 8 hr exposure
0.89 ± 0.02
0.88 ± 0.03
0.94 ±0.01
0.95 ±0.02
One-day post-exposure
0.87 ± 0.03
0.89 ±0.03
0.92 ±0.02
0.88 ±0.03
Percentage inter-trial intervals responded tod
One-day pre-exposure
12.88 ± 2.00
10.13 ± 1.56
10.75 ± 1.94
10.38 ± 1.84
First 8 hr exposure
12.50 ±2.12
8.88 ± 2.03
11.50 ±2.60
10.19 ± 1.28
Third 8 hr exposure
12.00 ± 1.65
8.88 ± 2.24
8.25 ± 1.71
5.75 ± 1.39
One-day post-exposure
10.88 ± 1.39
10.63 ± 1.81
11.25 ±0.92
8.50 ± 1.40
Repetitive errors6
One-day pre-exposure
8.25 ±3.71
7.63 ± 1.70
10.75 ±2.73
7.25 ± 1.75
First 8 hr exposure
2.00 ± 0.50
3.25 ± 1.47
4.63 ± 1.58
1.88 ±0.67
Third 8 hr exposure
2.63 ± 1.70
4.75 ± 1.81
3.00 ±0.78
1.25 ±0.73
One-day post-exposure
4.75 ±2.81
2.75 ± 1.35
4.63 ± 3.09
4.13 ± 1.38
Repetitive inter-trial responses'
One-day pre-exposure
3.63 ± 1.02
5.88 ± 1.33
7.25 ± 1.93
3.25 ± 1.35
First 8 hr exposure
6.13 ± 1.73
3.88 ± 1.22
5.63 ± 1.97
8.38 ±2.50
Third 8 hr exposure
7.25 ± 1.24
3.25 ±0.88
2.25 ± 1.52*
1.63 ±0.98*
One-day post-exposure
6.63 ± 1.94
2.88 ±0.83
5.13 ± 1.54
2.63 ±0.68
Trial response latency8
One-day pre-exposure
1.83 ±0.18
2.25 ±0.55
2.06 ± 0.40
2.28 ±0.43
First 8 hr exposure
1.70 ±0.18
2.38 ±0.43
2.52 ±0.40
3.91 ±0.73*
Third 8 hr exposure
1.91 ±0.23
2.69 ±0.69
2.75 ±0.94
1.82 ±0.13
One-day post-exposure
1.68 ±0.16
2.70 ±0.60
2.18 ±0.73
1.45 ± 0.06
Standard deviation of response latency
One-day pre-exposure
2.16 ±0.38
3.82 ± 1.57
3.33 ± 1.42
4.65 ±2.23
First 8 hr exposure
2.06 ±0.38
3.64 ± 1.32
4.19 ± 1.65
7.33 ±3.43
Third 8 hr exposure
2.74 ±0.71
4.03 ± 1.50
5.25 ± 3.04
2.34 ± 0.40
One-day post-exposure
1.84 ±0.38
5.95 ±2.40
5.88 ±4.21
1.81 ±0.38
Latency <2 sech
One-day pre-exposure
61.75 ±4.55
70.13 ±2.23
67.75 ±66.88
66.88 ±3.22
First 8 hr exposure
68.50 ±3.84
69.75 ±3.75
65.76 ±3.13
52.13 ±3.96
Third 8 hr exposure
70.38 ±4.34
64.13 ±4.35
74.88 ± 1.75
79.00 ± 2.32
One-day post-exposure
69.38 ±2.98
67.63 ±3.20
78.13 ±3.05
78.00 ± 2.34
Latency >6 sec1
One-day pre-exposure
3.38 ±0.71
5.38 ± 1.48
4.63 ± 1.15
4.00 ± 1.05
First 8 hr exposure
3.88 ±0.58
5.00 ± 1.69
6.00 ± 1.34
10.63 ± 1.80*
Third 8 hr exposure
4.25 ±0.98
5.63 ± 2.44
5.63 ± 1.92
3.13 ±0.61
One-day post-exposure
2.13 ±0.67
6.00 ± 1.68
3.38 ± 1.40
1.88 ±0.35
This document is a draft for review purposes only and does not constitute Agency policy.
B-134 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-37 (Continued): Characteristics and quantitative results for McKee etal.
	(2010)	
Drink response latencyJ
One-day pre-exposure
0.29 ±0.01
0.32 ±0.02
0.38 ±0.03*
0.33 ±0.02
First 8 hr exposure
0.26 ±0.01
0.30 ± 0.02
0.43 ± 0.03*
0.49 ±0.03*
Third 8 hr exposure
0.30 ± 0.02
0.32 ±0.03
0.37 ±0.02
0.34 ± 0.03
One-day post-exposure
0.27 ±0.01
0.34 ±0.03
0.36 ±0.03
0.30 ±0.02
Health Effect at LOAEL
NOAEL
LOAEL
n/a
n/a
n/a
Comments: This study observed alterations in a number of parameters, including forelimb grip strength, total distance traveled,
number of movements, and several visual discrimination performance tests. LOAEL and NOAEL cannot be determined because a
dose-response relationship was not apparent. Statistically significant results occurred in a low exposure group and not others, while
forelimb grip was found to be significantly increased in the highest exposure group on day 3. Acute duration of exposure (exposure
on 3 consecutive days). Generally, acute exposure studies have limited utility in quantitation of human health reference values.
aTotal number of trials completed during each session, maximum = 100.
bNumber of reinforcements obtained divided by the number of reinforcements delivered (xlOO).
cNumber of correct trial responses divided by the number of trial responses.
dThe number of inter-trial intervals in which at least 1 response was made divided by the total number of ITI (xlOO).
0The total number of incorrect trial responses following an initial incorrect response.
fThe total number of ITI responses following an initial ITI response.
gThe latency (seconds) to make a correct trial response.
hThe number of responses within 2 seconds.
'The number of responses taking more than 6 seconds.
JThe mean latency (seconds) to obtain reinforcement.
^Statistically significant from controls at p < 0.05.
Source: McKee et al. (2010)
This document is a draft for review purposes only and does not constitute Agency policy.
B-135 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-38. Characteristics and quantitative results for Saillenfait et al. (2005)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Sprague-
Dawley rats
F &
M
24 dams
per dose
Inhalation (6 hr/day
GD6-GD20)
0, 100, 300, 600, 900 ppm
(0, 492, 1,476, 2,952, or
4,428 mg/m3) 1,2,4-TMB;
0, 100, 300, 600,1,200 ppm
(0, 492, 1,476, 2,952, or
5,904 mg/m3) 1,3,5-TMB
Gestational days
GD6-GD20
Additional study details








• Animals were exposed to 1,2,4- or 1,3,5-TMB in 200 L glass/steel inhalation chambers for 6 hrs/day starting on GD6
and ending on GD20.
• Animals were randomized and assigned to the experimental groups.



• After GD20, dams were sacrificed and weighed, as were their uteri and any fetuses.


• Decreases in maternal body weight and fetal toxicity were observed.






Exposure concentration to 1,3,5-TMB
Observation


0 ppm
100 ppm
(492mg/ms)
300 ppm
(l,476mg/m3)
600 ppm
(2,952 mg/m3)
1,200 ppm
(5,904 mg/m3)
Maternal parameters
No. treated
24
24
24
24
24
No. (%) pregnant at
euthanization
21(87.5)
22 (91.7)
21(87.5)
17 (70.8)
18 (75.0)
No. deaths
0
0
0
0
0
Body weight (g) on day 6
274 ±17s
273 ± 16
274 ± 21
270 ± 17
275 ± 14
Body weight change (g)
Days 0-6
31 ± 11
31 ±8
31 ±7
29 ±8
28 ±8
Days 6-13
25 ± 12
29 ±4
23 ±6
16 ± 8**
10 ±7
Days 13-21
110 ± 14
109 ± 10
95 ±21*
80 ±20**
63 ±26**
Days 6-21
135 ± 15
138 ± 11
118 ± 24*
95 ± 24**
73 ±28**
Corrected weight gain3
29 ± 14
30 ±9
20 ± 12
7 ± 20**
-12±19**
Food consumption (g/day)
Days 0-6
22 ±2
22 ±3
22 ±2
22 ±2
23 ±2
Days 6-13
22 ±2
22 ±2
20 ± 1*
18 ± 2**
17 ± 2**
Days 13-21
26 ±2
25 ±2
24 ±2*
21 ± 3**
19 ± 3**
Days 6-21
24 ±2
24 ±2
22 ±2*
20 ± 2**
18 ±2**
This document is a draft for review purposes only and does not constitute Agency policy.
B-136
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-38 (Continued): Characteristics and quantitative results for Saillenfait et al.
(2005)
Observation
Exposure concentration to 1,3,5-TMB
0 ppm
100 ppm
(492mg/ms)
300 ppm
(l,476mg/m3)
600 ppm
(2,952 mg/m3)
1,200 ppm
(5,904 mg/m3)
Gestational parameters
All litters"
21
22
21
17
18
No. of corpora lutea per dam
15.3 ± 1.5s
15.4 ± 1.7
15.5 ± 1.7
14.9 ±2.1
15.2 ± 1.5
Mean no. of implantation
sites per litter
14.9 ± 1.5
14.9 ± 1.8
14.5 ± 3.4
13.0 ±5.1
13.6 ±3.7
Mean % post-implantation loss
per litter0
4.8 ±4.2
3.9 ±4.3
6.8 ± 8.5
1.6 ±3.7
4.4 ±6.9
Mean % dead fetuses per
litter
0.0 ± 0.0
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
Mean % resorption sites per
litter
4.8 ±4.2
3.9 ±4.3
6.3 ±6.5
1.6 ±3.7
4.4 ±6.9
Live littersd
21
22
21
17
18
Mean no. of live fetuses per
litter
14.1 ± 1.6
14.3 ± 1.7
13.4 ±3.4
12.8 ±5.0
13.1 ±3.7
Mean % male fetuses per
litter
49.3 ± 13.5
48.2 ± 16.3
52.1 ± 18.1
51.1 ±20.9
48.5 ± 18.2
Fetal body weight (g)
All fetuses
5.64 ±0.35
5.61 ±0.24
5.43 ± 0.45
5.36 ±0.68
4.98 ±0.56**
Male fetuses
5.80 ±0.41
5.76 ±0.27
5.50 ±0.31
5.39 ±0.55*
5.10 ±0.57**
Female fetuses
5.50 ±0.32
5.47 ±0.21
5.27 ± 0.47
5.18 ±0.68
4.81 ±0.45**
Observation
Exposure concentration to 1,3,5-TMB
0 ppm
100 ppm
(492mg/ms)
300 ppm
(l,476mg/m3)
600 ppm
(2,952 mg/m3)
1,200 ppm
(5,904 mg/m3)
Fetal variations and malformations
Total no. fetuses examined (litters)
External
297 (21)
314 (22)
282 (21)
217 (17)
236 (18)
Visceral
149 (21)
157 (22)
141 (20)
109 (15)
118 (18)
Skeletal
148 (21)
157 (22)
141 (21)
108 (17)
118 (18)
Malformations
Diaphragmatic hernia
0
1(1)
0
1(1)
0
Multiple skeletal
malformations6
1(1)
0
0
0
0
External variations
0
0
0
0
0
Club foot (bilateral)
0
1(1)
0
0
0
Visceral variations
Dilated renal pelvis
2(2)
0
5(4)
0
2(2)
Distended ureter
12 (9)
14 (8)
18 (8)
5(3)
11(6)
This document is a draft for review purposes only and does not constitute Agency policy.
B-137 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-38 (Continued): Characteristics and quantitative results for Saillenfait et al.
(2005)
Skeletal variations
Fifth sternebrae incomplete
ossification or unossifiedf
2(2)
2(2)
7(4)
7(5)
12 (7)
Fourth sternebrae, split
0
0
0
0
1(1)
Cervical rib, rudimentary
2(2)
0
5(5)
5(3)
2(2)
Fourteenth rib,
supernumerary
11 (8)
9(6)
11(6)
15 (8)
17 (8)
Thoracic vertebra centra,
incomplete ossification
10 (5)
8(6)
10 (7)
9(7)
9(7)
Observation
Exposure concentration to 1,2,4-TMB
0 ppm
100 ppm
(492mg/ms)
300 ppm
(l,476mg/m3)
600 ppm
(2,952 mg/m3)
900 ppm
(4,428 mg/m3)
Maternal parameters
No. treated
25
24
24
24
24
No. (%) pregnant at
euthanization
24 (96.0)
22 (91.7)
22 (91.7)
22 (91.7)
24 (100)
No. deaths
0
0
0
0
0
Body weight (g) on day 6
271 ± 18s
272 ±21
272 ± 22
275 ± 19
269 ± 18
Body weight change (g)

Days 0-6
27 ±8
28 ±6
28 ±7
28 ± 12
24 ±8
Days 6-13
27 ±8
27 ±6
26 ±6
19 ± 8**
14±12**
Days 13-21
105 ± 28
98 ± 16
100 ± 20
97 ± 17
82 ±14**
Days 6-21
131 ±33
124 ± 18
126 ± 24
116 ± 23
95 ±19**
Corrected weight gain3
29 ± 12
31 ± 14
27 ± 12
15 ±17**
0 ± 14**
Food consumption (g/day)
Days 0-6
23 ±2
23 ±2
23 ±2
23 ±3
23 ±3
Days 6-13
21 ± 3
20 ±2
20 ±2
18 ± 2**
17 ± 2**
Days 13-21
26 ±3
25 ±2
24 ±2
23 ± 3**
22 ± 3**
Days 6-21
24 ±3
23 ±2
22 ±2
21 ± 3**
20 ± 2**
Observation
Exposure concentration to 1,2,4-TMB
0 ppm
100 ppm
(492mg/ms)
300 ppm
(l,476mg/m3)
600 ppm
(2,952 mg/m3)
900 ppm
(4,428 mg/m3)
Gestational parameters
All litters"
24
22
22
22
24
No. of corpora lutea per dam
15.4 ± 2. lg
15.2 ± 1.3
15.2 ±2.1
15.8 ± 1.7
15.7 ±2.5
Mean no. of implantation
sites per litter
14.2 ±3.3
13.7 ±2.9
14.1 ±3.2
14.9 ±2.4
15.0 ±2.4
Mean % post-implantation loss
per litter0
10.0 ±22.1
8.6 ±8.9
5.8 ± 6.8
5.0 ±5.7
5.4 ±6.7
Mean % dead fetuses per
litter
0.0 ± 0.0
0.3 ± 1.5
0.0 ±0.0
0.0 ±0.0
0.0 ±0.0
Mean % resorption sites per
litter
10.0 ±22.1
8.3 ±9.1
5.8 ± 6.8
5.0 ±5.7
6.4 ±6.7
This document is a draft for review purposes only and does not constitute Agency policy.
B-138 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-38 (Continued): Characteristics and quantitative results for Saillenfait et al.
(2005)
Live littersd
23
22
22
22
24
Mean no. of live fetuses per
litter
13.9 ±2.5
12.5 ±3.0
13.3 ±3.2
14.1 ±2.3
14.3 ±2.6
Mean % male fetuses per
litter
46.6 ± 17.1
46.0 ± 14.1
49.9 ± 13.4
46.2 ± 15.4
50.4 ± 16.2
Fetal body weight (g)
All fetuses
5.71 ±0.34
5.64 ±0.31
5.56 ±0.47
5.40 ±0.39*
5.60 ±0.40**
Male fetuses
5.86 ±0.34
5.79 ±0.30
5.72 ±0.49
5.55 ±0.48*
5.20 ±0.42**
Female fetuses
5.57 ±0.33
5.51 ±0.31
5.40 ± 0.45
5.28 ±0.40*
4.92 ±0.40**
Observation
Exposure concentrations to 1,2,4-TMB
0 ppm
100 ppm
(492mg/ms)
300 ppm
(l,476mg/m3)
600 ppm
(2,952 mg/m3)
900 ppm
(4,428 mg/m3)
Fetal variations and malformations
Total no. fetuses examined (litters)
External
319 (23)
275 (22)
293 (22)
310 (22)
342 (24)
Visceral
160 (23)
137 (22)
147 (22)
155 (22)
171 (24)
Skeletal
159 (23)
138 (22)
146 (22)
155 (22)
171 (24)
Malformations

Diaphragmatic hernia
0
0
1(1)
0
1(1)
Multiple skeletal
malformations6
0
0
0
1(1)
0
External variations
Club foot (bilateral)
3(3)
0
0
0
0
Visceral variations
Dilated renal pelvis
3(3)
3(3)
3(3)
3(3)
3(2)
Distended ureter
7(4)
5(3)
8(5)
8(5)
2(2)
This document is a draft for review purposes only and does not constitute Agency policy.
B-139 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-38 (Continued): Characteristics and quantitative results for Saillenfait et al.
	(2005)	
Skeletal variations
Third sternebrae, incomplete
ossification
0
1(1)
0
0
0
Fifth sternebrae incomplete
ossification or unossifiedf
1(1)
0
4(4)
5(4)
6(6)
Extra ossification site
0
1(1)
0
0
0
Cervical rib, rudimentary
1(1)
2(2)
0
3(2)
2(2)
Fourteenth rib,
supernumerary
25 (10)
13 (8)
18 (12)
21(10)
34 (16)
Thirteenth rib, short
(unilateral)
1(1)
0
0
0
0
Thoracic vertebral centra,
incomplete ossification
8(6)
4(4)
7(4)
6(6)
7(5)
Health Effect at LOAEL
NOAEL
LOAEL
Maternal toxicity: decrease in
maternal body weight and
food consumption
Developmental toxicity:
significant reduction in fetal
body weight
Maternal toxicity: 300 ppm (1,476 mg/m3)
for 1,3,5-TMB and 1,2,4-TMB
Fetal toxicity: 300 ppm (1,476 mg/m3) for
1,2,4- and 1,3,5-TMB
Maternal toxicity: 600 ppm (2,952 mg/m3)
for 1,3,5-TMB and 1,2,4-TMB
Fetal toxicity: 600 ppm (2,952 mg/m3) for
1,2,4-and 1,3,5-TMB
Comments: This study observed alterations in a number of maternal and fetal parameters, including decreased maternal and fetal
weight. Values reported by authors can be used to determine NOAEL and LOAEL. There was no investigation of pre-implantation
developmental toxicity due to 1,2,4-TMB or 1,3,5-TMB exposure. 1,2,3-TMB maternal or developmental toxicity not investigated.
aBody weight gain during GD6-GD21 minus gravid uterine weight.
includes all animals pregnant at euthanization.
Resorptions plus dead fetuses.
includes all animals with live fetuses at euthanization.
eRunt showing skeletal alterations including missing ribs, missing thoracic vertebrae, incomplete ossification of sternebrae and skull
bones.
fUnossified = alizarine red S negative.
gMean ±SD.
*, ** Statistically significant from controls at p < 0.05 and 0.01, respectively.
Source: Saillenfait et al. (2005)
This document is a draft for review purposes only and does not constitute Agency policy.
B-140 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table B-39. Characteristics and quantitative results for Tomas et al. (1999a)
Study design	
Species
Sex
Exposure route
Dose range
Exposure duration
WAG/Rij
Rats
M
6 rats per
dose
Oral (gavage, in
olive oil)
0, 2, 8, or 32 mrnol/kg BW
(240, 960, 3,840
mg/kg BW).
1,2,3-, 1,2,4-, and
1,3,5-TMB
Acute
Additional study details
•	1,2,3-, 1,2,4-, and 1,3,5-TMB were tested for their effects on electrocortical arousai by an electrocardiogram before
and after oral administration (in olive oil) of 0, 0.002, 0.008, or 0.032 mol/kg BW of each isomer.
•	Solvent concentration in peripheral blood was determined via head space gas chromatography.
•	All three TMB isomers were found to cause a slight increase in locomotor activity.
450-i
IT 300 -
-------
Toxicological Review ofTrimethylbenzene
Table B-39 (Continued): Characteristics and quantitative results for Tomas etal.
	("1999a")	
40-1
C/>
>
2 20 -
oil
I
I
TOLUENE
0.002 mol/kg
1 * *
* *
Ir1! - Fl
0.008 mol/kg 0.032 mol/kg
* * *
m
* * *
S, s, s2 s3 s0 s, s2 s3 s0 s, s2 s3 s0 s, s2 s3
Figure 2. Changes in number
of high-voltage spindle
episodes following acute
exposure to toluene and
1,2,3-, 1,2,4-, and 1,3,5-TMB
at doses of 0.002, 0.008, and
0.032 mol/kg.
40 -i
CO
>
2 20 -
40-|
2 20 -
HEMIMELLITENE
1
* *

* * ;
£
* ;
i-
S. s, S3 s, s, s2 s3 s0 s, s2 s3
PSEUDOCUMENE
5

;;
ii
S. S, s2 s3 s0 s, s2 s3
Source: Reproduced from
Tomas et al. (1999a)
40-i
C/D
>
° 20 -
a
-Q
E
3
E
0
MESITYLENE
* *
I

* *
s s s s
S„ s, s2 s3
s s s s
S0 - preinjection -	^
S, - 20 min postinjection - ~
52-40	min postinjection - ~
53-60	min postinjection - ~
* - p<0.001 compare to oil group
¦- p<0.001 compare to control
measurement
Health Effect at LOAEL
NOAEL
LOAEL
Abnormal electrocortical
stimulation
n/a
2 mmol/kg 1,2,3-TMB,
1,2,4-TMB, and 1,3,5-TMB
Comments: Exposures were of an acute duration, and therefore not suitable for reference value derivation. However, qualitatively,
this study provided evidence of CNS disturbances that, when considered together with short-term and subchronic neurotoxicity
studies, demonstrate that TMB isomers perturb the CIMS of exposed animals.
Source: Tomas et al. (1999a)
This document is a draft for review purposes only and does not constitute Agency policy,
B-142 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table B-40. Characteristics and quantitative results for Tomas et al. (1999b)
Study design
Species
Sex
Exposure route
Dose range
Exposure
duration
WAG/Rij rats
M
10 rats per
dose
Oral (in olive oil)
0, 8,16, or 32 mmol/kg
BW (960,1,920, or 3,850
mg/kg BW)
1,2,4-TMB, 1,2,3-TMB, or
1,3,5-TMB
Acute
Additional study details
•	1,2,3-, 1,2,4-, and 1,3,5-TMB were tested for their effects on locomotor activity by an open field test following
oral administration (in olive oil) of 0, 8,16, or 32 mmol/kg BW of all isomers.
•	All three TMB isomers were found to cause a slight increase in locomotor activity.
a
g>
0.008 mol/kg
- p<0.0001 compare to time point 1, 2
injection
M.
I IlI J *
_lA_
01234567
Time points
0.015 mol/kg
injection
_Li
y
* - pcQ.0001 compare to time point 1, 2, 3
Jitl	I;'n I	iff11	
Figure 1. Locomotor activity
following acute exposure to
toluene and TMB isomers at
doses of 0.008 mol/kg, 0.016
mol/kg, and 0.032 mol/kg.
Source: Reproduced from Tomas et
al. (1999b)
S 100
injection
1	2	3	4	5	6	?
Time points
0.032 mol/kg	* " P<0-0001 compare to time point 1
|j ii i l 11. 1 ¦ !
tKB l
Time points
control group (oil) pseudocumene i
homimellitene t
mesitylene en toluene i
Health Effect at LOAEL
NOAEL
LOAEL
Increased locomotor activity
16 mmol/kg 1,2,3-TMB
16 mmol/kg 1,2,4-TMB
8 mmol/kg 1,3,5-TMB
32 mmol/kg 1,2,3-TMB
32 mmol/kg 1,2,4-TMB
16 mmol/kg 1,3,5-TMB
Comments: Exposures were of an acute duration, and therefore not suitable for reference value derivation. However,
qualitatively, this study provided evidence of CNS disturbances that, when considered together with short-term and subchronic
neurotoxicity studies, demonstrate that TMB isomers perturb the CNS of exposed animals.
Source: Tomas et al. (1999b)
This document is a draft for review purposes only and does not constitute Agency policy.
B-143 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-41. Characteristics and quantitative results for Tomas et al. (1999c)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
4 rats per
dose
i.p. injection
6.6 mmol/kg BW
1,2,3-, 1,2,4-, and
1,3,5-TMB
Acute
Additional study details
• 1,2,3-, 1,2,4-, and 1,3,5-TMB were tested for their effects on the CNS by monitoring evoked hippocampal and
cortical activity following i.p. injection of 6.6 mmol/kg BW of any isomer.
•	Solvent concentration in peripheral blood was determined via head space gas chromatography.
•	Significant differences in hippocampal and cortical activity occurred following injection.
Figure 1. Amplitude
abnormalities of the
cortical N1 wave 30
and 60 min after i.p.
solvent injection.
Source: Reproduced
from Tomas et al.
(1999c)
%
15
110 H
5
0
5
10
TOLUENE
I
MESITYLENE
JL_
PSEUDOCUMENE HEMIMELLITENE
111
¦ 30 min
~ 60 min
¦
X
This document is a draft for review purposes only and does not constitute Agency policy.
B-144
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-41 (Continued): Characteristics and quantitative results for Tomas etal.
	(1999c)	
0/
/O
12
10
8
6
4
2
0
2
4
6
TOLUENE
MESITYLENE
PSEUDOCUMENE
HEMIMELLITENE
;JCLj
1
¦ 30 min
~ 60 min

Figure 2. Amplitude
abnormalities of the
cortical Pl-Nl wave
30 and 60 minutes
after i.p. solvent
injection.
Source: Reproduced
from Tomas et al.
(1999c)
o/
/o
TOLUENE
MESITYLENE
PSEUDOCUMENE
HEMIMELLITENE
30 min
~ 60 min
Figure 3. Amplitude
abnormalities of the
hippocampal Nl wave 30
and 60 min after i.p.
solvent injection.
Source: Reproduced from
Tomas et al. (1999c)
%
01
5-
10-
15-
20-
25-
30-
35-
40-
45-
50_
TOLUENE
T
~	30 min
~	60 min
MESITYLENE
PSEUDOCUMENE
HEMIMELLITENE
T
T
Figure 4. The effect of
i.p. solvent injection on
the cortical EEG in the
13-20.75 Hz frequency
band.
Source: Reproduced from
Tomas et al. (1999c)
This document is a draft for review purposes only and does not constitute Agency policy,
B-145 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-41 (Continued): Characteristics and quantitative results for Tomas etal.
	(1999c)	
60
50
40 -I
30
20
10
% 0
TOLUENE
e 30 min
~ 60 min
MESITYLENE
PSEUDOCUMENE HEMIMELLITENE
T T
Figure 5. The effect of
i.p. solvent injection on
the hippocampal EEG
in the 1-3.75 Hz
frequency band.
Source: Reproduced from
Tomas et al. (1999c)
TOLUENE
MESITYLENE
PSEUDOCUMENE HEMIMELLITENE
%
0
10 -I
20
i 30-
40"
50-
60
T
T
T
~	30 min
~	60 min
X
X
Figure 6. The effect of
i.p. solvent injection on
the hippocampal EEG in
the 7-9.75 Hz
frequency band.
Source: Reproduced from
Tomas et al. (1999c)
Health Effect at LOAEL
NOAEL
LOAEL
n/a (acute exposure study, one
dose level)
n/a
6.6 mmol/kg 1,2,3-TMB,
1,2,4-TMB, and 1,3,5-TMB
Comments: Unable to quantify dose-response relationship from data because only one dose group used. Exposures were of an
acute duration, and therefore not suitable for reference value derivation. However, qualitatively, this study provided evidence of
CNS disturbances that, when considered together with short-term and subchronic neurotoxicity studies, demonstrate that TMB
isomers perturb the CNS of exposed animals.
Source: Tomas et al. (1999c).
This document is a draft for review purposes only and does not constitute Agency policy.
B-146 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Trimethylbenzene
Table B-42. Characteristics and quantitative results for Wiaderna et al. (1998)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
13 or 14
rats/ dose
Inhalation (6 hr/day,
5 days/week)
0 or 25,100, or 250 ppm
(0, 123, 492, or 1,230
rng/m3) 1,2,3-TMB
4 weeks
Additional study details
•	Animals were exposed to 1,2,3-TMB in 1.3 rrf dynamic inhalation exposure chambers for 6 hrs/day, 5 days/week
for 4 weeks. Food and water was provided ad libitum.
•	Animals were randomized and assigned to the experimental groups.
•	Rats were tested with a variety of behavioral tests, including radial maze performance, open field activity, passive
avoidance, and active two-way avoidance.
•	Tests were performed on days 14-18 following exposure.
•	Neurobehavioral effects were observed at 25 and 100 ppm (123 and 492 mg/m ) concentrations, but not at 250
ppm (1,230 mg/m3).
300
250
200
150
100
50
o

3.0 r
2.5 -
2.0
1.5
1.0
0.5 -
0.0
HMO
day
day
day
day
day
HM25 HM100 HM250
Figure 1, Radial maze performance of rats
exposed for 4 weeks to 1,2,3-TMB,
The test (one trial a day) was performed on days
14-18 after exposure. Upper diagram: changes
in trial duration, i.e., the time of successive eight
arm entries, during successive days of training.
Lower diagram: number of perseveration errors
in successive daily trials.
Denotation of groups: HMO-sham exposed
group (n = 13), HM25, HM100, HM250-groups
exposed to 1,2,3-TMB at concentrations of 25
ppm (123 mg/m3, n = 13), 100 ppm (492 mg/m3,
n = 14), and 250 ppm (1,230 mg/'m3, n = 13)
respectively. Bars represent group means and
standard error.
* p < 0.05 compared to trial 1 in the same
group.
Source: Wiaderna et al. (1998)
HMO
HMS5
HM100 HMS50
This document is a draft for review purposes only and does not constitute Agency policy.
B-14-7
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-42 (Continued): Characteristics and quantitative results for Wiaderna et al.
	(1998)	
100
Locomotion
Exploration
GroomingT
X//A group HMO
ISSS^ group HM25
group HM100
lllll group HM250
Figure 2. A comparison of spontaneous
locomotor (upper diagram),
exploratory (middle diagram), and
grooming (lower diagram) activity of
rats in an open field during a 5-min
observation period.
The test was performed 25 days after a 4-
week exposure to 1,2,3-TMB. Denotation of
groups as in Figure 1 (above). The bars
represent group means and SE,
Source: Wiaderna et ai. (1998)
0.0
This document is a draft for review purposes only and does not constitute Agency policy,
B-148 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-42 (Continued): Characteristics and quantitative results for Wiaderna et al.
	(1998)	
u
c
-------
Toxicological Review of Trimethylbenzene
Table B-42 (Continued): Characteristics and quantitative results for Wiaderna et al.
	(1998)	
c
50
o

"C
QJ
40


'C
o
30
o
-»->
20
so

"cO
io



0

50
0

o


40
CL>

-*->


30
o

o
20
CO

Is
10
"C

E—•


0

100
a

o

• r—*
80
a

ID

a>
60


V

o
e
40
a

T3

o
20
>•

¦«c


0
Training
HMO HM25 HMIOO HM350
Retraining

HMO HM25 HMIOO HMS60
Retention
HMO	HM25 HMIOO HM250
Figure 5. Active avoidance learning and retention in rats
after a 4-week exposure to 1,2,3-TMB.
Upper and middle diagrams: comparisons of the number of
trials to attain an avoidance criterion (four avoidance responses
during five successive trials) during the training (upper diagram
and retraining (middle diagram) session). Lower diagram: a
retention score calculated according to the formula: %Ret = (1 -
Resc/Tesc) x 100, where Resc and Tesc are numbers of escape
responses during retraining and training, respectively.
Denotation of groups as in Figure 1 (above). The bars represent
group means and SE.
* p < 0.05 compared to control group.
Source: Wiaderna et al. (1998)
Health Effect at LOAEL
NOAEL
LOAEL
Impaired learning of passive
avoidance
n/a
25 ppm (123 mg/m~
Comments: CIMS disturbances were observed up to 2 months after termination of exposure, indicating the persistence of effects
after metabolic clearance of 1,2,3-TMB from the test animals. No effects were observed in the 250 ppm (1,230 mg/m3) exposure
group. Duration of exposure only 4 weeks. Generally, short-term exposure studies have limited utility in quantitation of human
health reference values.
Source: Wiaderna et al. (1998)
This document is a draft for review purposes only and does not constitute Agency policy,
B-150 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-43. Characteristics and quantitative results for Wiaderna et al. (2002)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
LOD: Wistar
rats
MM
12 rats
per dose
Inhalation (6 hr/day,
5 days/week)
0 or 25,100, or 250 ppm
(0, 123, 492, or 1,230
mg/m3) 1,2,3-TMB
4 weeks
Additional study details
•	Animals were exposed to 1,3,5-TMB in 1.3 m3 dynamic inhalation exposure chambers for 6 hrs/day, 5 days/week
for 4 weeks. Food and water was provided ad libitum.
•	Animals were randomized and assigned to the experimental groups.
•	Rats were tested with a variety of behavioral tests, including radial maze performance, open field activity, passive
avoidance, active two-way avoidance, and shock-induced changes in pain sensitivity.
•	1,3,5-TMB-exposed rats showed alterations in performance in spontaneous locomotor activity, active and passive
avoidance learning, and paw-lick latencies.
100
160
140
3120
&
1100
I 80
f 60
40
20
0
£h.
i
~ p< 0.001 compared to MESO
A

I
~	trial 1
~	trial 2
¦	trial 3
~	trial 4
~	trial 5
¦	trial 6
Figure 1. Passive avoidance. The
comparison of the time of staying
on the platform in the consecutive
test trials.
The test was performed between days
35 and 45 after the exposure to
1,3,5-TMB. Leaving the platform in trial
3 was punished by an electric shock.
Trials 1, 2, 3, and 4 were performed at
24 hr intervals, while trials 5 and 6 were
effected 3 and 7 days after trial 3,
respectively.
The bars represent group means and
SE.
Source: Wiaderna et al. (2002)
MESQ
MES25
MES100
MES250
This document is a draft for review purposes only and does not constitute Agency policy.
B-151
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-43 (Continued): Characteristics and quantitative results for Wiaderna et al.
	(2002)	
60
50
>
I
CD 30
1
120
10
* p < 0.02 compared to MESD and MES25 in ttie same trial
~L1
~L2
~L3
MESO
MES25
MES100
MES250
Figure 2. Hot plate. The
comparison of latency of the
reaction (paw-lick) to the thermal
stimulus before (LI), immediately
after (L2) and 24 hr after (L3)
intermittent 2 min electric shock in
rats exposed to 1,3,5-TMB.
The test was performed on days 50 and
51 after the exposure.
The bars represent group means and
SE.
Source: Wiaderna et al. (2002)
35
30
i 25
O
k_
£
k_
U
o
¦S 20
v>
-3
B 15
10
~ p < 0.02 compared to MESO
~ *
JL_
MESO
MES25
MES100
MES250
Figure 3. Active avoidance. The
comparison of the rat groups
exposed to 1,3,5-TMB for the
number of trials (attempts)
required to reach the avoidance
criterion (4 shock avoidances) in
5 consecutive trials (attempts)
during the training session.
The test was performed on day 54
(training) and day 60 (retraining) after
the exposure.
The bars represent group means and
SE. Source: Wiaderna et al. (2002)
Health Effect at LOAEL
NOAEL
LOAEL
Shorter retention of passive
avoidance reaction
n/a
25 ppm (123 mg/m
Comments: This study observed alterations in a number of behavioral tests. Values reported by authors can be used to determine
LOAEL and NOAEL. CNS disturbances observed up to 2 months after termination of exposure, indicating the persistence of effects
following metabolic clearance of 1,3,5-TMB from the test animals. Unable to quantify dose-response relationship from data because
responses either equal at all exposure concentrations or elevated only at one exposure concentration. Duration of exposure only 4
weeks. Generally, short-term exposure studies have limited utility in quantitation of human health reference values.
Source: Wiaderna et al. (2002).
This document is a draft for review purposes only and does not constitute Agency policy.
B-152 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-44. Characteristics and quantitative results for Wiglusz et al. (1975b)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
5-8 per
dose
Inhalation
0,1.5, 3.0, or 6.0 mg/L
(0, 1,500, 3,000, or 6,000
mg/m3) 1,3,5-TMB
Acute study: 6 hrs
Short-term study: 6 hrs/day,
6 days/week for 5 weeks
Additional study details







• Male Wistar rats were exposed in a short-term study to 0,1.5, 3.0, or 6.0 mg/L 1,3,5-TMB.

• In a separate chronic study, male Wistar rats were exposed to 3.0 mg/L 1,3,5-TMB for 6 hrs/day, 6 days/week, for
5 weeks.
• Rats weighed 240-280 g and were housed in stainless steel wire mesh cages, with food and water provided
ad libitum.
• Blood samples were collected for 3 days before exposure then on days 1, 7,14, and 28.




1,3,5-TMB exposure concentration (mg/L)—
hematological parameters following single 6 hour exposure
Observation


0
1.5
3.0
6.0
Hemoglobin in g% (mean ± SD)
Day 0
14.1 ± 1.3
15.2 ±0.3
15.0 ±0.8
14.2 ± 1.1
Day 1
-
-
14.8 ± 1.0
13.9 ±2.1
Day 7
-
14.0 ± 0.5
13.5 ±0.5
13.5 ±0.8
Day 14
15.1 ±0.8
14.6 ± 0.5
13.6 ±0.6
13.1 ±0.4
Day 28
14.8 ±0.5
14.9 ± 0.7
13.6 ±0.8
14.8 ±0.4
Million erythrocytes per mm3 serum (mean ± SD)
Day 0
4.91 ±0.19
5.35 ±0.09
4.96 ±0.15
5.51 ±0.17
Day 1
-
-
5.32 ±0.02
5.31 ±0.11
Day 7
-
5.18 ±0.18
4.93 ±0.16
4.89 ±0.17
Day 14
5.37 ±0.90
4.99 ±0.11
5.09 ±0.10
4.77 ±0.10
Day 28
5.17 ±0.18
5.26 ±0.07
5.12 ±0.10
5.20 ±0.27
Thousand leukocytes per mm3 serum (mean ±SD)
Day 0
11.08 ±3.14
12.26 ±3.50
13.01 ±3.10
8.90 ± 3.88
Day 1
-
-
11.38 ± 1.37
8.24 ± 3.88
Day 7
-
11.70 ±2.97
11.66 ± 1.50
12.32 ±5.01
Day 14
8.0 ±2.16
12.06 ± 3.33
11.70 ± 1.05
10.68 ± 1.21
Day 28
6.83 ± 1.27
11.50 ± 10.48
11.96 ± 1.16
9.92 ±2.42
This document is a draft for review purposes only and does not constitute Agency policy.
B-153
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-44 (Continued): Characteristics and quantitative results for Wiglusz etal.
	("1975b")	
Percent segmented neutrophilic granulocytes (mean ± SD)
Day 0
8.5 ±4.1
13.5 ±3.6
18.5 ±2.3
16.6 ±2.8
Day 1
--
-
22.5 ±5.4
53.6 ±22.5
Day 7
--
20.2 ± 6.04
31.3 ± 10.3
26.7 ± 12.5
Day 14
10.6 ±2.5
12.2 ±5.9
30.1 ±6.2
20.6 ±23.7
Day 28
15.6 ±6.3
12.5 ±6.4
35.0 ±6.7
15.8 ±3.8
Percent bacciliform neutrophilic granulocytes (range)
Day 0
0.6 (0-1)
0.0
0.0
0.0
Day 1
--
--
0.0
0.0
Day 7
--
0.0
0.0
0.0
Day 14
0.0
0.16 (0-1)
0.0
0.0
Day 28
0.0
1 (0-2)
0.0
0.0
Percent acidophilic granulocytes (mean ± SD)
Day 0
1.1 ±0.7
2.6 ± 1.9
0.5 ±0.5
1.8 ± 1.7
Day 1
-
-
0.0
0.14 ±0.3
Day 7
-
1.1 ± 1.1
3.1 ±0.5
0.0
Day 14
2.8 ± 1.3
5.1 ±3.2
4.8 ± 1.0
2.6 ±2.6
Day 28
4.1 ±2.9
3.1 ± 1.7
6.0 ±4.1
2.2 ±2.8
Percent lymphocyte (mean ± SD)
Day 0
88.6 ±4.4
82.8 ±4.8
67.8 ±2.3
79.4 ±4.3
Day 1
--
-
73.3 ±5.4
44.0 ± 21.3
Day 7
--
77.6 ±4.8
65.0 ±7.9
71.2 ± 12.5
Day 14
85.4 ± 1.5
82.0 ± 3.8
64.3 ±5.8
75.0 ±23.0
Day 28
78.6 ±8.3
81.8 ± 7.6
57.1 ±4.1
81.2 ±5.8
Percent monocyte (mean ± SD)
Day 0
1.6 ±0.8
1.0 ±0.6
1.1 ±0.9
2.2 ± 1.0
Day 1
--
-
1.1 ±0.4
2.3 ± 1.8
Day 7
--
0.8 ± 1.1
0.3 ±0.5
1.7 ± 1.9
Day 14
0.8 ±0.4
0.6 ±0.5
0.3 ±0.8
1.2 ±0.4
Day 28
1.6 ± 1.0
1.6 ± 1.0
1.6 ± 1.2
1.0 ±0.8
This document is a draft for review purposes only and does not constitute Agency policy.
B-154 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-44 (Continued): Characteristics and quantitative results for Wiglusz etal.
	("1975b)	
Qj
¦*>»
"3-
Qi
CLl
c:
SO
30
iO
O
• control
mesltylene J.Omgli
mesitylene. 3.0 mg/l
mesitylene. 6.0 mg/i
Figure 1. Percentage
of segmented
neutrophilic
granulocytes after
6 hrs exposure to
1,3,5-TMB.
Source: Wiglusz et al.
(1975b)
0 1
~r
7
~t~
14
days after exposure
i—
26
-Ij
o
c?
a
o,
3
• >4
10
\ 10
o
control
mesitylene 3.0mg/l
£
fimn iiiiii mum	lsjli.i i
Figure 2.
Percentage of
segmented
neutrophilic
granulocytes
during
exposure to
1,3,5-TMB
3,0 mg/L for
6 hrs/day,
6 days/week,
for 5 weeks.
Source: Wiglusz
et al. (1975b)
O i
i	i	r
i*	days of exposure	I
Observation
Hematological parameters during 5 week exposure to 1,3,5-TMB (means ± SD)
Day 0
Day 1
Day 7
Day 14
Day 28
Hemoglobin in g%
Control group
13.0 ±4.7
14.6 ±2.5
14.6 ±2.5
15.6 ±3.2
14.2 ± 5.0
1,3,5-TMB group
14.6 ±0.7
15.5 ±0.6
14.8 ± 1.1
14.5 ±0.9
13.8 ±0.5
Million erythrocytes per mm Serum
Control group
5.42 ±0.78
6.12 ±04
6.40 ±0.25
6.46 ±0.39
6.18 ±0.61
1,3,5-TMB group
6.08 ± 1.18
6,35±0.38
6.11 ±0.63
5.74 ± 1.1
5.05 ±2.2
Thousand leukocytes per mm Serum
Control group
10.63 ±4.27
13.66 ±2.91
11.13 ±2.52
14.53 ± 2.64
11.46 ±2.74
1,3,5-TMB group
13.76 ±3.70
11.43 ±4.0
9.53 ±2.55
12.23 ±4.04
13.40 ±5.18
% Segmented neutrophilic Granulocytes
Control group
17.1 ± 11,9
14,5 ±8.1
12.1 ±2.5
13.6 ±6.3
15.6 ±3.2
1,3,5-TMB group
14.0 ± 5.0
17.0 ± 9.4
16.6 ±5.0
21.5 ±7.4
18.4 ±8.6
This document is a draft for review purposes only and does not constitute Agency policy,
B-155 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-44 (Continued): Characteristics and quantitative results for Wiglusz etal.
	("1975b")	
% Bacciliform neutrophilic granulocytes
Control group
0.83 (1-2)
0.66 (1-2)
1.33 (1-3)
1.33 (1-2)
1.0 (0-1)
1,3,5-TMB group
0.6 (1-2)
0.4 (0-1)
1(1-2)
1.8 (2-5)
1.4 (1-2)
% Acidophilic granulocytes
Control group
1 (1-4)
2.1 (1-4)
3.3 (1-7)
1.8 (1-4)
1.6 (1-4)
1,3,5-TMB group
1.5 (1-3)
1.0 (1-3)
0.8 (1-2)
1.0 (1-2)
0.8 (0-1)
% Lymphocyte
Control group
79.6 ± 11.7
81.6 ±8.6
81.8 ±4.7
81.1 ±5.2
80.0 ± 2.4
1,3,5-TMB group
79.8 ±5.5
81.0 ±7.7
80.5 ±6.5
74.0 ±9.4
77.2 ±8.4
% Monocyte
Control group
1.1 (1-3)
1.0 (0-2)
1.5 (1-4)
1.0 (1-2)
1.5 (1-3)
1,3,5-TMB group
0.6 (1-3

0.8 (1-2)
0.8 (1-2)
1.3 (1-3)
2.7 (2-4)
Health Effect at LOAEL
NOAEL
LOAEL
Increase in percent segmented
neutrophilic granulocytes
1.5 mg/L
3.0 mg/L
Comments: This study slight increases in percent segmented neutrophilic granulocytes on day 14 of the short-term exposure study.
Authors do not report statistical significance of results. Only one dose group used in chronic study.
Source: Wiglusz et al. (1975b)
This document is a draft for review purposes only and does not constitute Agency policy.
B-156 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-45. Characteristics and quantitative results for Wiglusz et al. (1975a)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
6/dose
Inhalation
0, 0.3,1.5, or 3.0 mg/L
(0, 300, 1,500, or 3,000
mg/m3) 1,3,5-TMB
Acute study: 6 hrs
Short-term study: 6 hrs/day,
6 days/week for 5 weeks
Additional study details







• Male Wistar rats were exposed in a short-term study to 0, 0.3,1.5, or 3.0 mg/L 1,3,5-TMB.

• In a separate chronic study, male Wistar rats were exposed to 3.0 mg/L 1,3,5-TMB for 6 hrs/day, 6 days/week,
for 5 weeks.
• Rats weighed 240-280 g and were housed in stainless steel wire mesh cages, with food and water provided ad
libitum.
• Blood samples were collected for 3 days before exposure then on days 1, 7,14, and 28.




1,3,5-TMB exposure concentration (mg/L)—hematological parameters
following single 6 hour exposure (means ± SE)
Observation
0
0.3
1.5
3.0
Aspartate amino transferase activity
Day 0
79.0 ±7.9
78.0 ±7.7
75.3±7.3
81.6 ±4.2
Day 2
81.8 ±6.2
90.0 ±5.7
71.8±3.3
74.6 ±4.5
Day 7
82.2 ±4.3
76.8 ±4.2
71.2±2.2
84.1 ±5.6
Day 14
82.6 ±8.5
73.0 ±4.2
76.3±6.7
76.1 ±3.9
Day 28
79.6 ±7.6
72.6 ±7.2
84.2±7.9
79.5 ± 10.6
Alanine amino transferase activity
Day 0
34.0 ±4.5
35.6 ±4.1
32.6 ±4.5
29.1 ±3.6
Day 2
34.0 ±4.6
30.8 ±2.7
30.6 ±8.3
26.5 ± 1.2
Day 7
31.0 ±3.1
37.5 ±5.6
29.3 ±4.5
39.5 ±3.0
Day 14
32.0 ±3.2
31.4 ±2.5
34.6 ±5.3
36.3 ± 1.7
Day 28
34.0 ±3.8
31.3 ±5.2
30.4 ± 9.4
39.3 ±2.7
Alkaline phosphatase activity
Day 0
28.6 ±9.6
30.9 ±3.3
27.4 ±6.4
37.3 ±5.6
Day 2
27.8 ±5.1
26.0 ±7.2
29.7 ±2.6
30.5 ±6.5
Day 7
31.8 ±5.8
28.1 ±5.9
32.8 ± 1.8
58.7 ±8.9*
Day 14
27.0 ±4.7
33.6 ±2.4
28.9 ±5.2
42.1 ±2.9
Day 28
30.5 ±3.2
28.0 ±6.9
23.0 ±4.7
-
This document is a draft for review purposes only and does not constitute Agency policy.
B-157
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-45 (Continued): Characteristics and quantitative results for Wiglusz etal.
	("1975a")	
i20
<0
<3

too
SO
control
mesitylene 0.3 Mf/t
mesitylene 1.5 mgfl
mesitylene 3.0 mgjl
> a 2
~r
7
U days after exposure
1
Figure 1.
Serum
activity of
aspartate
amino
transferase
after 6 hrs
exposure to
1,3,5-TMB;
values are
expressed in
% of initial
values.
Source:
Wiglusz et a!.
(1975a)
2 w
—-i
-------
Toxicological Review ofTrimethylbenzene
Table B-45 (Continued): Characteristics and quantitative results for Wiglusz etal.
	("1975a")	
*0
Q)
1
«
¦v.
.5
i4o
ISO
too
¦80
o 1
control
	mesitglene 0.3 mg/t
		mesilylene 1.0 mq/L
mesitylene 3.0 mg/L
Figure 3.
Serum activity
of alkaline
phosphatase
after 6 hrs
exposure to
1,3,5-TMB;
values are
expressed in %
of initial
values.
Source: Wiglusz
et al. (1975a)
r~
2
—j—
7
ii dags after exposure
1
2B~
Observation
Hematological parameters during 5 week exposure to 1,3,5-TMB (means ± SD)
Day 0
Day 1
Day 3
Day 7
Day 14
Day 28
Aspartate amino transferase activity
Control group
89.5 ±2.3
74.5 ±6.9
79.6 ± 10.5
83.2 ± 10.6
83.5 ±7.3
82.2 ±6.3
1,3,5-TMB group
72.0 ±5.1
70.8 ±5.2
81.3 ±9.1
80.0 ± 6.3
93.4 ± 1.4*
79.6 ±9.4
Alanine amino transferase activity
Control group
34.0 ±4.1
33.8 ±5.0
35.6 ±2.6
30.5 ±4.9
30.0 ±4.5
35.6 ±4.6
1,3,5-TMB group
34.8 ±3.6
28.0 ±6.32
3.33 ±3.8
35.1 ±3.9
36.4 ±4.0
36.5 ±5.0
Ornithite carbamyl transferase activity
Control group
2.7 ±0.2
2.6 ±0.2
3.1 ±0.2
2.8 ±0.1
2.6 ±0.3
3.6 ±0.3
1,3,5-TMB group
2.6 ±0.4
2.5 ±0.6
3.8 ±0.4
3.5 ±0.2
2.6 ±0.2
3.7 ±0.4
Alkaline phosphatase activity
Control group
27.8 ±4.0
28.8 ±3.8
28.5 ±6.8
26.5 ±3.9
27.2 ±8.8
25.8 ±3.0
1,3,5-TMB group
32.4 ± 1.8
23.6 ±3.6
22.2 ±3.6
30.2 ±6.9
25.6 ±5.9
32.6 ±4.8
This document is a draft for review purposes only and does not constitute Agency policy.
B-159 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-45 (Continued): Characteristics and quantitative results for Wiglusz etal.
	("1975a")	
^ tZO -
control
m€» sit y /£* ne, 3. O mg/Z
cfcu/r of exposure
Figure 4, Serum
activity of
aspartate
amino
transferase
during
exposure to
1,3,5-TMB at
3,0 mg/Lfor6
hrs/day, 6
days/week, for
5 weeks; values
are expressed
in % of initial
values.
Source: Wiglusz t al. (1975a)
^ 120
"a
¦5 too

	 control	-
	 mesitylene 3.0 mg/l
ijiipii ¦ ¦¦¦!¦ ¦¦¦¦¦¦¦ mm
Figure 5. Serum
activity of
alanine amino
transferase
during exposure
to 1,3,5-TIVIB at
3.0 mg/L for 6
hrs/day, 6 days
per week, for 5
weeks; values
are expressed in
% of initial
values.
c i
r
14
cfat/s of exposure
26
Source: Wiglusz et al. (1975a)
This document is a draft for review purposes only and does not constitute Agency policy,
B-160 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-45 (Continued): Characteristics and quantitative results for Wiglusz etal.
	("1975a")	
control
mestti/lene S-Omqll
f20

.<3
days of exposure
Figure 6. Serum
activity of alkaline
phosphatase
during exposure to
1,3,5-TMB at 3.0
mg/L for 6 hrs/day,
6 days/week, for 5
weeks; values are
expressed in % of
initial values.
Source: Wiglusz et al.
(1975a)
Health Effect at LOAEL
NOAEL
LOAEL
Increase in alkaline phosphatase
activity
1.5 mg/L
3.0 mg/L
Comments: This study observed increases in alkaline phosphatase activity on day 7 of the short-term exposure study. Only one
dose group used in chronic study. Data not recorded daily; significant gaps exist between sampling days.	
*Statistically significant in relation to initial values (p < 0.05).
Source: Wiglusz et al. (1975a)
This document is a draft for review purposes only and does not constitute Agency policy.
B-161 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
B.6. HUMAN TOXICOKINETIC STUDIES
Table B-46. Characteristics and quantitative results for Jarnberg et al. (1996)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Caucasian
humans
M
9 per
dose
Inhalation
2 ppm and 25 ppm
(~10 and 123 mg/m3)
1,2,3-, 1,2,4-, or 1,3,5-TMB
2 hrs exposure, followed by
4 hrs observation
Additional study details
•	Caucasian males were exposed to 2 ppm (~ 10 mg/m3) 1,2,4-TMB and 25 ppm (123 mg/m3) 1,2,3-, 1,2,4-, or
1,3,5-TMB in an inhalation chamber for 2 hrs.
•	Study subjects were asked to perform light cycling to simulate a work environment, with participants generating
50 W power during 2 hr exposure.
•	1,2,3-, 1,2,4-, and 1,3,5-TMB concentrations in exhaled air, blood, and urine were determined via gas
chromatography.
•	No significant irritation or CNS effects were observed.
•	Results imply extensive deposition in adipose tissue.
•	Exhalation accounted for 20-37% of absorbed amount while urinary excretion of unchanged TMBs accounted for
<0.002%.
•	The study was approved by the Regional Ethical Committee at the Karolinska Institute
Respiratory uptake and urinary excretion of TMB isomers following 2 hour inhalation exposure (mean ± 95%CI)
Exposure
25 ppm (123
mg/m3)l,2,3-TM
B
25 ppm (123
mg/m3)l,3,5-TM
B
25 ppm (123
mg/m3)
1,2,4-TMB
2 ppm (~10
mg/m3)
1,2,4-TMB
Respiratory uptake (%)a
56 ±4
62 ±3
64 ±3
63 ±2
Net respiratory uptake (%)b
48 ±3
55 ±2
60 ±3
61 ±2
Respiratory uptake (mmol)a
1.4 ±0.1
1.6 ±0.1
1.6 ±0.1
0.16 ±0.01
Net respiratory uptake (mmol)b
1.2 ±0.1
1.4 ±0.1
1.5 ±0.1
0.15 ±0.01
Respiratory excretion (%)c
37 ±9
25 ±6
20 ±3
15 ±5
Net respiratory excretion (%)d
28 ±8
16 ±4
14 ±2
9 ±4
Urinary excretion (%)e
0.0023 ± 0.0008
0.0016 ± 0.0015
0.0010 ± 0.0004
0.0005 ± 0.0002
This document is a draft for review purposes only and does not constitute Agency policy.
B-162
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-46 (Continued): Characteristics and quantitative results for Jarnberg et al.
	(1996)	
Kinetic parameter
25 ppm (123
mg/m3)l,2,3-TM
B
25 ppm (123
mg/m3)l,3,5-TM
B
25 ppm (123
mg/m3)
1,2,4-TMB
2 ppm (~10
mg/m3)
1,2,4-TMB
Total calculated blood clearance
(L/hr/kg)f
0.63 ±0.13
0.97 ±0.16
0.68 ±0.13
0.87 ±0.37
Total apparent calculated blood
clearance (L/hr/kg)8
0.54 ±0.11
0.86 ±0.12
0.63 ±0.11
0.82 ±0.32
Exhalatory blood clearance (L/hr/kg)f
0.23 ±0.07
0.24 ±0.10
0.14 ±0.04
0.14 ±0.10
Metabolic blood clearance (L/hr/kg)f
0.39 ±0.11
0.72±0.11
0.54 ±0.10
0.74 ±0.29
1st Phase half-life (min)
1.5 ±0.9
1.7 ±0.8
1.3 ±0.8
1.4 ± 1.8
2nd Phase half-life (min)
24 ±9
27 ±5
21 ±5
28 ± 14
3rd Phase half-life (min)
4.7 ± 1.6
4.9 ± 1.4
3.6 ± 1.1
5.9 ±2.5
4th Phase half-ife (min)
78 ±22
120 ±41
87 ±27
65 ±20
AUC (nM x hrs)
32 ±6
22 ±4
35 ± 10
3.6 ±2.0
Volume of distribution (L/kg)
30 ±6
39 ±8
38 ± 11
28 ±3
Mean residence time (hrs)
57 ±22
42 ± 11
69 ±32
47 ±22
Kinetic values of TMB isomers following 2 hour inhalation exposure (mean ± 95%CI)
Figure 1. Concentration of 1,2,4-TMB in capillary blood during and after 2 hr exposure to 25 ppm (123 mg/mB
1,2,4-TMB (mean values ± 95% CI).
124TMB in blood
(MM)
60	120	180	240
Time (min)
300
360
Source: Jarnberg et al. (1996)
This document is a draft for review purposes only and does not constitute Agency policy.
B-163 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-46 (Continued): Characteristics and quantitative results for Jarnberg et al.
	(1996)	
Figure 2. Concentration of 1,3,5-TMB in capillary blood during and after 2 hr exposure to 25 ppm (123 mg/mB)
1,3,5-TMB (mean values ± 95% CI).
c
135TMB in blood
S
V
6
5

3
o
1
O
6 O
1 SO
300
O
Time (miri)
Source: Jarnberg et al. (1996)
Figure 3. Concentration of 1,2,3-TMB in capillary blood during and after 2 hr exposure to 25 ppm (123 mg/mB)
1,2,3-TMB (mean values ± 95% CI).
123TMB in blood
(^M)
9 n
7 -
5 -
360
300
1 80
240
1 20
0
60
Time (miri)
Source: Jarnberg et al. (1996)
This document is a draft for review purposes only and does not constitute Agency policy.
B-164 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-46 (Continued): Characteristics and quantitative results for Jarnberg et al.
	(1996)	
Figure 4. Concentration of 1,2,4-TMB in capillary blood from 10 subjects exposed to 2 and 25 ppm (~10 and 123
mg/mB) of 1,2,4-TMB (mean values ± 95% CI)
124TMB in blood
CjM)
25 ppm
2 ppm
0.01
240
360
0
60
120
1 80
300
Time (min)
Source: Jarnberg et al. (1996)
Comments: Exposure duration possibly not sufficient to detect metabolic changes. Metabolites not measured.
aPercent of dose calculated as net uptake + amount cleared by exhalation during exposure .
Percentage of dose calculated as net uptake.
cDuring and post-exposure, percentage of the respiratory uptake.
dPost-exposure, percentage of net respiratory uptake.
0Post-exposure, percentage of respiratory uptake.
'Calculated from respiratory uptake.
Calculated from net respiratory uptake.
Source: Jarnberg et al. (1996)
This document is a draft for review purposes only and does not constitute Agency policy.
B-165 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-47. Characteristics and quantitative results for Jarnberg et al. (1997a)
Study design
Species
Sex
Exposure route
Dose range
Exposure duration
Caucasian
Human
M
Inhalation
11 mg/m 1,2,4-TMB
2 hrs
Additional study details
•	Nine Caucasian males were exposed to 11 mg/m31,2,4-TMB alone or 11 mg/m31,2,4-TMB as a component of 300
mg/m3 WS.
•	Exposure lasted 2 hrs, during which study subjects were required to cycle producing 50 W continuously to simulate
a work environment.
•	Gas chromatography was used to measure 1,2,4-TMB levels in air.
•	HPLC was used to measure urinary metabolites.
•	Irritation was not reported amongst subjects at these exposure levels.
•	The study was approved by the Regional Ethical Committee at the Karolinska Institute and was only performed
after informed consent.
Figure 1. Mean (± SD) capillary blood concentration of 1,2,4-TMB during and after exposure to 1,2,4-TMB alone and
1,2,4-TMB as a component of WS.
Source: Jarnberg et al. (1997a)
1,2,4-TMB in blood (|iM)
1.01
0,8-
0,6-
~	 Exposure to white spirit
Exposure to 1,2,4-TMB
exposure
60	120
Time (min)
180
This document is a draft for review purposes only and does not constitute Agency policy.
B-166
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-47 (Continued): Characteristics and quantitative results for Jarnberg et al.
	("1997a")	
Results from 2 hour exposure to 1,2,4-TMB alone or 1,2,4-TMB as a component of WS (mean ± SD)
Exposure
1,2,4-TMB alone
1,2,4-TMB in WS
p-value
Net respiratory uptake (mmol)
0.15 ±0.01
0.14 ±0.02
0.5a
AUC (nM x min), 0-3 hr
53 ±4
86 ±9
<0.0001a
Half-life of 3,4-DMHA (hr)
3.7 ±0.4b
3.0 ±0.7
0.2°
Excretion of 3,4-DMHA (%d), 0-6 hr
11 ±2
18 ±3
0.007°
Figure 2. Urinary excretion rate of 3,4-dimethylhippuric acid against the midpoint time of urine collection in 9 male
volunteers exposed to 11 mg/mBof 1,2,4-TMB, either alone or as a component of WS (mean ± 95% CI).
Urinary
of 3,4-
6 "|
5"
4"
3"
2-
1 -
0
Source: Jarnberg et al. (1997a)
Comments: Metabolites (DMBAs) measured in urine. Exposure duration possibly not sufficient to detect other metabolic changes.
Only one exposure group; multiple concentrations not tested.
excretion rate
DMHA (|imol/h)
Exposure to white spirit
Exposure to 1,2,4-TMB
exposure
T
T
T
4	8	12	16	20
Time (h from onset of exposure)
a Student's t-test
b Recalculated for 9 subjects form a 120 mg/mB exposure to 1,2,4-TMB
c Analysis of variance
d 5 of net respiratory uptake
Source: Jarnberg et al. (1997a)
This document is a draft for review purposes only and does not constitute Agency policy.
B-167 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-48. Characteristics and quantitative results for Jarnberg et al. (1997b)
Study design
Species
Sex
Exposure route
Dose range
Exposure duration
Caucasian
Humans
M
10
Inhalation
25 ppm (123 mg/m )
1,2,3-TMB, 1,2,4-TMB, or
1,3,5-TMB
2 hrs
Additional study details
•	Ten males were exposed to 25 ppm (123 mg/m3) 1,2,3-TMB, 1,2,4-TMB or 1,3,5-TMB for 2 hrs or 2 ppm (~10
mg/m3) 1,2,4-TMB for 2 hrs.
•	Study subjects were asked to perform light cycling to simulate a work environment, with participants generating
50 W power during 2 hr exposure.
•	Isomers of all DMHA metabolites in urine were detected via HPLC.
•	Approximately 22% of inhaled 1,2,4-TMB, 11% of inhaled 1,2,3-TMB, and 3% of inhaled 1,3,5-TMB was found to be
excreted as DMHAs in urine within 24 hrs following exposure.
•	The study was approved by the Regional Ethical Committee at the Karolinska Institute and only with the informed
consent of the subjects and according to the 1964 Declaration of Helsinki
Half-times of urinary excretion rate, recoveries, and rates of urinary DMHA isomer excretion (mean ± 95% CI)
Exposure
Isomer
Half-time (hr)
Urinary recovery %
(24 hrs)
Excretion rate,
Hg/min, 0-24 hrs
1,2,3-TMB
2,3-DMHA
4.8 ±0.8
9 ± 3
19 ±3
1,2,3-TMB
2,6-DMHA
8.1 ± 1.5
2 ± 2
4.2 ± 1.7
1,2,4-TMB
3,4-DMHA
3.80 ±0.4
18 ± 3
44 ± 6
1,2,4-TMB
2,4-DMHA
5.8 ±0.9
3 ±0.8
8.2 ± 1.4
1,2,4-TMB
2,5-DMHA
5.3 ± 1.5
<1 ± 0.2
1.6 ±0.5
1,3,5-TMB
3,5-DMHA
16 ±6
3 ± 2
8.9 ±2.1
Comments: Metabolites (DMBAs) measured in urine. Exposure duration possibly not sufficient to detect metabolic changes
associated with longer time points. Toxicokinetics studied at only one concentration.
Source: Jarnberg et al. (1997b)
This document is a draft for review purposes only and does not constitute Agency policy.
B-168 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-49. Characteristics and quantitative results for Jarnberg et al. (1998)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Caucasian
humans
M
9
subjects
Inhalation
2 ppm (~10 mg/m3)
1,2,4-TMB,
2 ppm (~ 10 mg/m3) in
WS,
25 ppm (123 mg/m3)
1,2,4-TMB
2 hrs exposure, followed by 6
hrs observation
Additional study details






• Caucasian males were exposed to 2 ppm (~10 mg/m3) 1,2,4-TMB, 2 ppm (~ 10 mg/m3) in WS, 25 ppm (123
mg/m3) 1,2,4-TMB in an inhalation chamber for 2 hrs.
• Study subjects were asked to perform light cycling to simulate a work environment.

• 1,2,4-TMB concentration was determined via gas chromatography.



• DMHA metabolites were measured with HPLC.




• Blood levels of 1,2,4 TMB and its urinary metabolites were found to be higher in the WS exposure group
suggesting that components of WS could interfere with TMB metabolism.
• No significant irritation or CNS effects were observed.



• The study was approved by the Regional Ethics Committee of the Karolinska Institute and was only performed
after informed consent.
Kinetic results following 2 hour inhalation exposure to 1,2,4-TMB and 1,2,4-TMB in WS—mean values (95% CI)
Kinetic parameter
2 ppm (~ 10 mg/m3)
group
2 ppm
(~10 mg/m3)
n WS
25 ppm (123
mg/m3) alone
Actual [TMB] (ppm)
2.22 (2.13-2.31)
2.26 (2.20-2.32)
23.9 (22.7-25.1)
Respiratory uptake (mmol)a
0.16 (0.14-0.18)
0.16 (0.14-0.18)
1.73 (1.61—1.85)
Net respiratory uptake
0.15 (0.14-0.16)
0.14 (0.12-0.16)
1.52 (1.37-1.67)
AUCb|00d (nM x min)
95 (54-137)
157 (136-178)*
1286 (1131-1441)
Total blood clearance (L/min)
2.09 (1.52-2.66)
1.06 (0.89-1.23)**
1.38 (1.23-1.53)*
Metabolic blood clearance (L/min)
1.71 (1.15-2.26)
0.79 (0.62-0.96)*
1.06 (0.87-1.25)*
Exhalatory blood clearance (L/min)
0.39 (0.28-0.50)
0.28 (0.20-0.36)
0.32 (0.24-0.40)
Mean residence time (hr)
4.6 (-1.3-10.5)
4.8 (2.1-7.5)
3.8(1.8-5.8)
Volume of distribution, steady state (L)
293 (69-517)
271 (139-403)
294 (165-423)
This document is a draft for review purposes only and does not constitute Agency policy.
B-169
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-49 (Continued): Characteristics and quantitative results for Jarnberg et al.
	(1998) 			
Half-life in blood, TMB, 1st phase (min)
3.9 (1.4-6.4)
5.9 (3.1-8.7)
6.1 (5.3-6.9)
Idem, TMB, 2nd phase (hr)
4.3 (-0.5-9.0)
4.8 (2.1-7.5)
4.0(2.2-5.8)
Half-life in urine, 3,4-DMHA (hr)
NDC
3.0 (2.3-3.7)
3.8 (3.4-4.2)
Urinary recovery, 3,4-DMHA (%)b, 0-6 hr
11 (9-13)
18( 15-21) *
14 (12-16)
Idem (%)b, 0-22 hR
ND
27 (23-31)
18 (15-21)
Comments: Multiple exposure concentrations were tested and multiple tissues were analyzed. Study of 1,2,4-TMB as a
component of WS. Toxicokinetics of 1,2,3- and 1,3,5-TMB not studied.
aNet respiratory uptake + amount cleared by exhalation during exposure.
b% of net respiratory uptake.
cNot determined.
*p < 0.05, **p < 0.01, compared to 2 ppm (~10 mg/mB) alone by repeated measures ANOVA.
Source: Jarnberg et al. (1998).
This document is a draft for review purposes only and does not constitute Agency policy.
B-170 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-50. Characteristics and quantitative results for Jones et al. (2006)
Study design
Species
Sex
Exposure route
Dose range
Exposure duration
Human
M/F
2 per sex
Inhalation
25 ppm (1,2,3-TMB
mg/m3) 1,3,5-TMB
4 hrs
Additional study details
•	Two males and two females were exposed to 25 ppm (1,2,3-TMB mg/m3) 1,3,5-TMB in an inhalation chamber for 4
hrs.
•	1,3,5-TMB concentration in exhaled air, venous blood, and urine was determined via gas chromatography.
•	No significant irritation or CNS effects were observed during the inhalation study, although one volunteer was
treated with a 2 cm2gauze patch soaked with liquid 1,3,5-TMB and reported mild itching, erythema, and oedema
where gauze contacted skin.
•	Authors conclude that urinary DMBA and breath TMB are suitable markers of TMB exposure, and that repeated
exposures during work week can result in significant accumulation in tissues.
•	The study was approved by the Health and Safety Executive's Research Ethics Committee
Figure 1. Mean ± SD urinary total DMBAs. Black and grey arrows represent 24 and 48 hrs respectively, following a
single 4 hr exposure to 25 ppm (1,2,3-TMB mg/mB) 1,3,5-TMB.
50	100
Time (hours)
¦Exposure Time
E 30 0
< 20 0
150
Source: Jones et al. (2006)
This document is a draft for review purposes only and does not constitute Agency policy.
B-171
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-50 (Continued): Characteristics and quantitative results for Jones et al.
	(2006)	
Figure 2. Mean ± SD blood levels of 1,3,5-TMB during and after 4 hr exposure to 25 ppm (1,2,3-TMB mg/mB)
1,3,5-TMB.
Exposure Time
o 0.40
Time (hours)
Source: Jones et al. (2006)
Figure 3. Mean ± SD breath levels of 1,3,5-TMB during and after 4 hr exposure to 1,3,5-TMB.
Exposure Time
5 120
£ AO
Time (hours)
Source: Jones et al. (2006)
Comments: Metabolite (DMBA) concentration measured in urine. Subjects tested included males and females. Small number of
study subjects (n = 4). Exposure duration possibly not sufficient to detect metabolic changes. Other metabolites not measured.
Source: Jones et al. (2006)
This document is a draft for review purposes only and does not constitute Agency policy.
B-172 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-51. Characteristics and quantitative results for Kostrzewski et al. (1997)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Human
M/F
5
Inhalation
Between 5 and 150
mg/m31,2,4-TMB,
1,3,5-TMB, and 1,2,3-TMB
4 or 8 hrs
Additional study details
•	Five humans were exposed to 1,2,4-TMB, 1,3,5-TMB, and 1,2,3-TMB at concentrations between 5 and 150
mg/m3.
•	Exposure durations were either 4 or 8 hrs.
•	TMBs were measured in blood and urine, via gas chromatography.
•	DMBA excretion was found to follow an open, two-compartment model.
1,2,3-, 1,2,4-, and 1,3,5-TMB concentration in blood before, during, and after exposure
Sampling time
(hrs)
1,2,3-TMB
1,2,4-TMB
1,3,5-TMB
Blood
concen-
tration
(Hg/dm3
[Hg/L])
SD
Blood
concen-
tration
(Hg/dm3
[Hg/L])
SD
Blood
concen-
tration
(Hg/dm3
[Hg/L])
SD
0
0
0
0
0.00
0
0.00
0.25
259
94.5
194
19.80
181
25.01
0.50
290
91.54
460
57.36
308
5.29
1
295
57.11
533
46.61
355
44.80
2
380
93.17
730
128.89
482
201.57
4
341
186.94
810
112.40
603
184.13
8
520
129.42
979
171.12
751
122.87
0.05
261
50.36
580
36.2
434
36.40
0.10
277
57.89
496
85.03
388
64.16
0.15
287
38.18
447
106.69
309
38.78
0.25
277
35.47
387
65.83
298
65.48
0.50
-
-
246
128.54
247
34.00
1
204
17.78
131
19.87
190
41.13
2
133
38.55
101
14.17
121
24.60
4
85
8.96
85
13.65
94
16.52
6
65
23.69
63
11.03
76
25.81
8
64
11.59
69
7.09
74
20.16
25
54
14.57
54
3.74
45
13.93
32
29
3.51
48
10.24
44
20.19
49
19
13.01
46
9.98
42
7.93
56
21
11.31
31
9.32
42
9.81
73
14
3.50
26
9.49
-
-
This document is a draft for review purposes only and does not constitute Agency policy.
B-173
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-51 (Continued): Characteristics and quantitative results for Kostrzewski
etal. (1997)
Excretion rate (V[mg/hr]) of dimethylbenzoic acid (DMBA) in urine during and after exposure to 1,2,3-TMB,
1,2,4-TMB, or 1,3,5-TMB
Sampling time (hr)
1,2,3-TMB exposure
2,3-DMBA
2,6-DMBA
V(mg/hr)
SD
V (mg/hr)
SD
0
0.000
0.000
0.000
0.000
0-2
3.518
0.852
0.099
0.097
2-4
10.745
1.856
0.097
0.084
4-6
16.594
5.028
0.146
0.039
6-8
23.468
5.291
0.202
0.070
8-10
16.874
2.353
0.160
0.004
10-12
14.769
1.964
0.150
0.035
12-14
11.929
2.070
0.161
0.048
14-16
7.715
2.236
0.129
0.038
16-23
3.976
0.782
0.110
0.042
23-27
1.876
0.213
0.067
0.021
27-31
1.822
0.893
0.079
0.052
31-35
1.471
0.551
0.081
0.055
35-39
2.292
0.998
0.143
0.032
39-47
1.388
0.660
0.102
0.037
47-51
1.125
0.414
0.109
0.041
51-55
1.543
0.468
0.172
0.058
55-59
1.505
0.683
0.139
0.050
59-63
1.154
0.481
0.055
0.063
63-71
0.535
0.119
0.031
0.030
71-75
0.802
0.383
0.053
0.001
75-79
0.999
0.712
0.059
0.030
79-83
0.886
0.343
0.086
0.078
83-87
0.349
0.165
0.046
0.050
87-95
0.365
0.163
0.000
0.000
This document is a draft for review purposes only and does not constitute Agency policy.
B-174 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-51 (Continued): Characteristics and quantitative results for Kostrzewski
	 etal. ("1997")	
Sampling time (hr)
1,2,4-TMB exposure
2,4- and 2,5-DMBA
3,4-DMBA
V(mg/hr)
SD
V (mg/hr)
SD
0
0.000
0.000
0.000
0.000
0-2
6.632
3.069
19.949
5.489
2-4
12.931
4.315
22.731
4.536
4-6
21.148
7.067
26.906
6.525
6-8
29.263
9.240
35.346
11.017
8-10
16.616
11.451
12.082
10.205
10-12
15.619
2.935
6.198
2.325
12-14
17.328
2.218
6.029
2.135
14-16
13.832
2.176
4.415
1.372
16-23
7.023
2.565
2.520
1.043
23-27
4.052
0.674
1.870
0.525
27-31
2.570
0.760
2.005
0.460
31-35
2.209
0.666
1.523
0.610
35-39
1.211
1.075
1.247
0.895
39^47
1.262
0.256
0.957
0.099
47-51
1.174
0.459
0.953
0.623
51-55
0.370
0.228
0.659
0.231
55-59
0.928
0.327
0.936
0.515
59-63
1.591
1.162
1.286
0.391
63-71
0.948
0.276
0.869
0.141
71-75
1.122
0.049
0.851
0.246
75-79
0.748
0.441
0.422
0.231
79-83
1.082
0.733
0.744
0.328
83-87
-
-
-
-
87-95
-
-
-
-
This document is a draft for review purposes only and does not constitute Agency policy.
B-175 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-51 (Continued): Characteristics and quantitative results for Kostrzewski
	 etal. ("1997")	
Sampling time (hr)
1,3,5-TMB exposure
3,5-DMBA
V(mg/hr)
SD
0
0.000
0.000
0-2
3.538
0.833
2^4
8.854
2.955
4-6
12.334
3.905
6-8
19.204
6.092
8-10
19.413
6.329
10-12
23.535
7.606
12-14
22.460
3.254
14-16
16.941
4.350
16-23
10.790
3.116
23-27
6.908
2.691
27-31
6.558
3.657
31-35
3.983
2.367
35-39
3.946
2.073
39^47
3.110
0.838
47-51
3.244
1.140
51-55
2.343
1.355
55-59
3.669
1.882
59-63
2.436
1.303
63-71
1.600
1.305
71-75
1.025
0.639
75-79
1.044
0.825
79-83
0.750
0.645
83-87
-
-
87-95
-
-
Comments: Metabolites (DMBAs) measured in urine. Toxicokinetics studied over a range of exposures. Exposure duration
possibly not sufficient to detect other metabolic changes. Only one study subject per exposure group.
Source: Kostrzewski et al. (1997)
This document is a draft for review purposes only and does not constitute Agency policy.
B-176 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
B.7. ANIMAL TOXICOKINETIC STUDIES
Table B-52. Characteristics and quantitative results for Dahl et al. (1988)
Study Design
Species
Sex
Exposure route
Dose range
Exposure duration
F344 Rats
M
2 rats
Inhalation
1-5,000 ppm 1,2,4-TMB
80 minutes per day
for 5 consecutive days
Additional study details
•	Male F344 rats weighing between 264 and 339 g were housed in polycarbonate cages for the duration of the
experiment.
•	Vapors were pumped into exposure chamber at flow rate of 400ml/min past the nose of each rat in the nose-only
exposure tube.
•	The amount of absorbed hydrocarbon vapor was calculated from the flow rate and the output from the nose-only
tube as measured by gas chromatography every minute during each 80 minute exposure.
•	Concentrations were increased each day. Days 1-5 concentrations were lppm, lOppm, lOOppm, lOOOppm, and
5000ppm respectively.
•	1,2,4-TMB uptake in one rat was observed to be 11.5±2 nmol/kg/min/ppm. For the second rat, uptake was
observed to be 15.7±2.4 nmol/kg/min/ppm.
Comments: Study duration was short term (5 days). Reported values for uptake represent averages of uptake throughout
experiment, despite the widely differing doses administered. This makes it difficult to quantify dose-specific uptake. Statistical
power is limited because only two rats were used.
Source: Dahl et al. (1988)
This document is a draft for review purposes only and does not constitute Agency policy.
B-177 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-53. Characteristics and quantitative results for Eide and Zahlsen et al. (1996)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Sprague-
Dawley rats
M
4 per
dose
Inhalation
0, 75,150, 300, 450 ppm
(0, 369, 738,1,476, or
2,214 mg/m3) 1,2,4-TMB
12 hr exposures in
inhalation chamber
Additional study details
•	Male Sprague-Dawley rats were exposed to 75,150, 300, or 450 ppm (0, 369, 738,1,476, or 2,214 mg/m3)
1,2,4-TMB in an inhalation chamber for 12 hrs.
•	Food and water was give ad libitum except during exposure, and animal weight ranged between 200 g and 250 g
prior to exposure.
•	Hydrocarbon concentration tissue concentrations were determined via head space gas chromatography. Daily
mean concentrations did not vary by more than ±5.3% from nominal concentrations.
•	1,2,4-TMB was found in higher concentrations in blood than n-nonane and trimethylcyclohexane.
Tissue 1,2,4-TMB concentrations following 12 hour 1,2,4-TMB inhalation exposure	
Exposure
Blood
(|imol/kg)
Brain
(|imol/kg)
Liver
(|imol/kg)
Kidneys
(|imol/kg)
Fat
(|imol/kg)
75ppm (369 mg/m
14.1
23.6
53.4
53.4
516
150 ppm (738 mg/m
Kr1" \' '"to/ ¦¦¦ /
300 ppm (1,476 mg/m3
450 ppm (2,214 mg/m3
57.5
97.5
123.1
168.5
3,806
115.5
220.9
256.3
282.4
12,930
221.3
400.2
468.6
492.5
19,270
Comments: Fat was analyzed and shown to retain higher concentrations of 1,2,4-TMB than all other tissues. Multiple exposure
concentrations were tested and multiple tissues were analyzed. No data on urinary elimination. No data on metabolites of
1,2,4-TMB.
Source: Eide and Zahlsen et al. (1996).
This document is a draft for review purposes only and does not constitute Agency policy.
B-178 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-54. Characteristics and quantitative results for Huo et al. (1989)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
3 rats per
dose
Oral, in olive oil
0.08 mmol/kg,
0.8 mmol/kg,
0.49 nCi/kg 1,2,4-TMB
3, 6,12, and 24 hrs
Additional study details






• Single doses of 14C labeled 1,2,4-TMB administered orally to rats.



• Tissues were analyzed at 3, 6,12, and 24 hr time points for the tissue distribution study and continuously for 24 hrs
in the metabolism study.
• Percent 1,2,4-TMB distributed to individual tissues determined via liquid scintillation counter, concentration of
metabolites analyzed via gas chromatography.
• 1,2,4-TMB was distributed widely throughout the body, though particularly high levels were found in adipose
tissue.
• Over 99% of radio-labeled material was recovered from urine within 24 hrs.


• Three most common metabolites were 3,4-DMHA (30.2%), 2,4-DMBA (12.7%), and 2,5-DMBA (11.7%).
Tissue distribution and urinary excretion following single oral dose of
14C-1,2,4-TMB

% Dose of radioactivity in tissue and urine (mean ± SD for three rats)
Tissue/Urine
3 hrs
6 hrs
12 hrs
24 hrs
Liver
2.76 ±0.39
2.69 ±0.60
1.54 ±0.38
0.13 ±0.04
Kidney
0.56 ±0.11
0.52 ±0.12
0.14 ±0.10
0.06 ± 0.05
Lung
0.10 ±0.03
0.06 ± 0.03
0.03 ± 0.03
0.01 ±0.01
Heart
0.03 ±0.01
0.01
-
-
Testis
0.09 ± 0.04
0.12 ±0.03
0.04 ± 0.04
-
Spleen
0.03 ± 0.02
0.03 ±0.01
0.01 ±0.01
-
Brain
0.08 ± 0.04
0.03 ± 0.02
0.03 ± 0.03
-
Stomach
2.39 ± 1.47
1.33 ±0.98
0.09 ± 0.06
0.04 ± 0.03
Intestine
2.96 ± 1.82
3.33 ± 1.31
1.39 ± 1.03
0.25 ±0.35
Serum
0.67 ±0.14
0.57 ±0.09
0.26 ±0.15
0.12 ±0.21
Muscle
2.38 ±0.23
1.88 ± 1.63
0.64 ±0.10
-
Skin
3.99 ± 1.51
2.29 ±0.98
0.16 ±0.25
-
Adipose Tissue
28.05 ± 9.28
26.31 ± 18.18
4.97 ±0.97
0.67 ±0.15
Urine
15.0 ± 1.1
32.6 ±7.9
50.7 ±7.9
99.8 ±4.1
This document is a draft for review purposes only and does not constitute Agency policy.
B-179
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-54 (Continued): Characteristics and quantitative results for Huo et al. (1989)
Concentration (|ig/g) radioactive material in tissue (mean ± SD)
Tissue
3 hrs
6 hrs
12 hrs
24 hrs
Liver
72 ±9
81 ±20
45 ± 12
5 ± 2
Kidney
68 ± 16
60 ± 13
17 ± 12
7 ± 6
Lung
17 ±9
12 ±6
4 ± 4
2 ±4
Heart
8 ± 2
2 ± 1
-
-
Testis
8 ± 4
11 ±2
3 ±4
-
Spleen
11 ±5
13 ±5
5 ± 5
-
Brain
11 ±5
6 ± 2
4 ± 4
-
Stomach
509 ±313
263 ± 218
18 ± 11
10 ± 7
Intestine
35 ±22
47 ± 17
21 ± 15
4 ± 6
Serum
17 ±3
15 ± 1
6 ± 3
3 ± 6
Muscle
6 ± 1
5 ± 4
1±0
-
Skin
20 ±7
12 ±4
1± 1
-
Adipose Tissue
200 ± 64
193 ±125
33 ±8
5 ± 1
Urinary metabolites of 1,2,4-TMB 24 hours after single oral dose in rats (values ± SD)
Metabolite
%Dose (0.08 mmol/kg
in urine
%Dose (0.8 mmol/kg) in urine
Free
Conjugated
Total
Free
Conjugated
Total
All rats
All rats
All rats
Rat 1
Rat 2
Rat 1
Rat 2
Rat 1
Rat 2
2,3,5-AND 2,4,5-TMPa
2.6 ± 1.2
5.1 ± 1.4
7.7 ±2.2
2.5
1.5
4.3
2.0
6.7
3.5
2,3,6-TMP
-
3.9 ±0.7
4.0 ±0.6
0.1
0.4
2.1
1.5
2.1
1.8
Total phenols
2.7 ± 1.1
9.0 ±2.0
11.8 ±2.9
2.6
1.9
6.3
3.5
8.8
5.3
2,4-DMBOHb
0.1 ±0.1
12.5 ±2.6
12.7 ±2.6
0.1
0.4
11.5
7.2
11.6
7.6
2,5-DMBOH
0.1 ±0.0
11.6 ±2.7
11.7 ±2.7
0.1
0.2
8.7
8.7
8.8
8.9
3,4-DM BOH
-
1.9 ±0.9
1.9 ±0.8
-
0.1
0.9
0.8
0.9
0.9
Total alcohols
0.2 ±0.1
26.0 ±5.5
26.3 ±5.4
0.1
0.7
21.1
16.8
21.2
17.5
2,4-DMBAc
0.8 ±0.1
5.2 ±2.0
6.0 ±2.0
0.8
2.5
6.8
1.5
7.6
4.0
2,5-D MBA
0.5 ±0.0
3.1 ± 1.3
3.6 ± 1.3
0.3
1.2
3.5
2.1
3.9
2.3
3,4-D MBA
0.2 ±0.1
0.7 ±0.2
0.8 ±0.2
0.1
0.2
0.5
0.2
0.5
0.4
Total benzoic acids
1.5 ±0.1
8.9 ±3.4
10.4 ± 3.3
1.2
3.9
10.8
3.8
11.9
6.7
2,4-DMHAd
5.0 ± 1.9
2.0 ± 1.0
7.0 ±2.6
3.3
2.7
4.8
1.2
8.1
3.7
2,5-DMAH
0.5 ±0.2
0.3 ±0.3
0.8 ±0.3
0.2
0.1
0.5
0.1
0.7
0.2
3,4-DMHA
27.3 ±8.4
3.3 ± 1.2
30.2 ±9.4
23.1
17.9
15.6
7.1
38.7
25.0
Total hippuric acids
32.7 ± 10.5
5.6 ±2.3
37.9 ± 12.1
26.6
20.8
20.9
8.4
47.5
28.9
Total metabolies
37.1 ± 11.4
49.5 ± 13.0
86.4 ± 23.0
30.4
27.2
59.1
32.4
89.5
58.4
Comments: Many tissues examined for radioactive and metabolite content. Multiple metabolites measured. Small numbers of rats
per dose group, particularly for the 0.8 mmol/kg group (n = 2). Time points only extend to 24 hours.
atrimethylphenol, bdimethylbenzoic alcohol, cdimethylbenzoic acid, ddimethylyhippuric acid.
Source: Huo et al. (1989)
This document is a draft for review purposes only and does not constitute Agency policy.
B-180 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-55. Characteristics and quantitative results for Mikulski and Wiglusz (1975)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
9 rats/dose
Unspecified
1.2 g/kg BW 1,2,3-,
1,2,4-, and 1,3,5-TMB
48 hrs
Additional study details
•	Rats weighing between 210 and 350 g were with treated with 1,2,3-, 1,2,4-, or 1,3,5-TMB at 1.2 g/kg body weight.
•	In one experiment, urine was collected every 4 hrs over a period of 3 days.
•	In a second experiment, metabolites were collected from rats were treated with mesitylene (1,3,5-TMB),
pseudocumene (1,2,4-TMB), or hemimellitene (1,2,3-TMB).
•	Phenobarbital was found to inhibits the metabolism of TMBs to dimethylhippuric acids
Urinary excretion of glycine, glucuronic, and sulphuric acid conjugates of TMBs
Not treated
% of dose
Glycine
conjugates
Glucuronides
mean ± SD)
Organic sulphates
Total
1,3,5-TMB
59.1 ±5.2
4.9 ± 1.0
9.2 ±0.8
73.2
1,2,4-TMB
23.9 ±2.3
4.0 ±0.5
9.0 ±2.1
36.9
1,2,3-TMB
10.1 ± 1.2
7.9 ± 1.3
15.0 ±3.5
33.0
Treated with Phenobarbital
1,3,5-TMB
35.1 ±3.4
9.8 ± 1.3
8.1 ± 1.4
53.0
1,2,4-TMB
30.6 ±2.5
12.2 ±2.8
17.4 ±3.6
60.2
1,2,3-TMB
5.7 ± 1.1
11.3 ±2.0
22.3 ±3.0
39.3
Comments; Kinetic data for all three TMB isomers and their metabolites were included in study. However, the authors did not
report method for dosing.
Source: Mikulski and Wiglusz (1975)
This document is a draft for review purposes only and does not constitute Agency policy.
B-181 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-56. Characteristics and quantitative results for Swiercz etal. (2002)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
lmp:DAK
Wistar rats
M
4/dose
Inhalation
25,100, or 250 ppm
(123, 492, 1,230 mg/m3)
1,2,4-TMB
6 hrs
Additional study details






• Two males and two females were exposed to 25,100, or 250 ppm (123,492,1,230 mg/m3
inhalation chamber for 6 hrs.
1,2,4-TMB in an
• 1,2,4-TMB concentration was determined via gas chromatography.


• Blood samples were taken from the tail vein at various time points up to 6 hrs after start of exposure.
• The half-life of 1,2,4-TMB elimination was found to increase with increasing exposure.

Air concentrations of 1,2,4-TMB and body mass of rats (means
± SD)


Biological material
1,2,4-TMB nominal
concentration
1,2,4-TMB actual
concentration (ppm)
Rat body weight (g)



25 ppm (123 mg/m3)
25 ±2
200 ± 10
Blood during 6 hr exposure
100 ppm (492 mg/m3)
109 ± 10
228 ± 10



250 ppm (1,230 mg/m3)
262 ±21
190 ± 12



25 ppm (123 mg/m3)
26 ±3
349 ±6
Blood after 6 hr exposure
100 ppm (492 mg/m3)
101 ±3
333 ± 18



250 ppm (1,230 mg/m3)
238 ±9
336 ±5



25 ppm (123 mg/m3)
27 ±3
355 ± 10
Urine after 6 hr exposure
100 ppm (492 mg/m3)
98 ±3
338 ± 10



250 ppm (1,230 mg/m3)
240 ±7
330 ± 12
Blood 1,2,4-TMB concentration: During 6 hour inhalation exposure (mean ± SD)

1,2,4-TMB concentration
Time
25 ppm
(123 mg/mg3)
100 ppm
(492 mg/mg3)
250 ppm
1,230 mg/mg3)
15 (min)
0.22 ±0.07
1.12 ±0.80
4.02 ± 0.85
30
0.33 ± 0.08
1.99 ± 1.09
4.87 ± 1.61
45
0.49 ±0.16
3.56 ±0.49
6.97 ± 1.22
1 (hrs)
0.53 ±0.14
4.29 ±0.60
8.67 ±0.54
2
0.73 ±0.16
5.10 ±0.34
14.5 ±2.6
3
0.80 ±0.17
6.22 ±0.70
17.8 ± 1.6
4
0.72 ±0.15
7.40 ± 1.05
20.0 ±0.5
5
0.79 ±0.22
7.72 ± 1.48
23.3 ±2.6
6
0.94 ±0.16
8.32 ± 1.34
23.6 ± 1.8
This document is a draft for review purposes only and does not constitute Agency policy.
B-182
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-56 (Continued): Characteristics and quantitative results for Swiercz etal.
	(2002)	
Blood concentrations of 1,2,4-TMB: Following 6 hour exposure (mean ± SD)

1,2,4-TMB concentration
Time
25 ppm
(123 mg/mg3)
100 ppm
(492 mg/mg3)
250 ppm
1,230 mg/mg3)
3 (min)
0.68 ±0.09
4.44 ± 1.54
20.9 ± 4.03
15
0.47 ± 0.04
3.72 ±0.96
20.7 ±5.13
30
0.40 ± 0.05
2.98 ±0.88
17.1 ±4.71
45
0.36 ±0.04
2.89 ±0.86
15.9 ±5.74
1 (hrs)
0.34 ±0.03
1.79 ±0.49
14.9 ± 3.77
2
0.23 ±0.04
1.25 ±0.33
10.2 ± 3.04
3
0.17 ±0.04
0.88 ±0.29
8.05 ± 2.25
4
0.12 ±0.02
0.61 ±0.20
6.13 ± 1.64
5
0.10 ±0.02
0.41 ±0.14
3.98 ±0.43
6
0.08 ± 0.02
0.33 ±0.06
3.20 ±0.52
Dimethylbenzoic acid (DMBA) urine concentrations: After 6 hour exposure to 1,2,4-TMB (mean ± SD)
1,2,4-TMB
2,5-DMBA (mg/L)
2,4-DMBA (mg/L)
3,4-DMBA (mg/L)
25 ppm (123 mg/m3)
23.6 ±8.6
37.6 ± 12.9
79.9 ±33.3
100 ppm (492 mg/m3)
54.0 ± 5.4
130.9 ±22.1
200.8 ± 25.8
250 ppm (1,230 mg/m3)
109.4 ±71.1
308.8 ±220.1
571.8 ±381.6
Comment: Metabolites (DMBAs) measured in urine. Appropriate number of animals per dose group (n =
possibly not sufficient to detect other metabolic changes.
4). Exposure duration
Source: Swiercz et al. (2002)



This document is a draft for review purposes only and does not constitute Agency policy.
B-183 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-57. Characteristics and quantitative results for Swiercz etal. (2003)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
4/dose
Inhalation
25,100, or 250 ppm
(123, 492,1,230 mg/m3)
1,2,4-TMB
6 hrs or 4 weeks
Additional study details
•	Male Wistar rats were exposed to either 25,100, or 250 ppm (123, 492,1,230 mg/m3) pseudocumene (1,2,4-TMB)
in an inhalation chamber for either 6 hrs or 4 weeks.
•	Rats were sacrificed following exposure period and tissues were analyzed 1,2,4-TMB content via gas
chromatography.
•	Venous elimination was found to follow an open two-compartment model.
•	Within brain structures, the brainstem was found to contain the highest levels of 1,2,4-TMB.
Air concentrations of 1,2,4-TMB in inhalation chamber and body weight (mean ± SD)
Biological material
1,2,4-TMB nominal
concentration in
inhaled air
1,2,4-TMB actual
concentration in
inhaled air (ppm)
Rat body weight (g)
Arterial blood and brain
structure from rats after
6 hrs
25 ppm (123 mg/m3)
21 ± 2
219 ± 13
100 ppm (492 mg/m3)
116 ±5
180 ± 28
250 ppm (1,230 mg/m3)
215 ± 15
220 ± 24
Arterial blood and brain
structure from rats after
4 weeks
25 ppm (123 mg/m3)
24 ±3
327 ±21
100 ppm (492 mg/m3)
99 ±7
295 ±31
250 ppm (1,230 mg/m3)
249 ± 19
268 ±21
Liver, lung, and brain
homogenate after 6 hrs
25 ppm (123 mg/m3)
28 ± 1
227 ± 15
100 ppm (492 mg/m3)
123 ±9
246 ± 11
250 ppm (1,230 mg/m3)
256 ±7
228 ± 12
Liver, lung, and brain
homogenate after 4 weeks
25 ppm (123 mg/m3)
25 ±2
310 ± 10
100 ppm (492 mg/m3)
103 ±8
328 ± 23
250 ppm (1,230 mg/m3)
249 ± 13
320 ± 20
Venous blood collected
following 4 week exposure
25 ppm (123 mg/m3)
24 ±3
321 ± 6
100 ppm (492 mg/m3)
99 ±7
300 ± 22
250 ppm (1,230 mg/m3)
249 ± 19
373 ± 48
This document is a draft for review purposes only and does not constitute Agency policy.
B-184
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-57 (Continued): Characteristics and quantitative results for Swiercz etal.
	(2003)	
Venous blood 1,2,4-TMB concentrations after 4 week inhalation exposure

1,2,4-TMB concentration mean ± SD
Time
25 ppm
(123 mg/mg3)
100 ppm
(492 mg/mg3)
250 ppm
1,230 mg/mg3)
3 (min)
0.56 ±0.18
4.06 ± 0.46
13.77 ±3.34
15
0.43 ±0.10
3.73 ± 1.21
11.82 ±3.05
30
0.33 ±0.03
3.02 ± 1.43
8.28 ±2.07
45
0.28 ±0.05
2.86 ±0.89
7.21 ± 1.84
1 (hr)
0.22 ±0.02
2.62 ±0.82
6.27 ± 1.72
2
0.17 ±0.06
1.83 ±0.17
4.50 ± 1.04
3
0.11 ±0.04
0.88 ±0.24
3.17 ±0.76
4
0.07 ± 0.04
0.64 ±0.21
1.73 ±0.37
5
0.07 ±0.01
0.39 ±0.11
1.30 ±0.22
6
0.06 ± 0.02
0.37 ±0.14
1.25 ±0.22
Liver, lung, and brain homogenates and arterial blood 1,2,4-TMB concentrations following inhalation exposure
(mean ± SD)
Exposure
25 ppm
(123 mg/mg3)
100 ppm
(492 mg/mg3)
250 ppm
1,230 mg/mg3)
Blood 6 hrs (mg/L)
0.31 ±0.12
1.24 ±0.41
7.76 ± 1.64
Blood 4 weeks (mg/L)
0.33 ±0.11
1.54 ±0.32
7.52 ±2.11
Brain 6 hrs (mg/kg)
0.49 ± 0.06
2.92 ±0.73
18.34 ± 1.92
Brain 4 weeks (mg/kg)
0.45 ± 0.05
2.82 ± 0.40
18.63 ±4.27
Liver 6 hrs (mg/kg)
0.44 ± 0.01
7.13 ± 1.31
28.18 ±5.34
Liver 4 weeks (mg/kg)
0.45 ±0.15
3.00 ± 0.49*
22.47 ±4.10
Lung 6 hrs (mg/kg)
0.43 ±0.11
4.14 ±0.54
18.90 ±3.72
Lung 4 weeks (mg/kg)
0.47 ± 0.20
3.74 ±0.82
22.47 ±4.10
1,2,4-TMB in various brain structures following 1,2,4-TMB inhalation exposure

1,2,4-TMB concentration (mg/kg), mean ± SD
Brain structure (time)
25 ppm
(123 mg/mg3)
100 ppm
(492 mg/mg3)
250 ppm
1,230 mg/mg3)
Brain stem (6 hrs)
0.54 ±0.11
3.38 ±0.84
26.91 ±5.33
Temporal cortex (6 hrs)
0.31 ±0.06*
2.30 ±0.71
13.54 ±2.33*
Hippocampus (6 hrs)
0.28 ±0.09*
1.89 ±0.29*
12.99 ±2.18*
Cerebellum (6 hrs)
0.32 ±0.09*
1.99 ± 0.40*
12.91 ±2.05*
Brain stem (4 weeks)
0.38 ±0.23
2.33 ± 1.24
21.95 ±3.81
Temporal cortex (4 weeks)
0.25 ±0.07
2.03 ±0.66
15.71 ±3.54
Hippocampus (4 weeks)
0.41 ±0.27
3.03 ± 0.48
12.44 ±2.63*
Cerebellum (4 weeks)
0.33 ±0.05
3.20 ± 0.40
10.85 ± 2.47*
Comments: Adipose tissue was not examined for 1,2,4-TMB content. Metabolite concentration was not measured. No control
group.
P < 0.05 in comparison to brainstem
Source: Swiercz et al. (2003).
This document is a draft for review purposes only and does not constitute Agency policy.
B-185 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-58. Characteristics and quantitative results for Swiercz et al. (2006)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
IMP:WIST
Wistar rats
M
5/dose
Inhalation
25,100, or 250 ppm
(123, 492, 1,230 mg/m3)
1,3,5-TMB
6 hrs or 4 weeks
Additional study details
•	Male Wistar rats were exposed to either 0, 25,100, or 250 ppm (123, 492,1.230 mg/m3) mesitylene (1,3,5-TMB) in
an inhalation chamber for either 6 hrs or 4 weeks.
•	Rats were sacrificed following exposure period and tissues were analyzed for 1,3,5-TMB content via gas
chromatography.
•	1,3,5-TMB was found in the lungs in greater quantities following repeated exposures at 100 ppm (492 mg/m3) and
250 ppm (1.230 mg/m3).
Air concentrations of 1,3,5-TMB in inhalation chamber and body weight (mean ± SD)
Biological material
1,3,5-TMB nominal
concentration in
inhaled air
1,3,5-TMB actual
concentration in
inhaled air (ppm)
Rat body weight (g)
Liver, lung, and kidney
homogenates after 6 hr
exposure
Control
0
246 ±9
25 ppm (123 mg/m3)
25 ±2
254 ± 11
100 ppm (492 mg/m3)
97 ± 14
242 ± 14
250 ppm (1,230 mg/m3)
254 ± 20
249 ±7
Liver, lung, and kidney
homogenates after 4 week
exposure
Control
0
331 ± 17
25 ppm (123 mg/m3)
23 ±2
311 ±26
100 ppm (492 mg/m3)
101 ±8
320 ± 38
250 ppm (1,230 mg/m3)
233 ± 16
328 ±21
Blood collected after 6 hr
exposure
Control
0
251 ± 7
25 ppm (123 mg/m3)
24 ±2
250 ± 5
100 ppm (492 mg/m3)
101 ±7
239 ±7
250 ppm (1,230 mg/m3)
240 ± 22
249 ± 10
Blood collected after
4 week exposure
Control
0
310 ± 9
25 ppm (123 mg/m3)
23 ±2
307 ± 15
100 ppm (492 mg/m3)
101 ±8
310 ± 33
250 ppm (1,230 mg/m3)
233 ± 16
309 ± 19
Urine collected after 6 hr
exposure
Control
0
280 ± 9
25 ppm (123 mg/m3)
25 ±2
278 ± 10
100 ppm (492 mg/m3)
102 ± 10
335 ± 15
250 ppm (1,230 mg/m3)
238 ±27
273 ± 18
Urine collected after
4 week exposure
Control
0
310 ± 10
25 ppm (123 mg/m3)
25 ±2
295 ± 15
100 ppm (492 mg/m3)
102 ± 10
331 ± 19
250 ppm (1,230 mg/m3)
238 ±27
320 ± 28
This document is a draft for review purposes only and does not constitute Agency policy.
B-186
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-58 (Continued): Characteristics and quantitative results for Swiercz etal.
	(2006)	
Concentrations of 1,3,5-TMB in various tissues after exposure to 1,3,5-TMB (mean ± SD)
1,3,5-TMB exposure
duration and target
concentration
Liver (|ig/g
tissue)
Lung (|ig/g tissue)
Kidney (|ig/g tissue)
Blood (|ig/g tissue)
6 Hrs—25 ppm
(123 mg/m3)
0.30 ± 0.07
0.31 ±0.12
4.49 ± 1.93
0.31 ±0.12
6 Hrs—100 ppm
(492 mg/m3)
3.09 ±0.50
2.87 ±0.57
13.32 ±2.58
3.06 ±0.65
6 Hrs—250 ppm
(1,230 mg/m3)
17.00 ± 6.08
17.36 ±5.56
31.80 ± 9.44
13.36 ± 1.54
4 Wks—25 ppm
(123 mg/m3)
0.22 ±0.01
0.42 ±0.12
1.73 ±0.30*
0.31 ±0.08
4 Wks—100 ppm
(492 mg/m3)
3.01 ±0.58
1.99 ±0.75
15.61 ±2.14
2.30 ±0.52
4 Wks—250 ppm
(1,230 mg/m3)
12.98 ±4.16
11.20 ±3.61
35.97 ±8.53
7.55 ± 1.43**
Concentrations of 3,5-DMBA in various tissues after exposure to 1,3,5-TMB (mean ± SD)
1,3,5-TMB exposure
duration and target
concentration (ppm)
Liver (|ig/g
tissue)
Lung (ng/g tissue)
Kidney (|ig/g tissue)
Urine (mg/18 hrs)
6 Hrs—25 ppm
(123 mg/m3)
12.62 ± 1.62
2.87 ±0.55
8.77 ±0.99
0.52 ±0.03
6 Hrs—100 ppm
(492 mg/m3)
26.05 ±2.77
5.50 ±0.55
27.01 ±9.86
3.66 ±0.57
6 Hrs—250 ppm
(1,230 mg/m3)
36.92 ± 1.61
13.39 ± 1.90
60.91 ± 19.78
10.99 ± 3.90
4 Wks—25 ppm
(123 mg/m3)
6.52 ±0.67**
3.69 ± 1.21
11.06 ±4.33
0.83 ±0.15*
4 Wks—100 ppm
(492 mg/m3)
21.67 ±3.14**
8.90 ±0.98**
31.03 ± 18.56
4.36 ±0.86
4 Wks—250 ppm
(1,230 mg/m3)
53.07 ±5.41**
19.79 ±2.70**
82.10 ± 14.48
11.92 ±3.05
Venous blood 1,3,5-TMB concentration following 6 hr 1,3,5-TMB inhalation exposure
Time
1,3,5-TMB (|ig/mL)
25 ppm
(123 mg/mg3)
100 ppm
(492 mg/mg3)
250 ppm
1,230 mg/mg3)
3 (min)
0.31 ±0.12
3.06 ±0.65
13.36 ± 1.54
15
0.26 ±0.13
2.51 ±0.17
13.05 ± 1.61
30
0.15 ±0.04
2.35 ±0.57
12.06 ± 1.23
45
0.10 ±0.03
1.41 ±0.27
10.53 ± 1.71
1 (hrs)
0.06 ± 0.02
1.35 ±0.30
8.85 ±0.90
2
0.04 ± 0.02
1.34 ±0.39
6.14 ±0.53
3
ND
0.79 ±0.30
4.54 ±0.67
4
ND
0.57 ±0.14
3.49 ± 1.16
5
ND
0.38 ±0.14
2.31 ±0.67
6
ND
0.20 ± 0.04
0.76 ±0.06
This document is a draft for review purposes only and does not constitute Agency policy.
B-187 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-58 (Continued): Characteristics and quantitative results for Swiercz etal.
	(2006)	
Venous blood 1,3,5-TMB concentration following 4 week 1,3,5-TMB inhalation exposure
Time
1,3,5-TMB (ng/mL)
25 ppm
(123 mg/mg3)
100 ppm
(492 mg/mg3)
250 ppm
1,230 mg/mg3)
3 (min)
0.31 ±0.08
2.30 ±0.52
7.55 ± 1.43
15
0.26 ±0.03
1.83 ± 0.47
6.51 ± 1.50
30
0.19 ±0.02
1.57 ±0.39
4.56 ±0.98
45
0.17 ±0.03
1.41 ±0.13
3.65 ±0.62
1 (hrs)
0.12 ±0.03
1.33 ±0.15
3.69 ± 1.25
2
0.05 ±0.01
0.95 ±0.22
3.14 ±0.64
3
ND
0.72 ±0.17
2.28 ±0.19
4
ND
0.41 ±0.11
1.74 ±0.17
5
ND
0.39 ±0.05
1.23 ±0.34
6
ND
0.29 ±0.13
1.14 ±0.20
Comments: Kinetics of 1,3,5-TMB elimination are reported and discussed in detail. Extensive analysis of 3,5-DMBA. Adipose tissue
was not examined for 1,3,5-TMB content.
*p < 0.05; ** p < 0.01 (respectively: significantly differed from the signal exposure (Student's t-test).
Source: Swiercz et al. (2006)
This document is a draft for review purposes only and does not constitute Agency policy.
B-188 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-59. Characteristics and quantitative results for Tsujimoto etal. (2000)
Study design
Species
Sex
Exposure route
Dose range
Exposure duration
Sic Wistar
rats
M
4 per dose
i.p. in corn oil
0,0.3,1, and 3
mmol/kg BW 1,2,4-TMB
2 days
Additional study details
•	Groups of four male Wistar rats dosed with 0,0.3,1, or 3 mmol/kg BW 1,2,4-TMB.
•	Urine samples collected for 2 days.
•	HPLC used to quantify amount of dimethylbenzyl mercapturic acid in urine.
Urinary excretion of dimethylbenzyl mercapturic acid in 1,2,4-TMB treated rats
Dose (mmol/kg)
% of dose ± SD
0-24 hr
24-48 hr
Total
0.3
14.0 ± 1.2
ND
14.0 ± 1.2
1.0
19.4 ± 1.8
ND
19.4 ± 1.8
3.0
16.7 ±6.2
2.5 ± 1.6
19.2 ±4.8
Comments: This study observed a marked decrease in dimethylbenzyl mercapturic acid excretion between 24 and 48 hours
following exposure. Authors do not report specific speciation data for 2,4-, 2,5-, or 3,4-dimethylbenzyl mercapturic acid.
Source: Tsujimoto et al. (2000)
This document is a draft for review purposes only and does not constitute Agency policy.
B-189 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-60. Characteristics and quantitative results for Tsujimoto etal. (2005)
Study Design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar
rats
M
4 per dose
i.p. in corn oil
0, 0.3,1, and 3
mmol/kg BW given 1,2,3
or 1,3,5-TMB
-
2 days
Additional study details










• Groups of four male Wistar rats were given 1,2,3- or 1,3,5-TMB intraperitoneal^ in doses of 0, 0.3,1, or
3 mmol/kg BW.
• Urine samples collected for 2 days, then analyzed for trimethylphenols (TMP) via GC-MS

Urinary excretion (% of dose
± SD) of phenolic metabolites in 1,2,3-TMB treated rats



Dose
(mmol/k

2,3,4-Trimethylphenol
3,4,5-T rimethylphenol
l)
0-24 hr
24-48 hr
Total
0-24 hr
24-48 hr
Total
0.3
5.90 ±2.62
0.46 ±0.34
6.36 ±2.92
ND
ND
ND
1.0
7.93 ± 5.00
0.35 ±0.16
8.28 ±4.85
<0.24
ND
<0.24
3.0
6.20 ± 3.45
0.57 ±0.34
6.77 ±3.60
<0.19
<0.04
<0.19

Urinary excretion (% of dose
± SD) of phenolic metabolites in 1,3,5-TMB treated rats



2,4,6-Trimethylphenol
Dose (mmol/kg)
0-24 hr
24-48 hr
Total
0.3
7.04 ± 1.24
0.53 ±0.29
7.57 ±0.99
1.0
4.39 ±0.61
0.51 ±0.12
4.90 ±0.64
3.0
3.32 ±0.58
0.82 ±0.34
4.14 ±0.67
Comments: This study observed a marked decrease in TMP excretion between 24 and 48 hours following exposure. This study does
not include data for 1,2,4 TMB and phenolic metabolites. Variation between rats (high standard deviation) within exposure groups.
ND - not detected
Source: Tsujimoto et al. (2005)
This document is a draft for review purposes only and does not constitute Agency policy.
B-190 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-61. Characteristics and quantitative results for Tsujino et al. (2002)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Wistar rats
M
3 for Experiment 1,
36 for Experiment 3
(shown below in Figure 3)
Dermal
(via saturated
cotton)
1 mL kerosene
0,1, 3, or 6 hrs
Additional study details
•	In first experiment, rats were dermally exposed to kerosene on a saturated, sealed piece of cotton for 1 hr to
analyze TMB and aliphatic hydrocarbon (AHC) dermal absorption.
•	In second experiment, 44 rats were divided into four groups which varied by exposure duration, post-exposure
time, and/or exposure either before or after death.
•	TMBs were detected at greater levels than AHCs, and were only detected in traces following post-mortem
exposure.
•	Trace concentrations of TMBs following post-mortem exposure suggest TMB must circulate in blood before being
distributed to organs.
1 hr exposure and ratio of TMBs to internal standard (o-xylene d10) (mean ± SD)
Tissue source
Post-mortem samples spiked with
kerosene (positive control)
Post-mortem samples following
dermal exposure
Blood
3.6 ± 1.6
0.4 ±0.4
Brain
1.2 ±0.5
0.14 ±0.05*
Lung
1.2 ±0.5*
0.09 ± 0.03
Liver
1.1 ±0.5
0.3 ±0.09*
Spleen
0.7 ±0.3
0.1 ±0.04
Kidney
1.0 ±0.4
0.5 ±0.1*
Muscle
1.2 ±0.5*
0.09 ± 0.02
Adipose
0.9 ±0.3*
0.15 ±0.07
OVERALL
1.4 ±0.3*
0.21 ±0.05*
This document is a draft for review purposes only and does not constitute Agency policy.
B-191
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-61 (Continued): Characteristics and quantitative results for Tsujino et al.
	(2002)	
1,2,4-TMB in Various Tissues following 1 hr Exposure and Ante vs. Post-Mortem Exposure
Figure 3.1,2,4-TMB levels in rats immediately after 1 hour of dermal exposure to kerosene are compared between
ante-mortem (group I) and post-mortem (group IV) groups.
Data represent mean ± SE. The data were analyzed using two-way ANOVA (* p < 0.05, ** p < 0.01)
600
O)
CD
CD
o
o
TD
13

CO
Q_
¦ Antemortem exposure
0 Postmortem exposure
300
Blood Brain Lung Liver Kidney Spleen Muscle Adipose
Source: Tsujino et al. (2002)
Comments: Number of tissues were tested and number of animals used in the ante- and post-mortem 1 hr exposure groups
(20 and 16 respectively). The authors conclude that their data shows that TMBs are dispersed throughout the body by circulation in
blood following dermal exposure. Small number of animals used to determine dermal absorption at 1 hour (n = 3). No data
provided for effects of exposure (if any).
*, **, *** p < 0.05, p < 0.01, p < 0.001
Source: Tsujimoto et al. (2005)
This document is a draft for review purposes only and does not constitute Agency policy.
B-192 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-62. Characteristics and quantitative results for Zahlsen et al. (1990)
Study design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Sprague-
Dawley rats
M
24
Inhalation
1,000 ppm (4,920 mg/m3)
1,2,4-TMB
12 hr exposures on days 1, 3, 7,
10, and 14
Additional Study details
• Male Sprague-Dawley rats were exposed to 1,000 ppm (4,920 mg/m3) 1,2,4-TMB in an inhalation for 12 hrs on
days 1, 3, 7,10, and 14.
•	Food and water was given ad libitum except during exposure, and animal weight ranged between 150 g and
200 g prior to exposure on day 1.
•	Hydrocarbon concentration in blood was determined via head space gas chromatography. Daily mean
concentrations did not vary by more than ±10% from nominal concentrations.
•	Multiple exposures to 1,2,4-TMB resulted in decreases in blood concentrations following subsequent
exposures, possibly due to the induction of metabolic enzymes that play a role in the metabolism of 1,2,4-TMB.
Figure 1. Blood concentrations (+SD) of n-nonane, 1,2,4-TMB, and 1,2,4-trimethylcyclohexane following 12 hr
exposures on days 1, 3, 7,10, and 14.
Source: Zahlsen et al. (1990)
g 700
8 500
® 400
Q
300
iTMB
£ 200
| 100
: n~C9
*TMCH
2
O
O
TIME OF EXPOSURE (days)
This document is a draft for review purposes only and does not constitute Agency policy.
B-193
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-62 (Continued): Characteristics and quantitative results for Zahlsen et al.
	(1990)	
Figure 2. Brain concentrations (+SD) of n-nonane, 1,2,4-TMB, and 1,2,4-trimethylcyclohexane following 12 hr
exposures on days 1, 3, 7,10, and 14.
Source: Zahlsen et al. (1990)
cn 1600
1400
1200
£ 1000
600
* 400
UJ
o 200
5	10
TIME OF EXPOSURE (days)
X n-C9
ITMCH
TMB
15
Figure 3. Perirenal fat concentrations (+SD) of n-nonane, 1,2,4-TMB, and 1,2,4-trimethylcyclohexane following 12
hr exposures on days 1, 3, 7,10, and 14.
Source: Zahlsen et al. (1990)
_ 70000
o>
g 60000
I? 50000
if
z 40000
g 30000-
<
^ 20000
•z.
O 10000
o
O
TMCH
5	10
TIME OF EXPOSURE (days)
This document is a draft for review purposes only and does not constitute Agency policy.
B-194 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-62 (Continued): Characteristics and quantitative results for Zahlsen et al.
	(1990)	
Brain:blood and fat:blood TMB distribution after 12 hr exposure at end of day 14
Compound
Concentration ratio
Brain:blood TMB ratio
2.0
Fat:blood TMB ratio
63
Comments: Perirenal fat was analyzed and shown to retain higher concentrations of 1,2,4-TMB than blood. Exposure was not
continuous (only occurred on days 1, 3, 7,10, and 15). Only one exposure concentration (1,000 ppm [4,920 mg/m3]) was tested,
and there were no control groups.
Source: Zahlsen et al. (1990).

This document is a draft for review purposes only and does not constitute Agency policy.
B-195 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table B-63. Characteristics and quantitative results for Zahlsen etal. (1992)
Study Design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Sprague-
Dawley rats
M
4/time
point
Inhalation
100 ppm C9-aromate
12 hours/day for 3 days
Additional study details




• Food and water was given ad libitum, except during exposure.

• Rats weighed between 150-200 g and were between 40 and 50 days of age.

• 4 rats were housed in each cage, and each exposure chamber contained 4 cages; 16 rats were present at the
beginning of exposure.
• At each time point, 4 rats were sacrificed and their tissues analyzed for C9-aromate presence.



C9-Aromate Concentration in Rat Tissues at Various Time Points (Mean ± S.D)
Observation


100 ppm C9 Exposure Group
Blood Day 1
14.2±0.7
Blood Day 2
12.6±0.9
Blood Day 3
17.1±2.2
Blood Reca
0.2±0.1
Brain Day 1
38.1±1.5
Brain Day 2
34.9±3.9
Brain Day 3
36.5±2.2
Brain Rec
nd
Liver Day 1
41.0±4.5
Liver Day 2
30.5±3.4
Liver Day 3
35.4±2.4
Liver Reca
0.6±0.1
Kidney Day 1
113.8±26.5
Kidney Day 2
142.0±35.2
Kidney Day 3
103.6±18.8
Kidney Reca
2.0±0.3
Fat Day 1
1741±329
Fat Day 2
1375±88
Fat Day 3
1070±93
Fat Reca
120±52
Comments: Data was collected immediately following exposure and 12 hours following exposure, providing insight into metabolic
clearance and excretion. Study duration was short term (5 days), making it difficult to determine if tissue concentration changes
following chronic exposure.
aRec=After 12 hour recovery
Source: Zahlsen et al. (1992)
This document is a draft for review purposes only and does not constitute Agency policy.
B-196 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
B.8. ANIMAL AND HUMAN TOXICOKINETIC STUDIES
Table B-64. Characteristics and quantitative results for Meulenberg and
Vijverberg (2000)
Study Design
Species
Sex
N
Exposure route
Dose range
Exposure duration
Rat and
Human
F &
M
Varies
n/a
Not given
Not given
Additional study details
•	Authors examined partition coefficients for many volatile organic compounds from multiple studies.
•	1,2,3-, 1,2,4-, and 1,3,5-TMB were among the volatile organic compounds considered for review.
•	Partition coefficients for blood, fat, brain, liver, muscle, and kidney were reported for both rats and humans.
Observation
Partition Coefficients for 1,2,3-, 1,2,4- and 1,3,5-TMB
1,2,3-TMB
1,2,4-TMB
1,3,5-TMB
Reported and predicted partition coefficients For oil, saline, and air
Poil:air
10,900a
10,200a
9,880a
~
' saline:air
2.73a
1.61a
1.23a
Reported and predicted PtiSSue:air values for various human tissues
Blood
66.5a
59.la
43a
Fat
4879b
4566
4423
Brain
220
206
199
Liver
306
286
277
Muscle
155
144
140
Kidney
122
114
110
Reported and predicted PtiSSue:air values for various rat tissues
Blood
62.6
55.7
55.7
Fat
6484
6068
5878
Brain
591
552
535
Liver
288
269
260
Muscle
111
104
100
Kidney
1064
995
963
Comment: This study evaluated a number of parameters, presenting predicted partition coefficients for blood, fat, brain, liver,
muscle, and kidney tissue in both humans and rats. Reported values based on single trial.
a Averaged values as reported bv Jarnberg and Johanson (1995).
bAII other values predicted bv Meulenberg and Viiverberg (2000).
This document is a draft for review purposes only and does not constitute Agency policy.
B-197 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Toxicological Review ofTrimethylbenzene
APPENDIX C. DOSE-RESPONSE MODELING FOR THE
DERIVATION OF REFERENCE VALUES FOR EFFECTS
OTHER THAN CANCER AND THE DERIVATION OF
CANCER RISK ESTIMATES
C.l. BENCHMARK DOSE MODELING SUMMARY
This appendix provides technical detail on dose-response evaluation and determination of
points of departure (POD) for relevant neurological, hematological, and developmental toxicity
endpoints in the TMB database. The endpoints were modeled using the U.S. EPA's Benchmark
Dose Software (BMDS, version 2.2). Sections C.l.1.1 and C.l.1.2 (non-cancer) describe the
common practices used in evaluating the model fit and selecting the appropriate model for
determining the POD, as outlined in the Benchmark Dose Technical Guidance Document fU.S.
EPA. 2012a). In some cases it may be appropriate to use alternative methods, based on
statistical judgement; exceptions are noted as necessary in the summary of the modeling
results.
C.l.l. Non-Cancer Endpoints
C.l.1.1. Evaluation of Model Fit
For each continuous endpoint, BMDS continuous models were fitted to the data using the
maximum likelihood method. Model fit was assessed by a series of tests as follows. For each
model, first the homogeneity of the variances was tested using a likelihood ratio test (BMDS
Test 2). If Test 2 was not rejected (x2 p-value > 0.10), the model was fitted to the data assuming
constant variance. If Test 2 was rejected (x2 p-value < 0.10), the variance was modeled as a
power function of the mean, and the variance model was tested for adequacy of fit using a
likelihood ratio test (BMDS Test 3). For fitting models using either constant variance or
modeled variance, models for the mean response were tested for adequacy of fit using a
likelihood ratio test (BMDS Test 4, withx2 p-value <0.10 indicating inadequate fit). Other
factors were also used to assess the model fit, such as scaled residuals, visual fit, and adequacy
of fit in the low-dose region and in the vicinity of the BMR.
This document is a draft for review purposes only and does not constitute Agency policy.
C-l	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
C.1.1.2. Model Selection
1	For each endpoint, the BMDL estimate (95% lower confidence limit on the BMD, as
2	estimated by the profile likelihood method) and AIC value were used to select a best-fit model
3	from among the models exhibiting adequate fit. If the BMDL estimates were "sufficiently close,"
4	that is, differed by at most threefold, the model selected was the one that yielded the lowest
5	Akaike Information Criterion (AIC) value. If the BMDL estimates were not sufficiently close, the
6	lowest BMDL was selected as the POD.
7
This document is a draft for review purposes only and does not constitute Agency policy.
C-2	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-l. Non-cancer endpoints selected for dose-response modeling for 1,2,3-TMB,
1,2,4-TMB, and 1,3,5-TMB
Study, Species (generation)
/ Sex, and Endpoint
Internal Doses, External Exposure Concentrations, and Effect Data
Korsak and Rvdzvriski (1996)
1,2,4-TMB
Rat (Wistar) / Male
Internal Dose (mg/L)
0
0.1272
0.8666
5.4424
CNS: Pawlick (seconds)
No. of animals
Mean ± SD
9
15.4 ±5.8
10
18.2 ±5.7
9
27.6 ±4.6
10
30.1 ±6.1
1,2,3-TMB
Rat (Wistar) / Male
Concentration (mg/mB)
0
123
492
1230
CNS: Pawlick (seconds)
No. of animals
Mean ± SD
30
9.7 ±2.1
20
11.8 ±3.8
10
16.3 ±6.3
10
17.3 ±3.4
Korsak et al. (2000a) - 1,2,4-TMB
Rat (Wistar) / Male
Internal Dose (mg/L)
0
0.1339
0.8671
5.2481
Decreased RBC
(106/cm3 [10s cells/mL])
No. of animals
Mean ± SD
10
9.98 ± 1.6
10
9.84 ± 1.82
10
8.50 ± 1.11
10
7.70+1.38
Rat (Wistar) / Female
Internal Dose (mg/L)
0
0.1335
0.8899
5.5189
Clotting time (seconds)
No. of animals
Mean ± SD
10
30 ± 10
10
23 ±4
10
19 ±5
10
22 ±7
Korsak et al. (2000b) - 1.2.3-TMB
Rat (Wistar) / Male
Concentration (mg/mB)
0
128
523
1269
Decreased segmented
neutrophils (%)
No. of animals
Mean ± SD
10
24.8 ±4.5
10
25.4 ±5.8
10
20.7 ±5.8
10
17.7 ±8.3
Increased reticulocytes
(%)
No. of animals
Mean ± SD
10
2.8 ± 1.3
10
2.1 ± 1.7
10
3.8 ±2.1
10
4.5 ± 1.8
Rat (Wistar) / Female
Concentration (mg/mB)
0
128
523
1269
Decreased segmented
neutrophils (%)
No. of animals
Mean ± SD
10
23.1 ±6.1
10
19.7 ±3.4
10
16.4 ±4.2
10
11.9 ±7.1
This document is a draft for review purposes only and does not constitute Agency policy.
C-3
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-l (Continued): Non-cancer endpoints selected for dose-response modeling
	for 1,2,3-TMB, 1,2,4-TMB, and 1,3,5-TMB	
Study, Species (generation)
/ Sex, and Endpoint
Internal Doses, External Exposure Concentrations, and Effect Data
Saillenfait et al. (2005)
1,2,4-TMB
F1 rat pups and Dams (SD)
Concentration (mg/mB)
0
492
1471
2913
4408
Male fetal weight (g)
Number of liters
Mean ± SD
23
5.86 ±0.34
22
5.79 ±0.30
22
5.72 ±0.49
22
5.55 ±0.48
24
5.20 ±0.42
Female fetal weight (g)
Number of liters
Mean ± SD
23
5.57 ±0.33
22
5.51 ±0.31
22
5.40 ± 0.45
22
5.28 ±0.40
24
4.92 ±0.40
Maternal weight gain (g)
Number of dams
Mean ± SD
24
29 ± 12
22
31 ± 14
22
27 ± 12
22
15 ± 17
24
0± 14
1,3,5-TMB
F1 rat pups and Dams (SD)
Concentration (mg/mB)
0
497
1471
2974
5874
Male fetal weight (g)
Number of liters
Mean ± SD
21
5.80 ±0.41
22
5.76 ±0.27
21
5.50 ±0.31
17
5.39 ±0.55
18
5.10 ±0.57
Female fetal weight (g)
Number of liters
Mean ± SD
21
5.50 ±0.32
22
5.47 ±0.21
21
5.27 ±0.47
17
5.18 ±0.68
18
4.81 ±0.45
Maternal weight gain (g)
Number of dams
Mean ± SD
21
29 ± 14
22
30 ±9
21
20 ± 12
17
7 ±20
18
-12 ± 19
This document is a draft for review purposes only and does not constitute Agency policy.
C-4	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Toxicological Review ofTrimethylbenzene
For all endpoints from Korsak et al. (2000a; 19971 and Korsak and Rydzynski (19961.
external exposure concentrations were first converted into the internal dose metric of weekly
average venous blood concentration (mg/L), and these dose metrics were used as the dose
inputs for BMD modeling. Due to PBPK model insufficiency at the high dose (i.e., estimating
higher internal blood metrics compared to observed blood data), all high doses were dropped
prior to modeling (see Dose-Response Analysis section in Volume 1 for more detail). Section C.2
is included for comparison at the end of this appendix that includes BMD modeling results when
the high doses were not dropped. All modeling results (i.e., BMDs and BMDLs) for the Korsak
studies are provided in mg/L. As a PBPK model was not applied to the endpoints from
Saillenfait et al. (2005), modeling results for these endpoints are provided in mg/m3.
Additionally, as no PBPK model was available for 1,2,3-TMB, all endpoints from Korsak etal.
(2000b) are provided in mg/m3.
Comprehensive modeling results for all endpoints are provided on EPA's Health Effects
Research Online (HERO) database fU.S. EPA. 2011bl.
C.1.1.3. Model Selection
Below are tables summarizing the modeling results for the non-cancer endpoints modeled.
The following parameter restrictions were applied, unless otherwise noted.
• Continuous models: For the polynomial models, restrict beta's in the appropriate direction
(i.e., > 0 for responses that increase with dose, and < 0 for responses that decrease with
dose); for the Hill, power, and exponential models, restrict power > 1.
This document is a draft for review purposes only and does not constitute Agency policy.
C-5	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-2. Summary of BMD modeling results for increased latency to paw-lick in
male Wistar rats exposed to 1,2,4-TMB by inhalation for 3 months;
BMR = 1 SD change from control mean (constant variance, high dose
dropped), (Korsakand Rydzynski, 1996)
Model3
Goodness-of-fit
BMD1sd
(mg/L)
BMDL1sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.5045
122.2153
0.42102
0.328286
Of the models that provided
an adequate fit and a valid
BMDL estimate, the
Exponential model 4 was
selected based on lowest
BMDL (BMDLs differed by at
least 3-fold)
Exponential (M4)
n/ac
123.7699
0.233402
0.0864608
Lineard
Polynomial 2°
Polynomial 3°
Power
0.6236
122.010727
0.354545
0.259068
a Constant variance case presented (Test 2 p-value = 0.169). Selected model in bold; scaled residuals for
selected model for concentrations 0, 0.1272, and 0.8666 mg/L were 6.09 x 10"08, -1.09 x 10"08, and -3.65 x 10"08
respectively.
b For Exponential model 3, the estimate of d was 1 (boundary). The models in this row reduced to exponential
model 2.
°X2 test had insufficient degrees of freedom (due to estimated model parameters = dose groups). However,
inspection of scaled residuals and visual fit indicated appropriate model fit.
d For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
Data Source: (Korsakand Rydzynski. 1996)
Exponential Model 4 with 0.95 Confidence Level
Exponential
30
25
20
15
10
BMDL
BMD
O
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
dose
	14:44 03/12 2012	
Note: BMR = 1 SD change from control mean; dose shown in mg/L 1,2,4-TMB (high dose dropped). (Korsakand Rydzynski. 1996)
Figure C-l. Plot of mean response by dose for increased latency to paw-lick in male
Wistar rats, with the fitted curve for Exponential model 4 with constant variance.
This document is a draft for review purposes only and does not constitute Agency policy.
C-6	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1	Exponential Model.
2	(Version: 1.7; Date: 12/10/2009)
3	The form of the response function is: Model 2: Y[dose] = a * exp{sign * b * dose}
4	A constant variance model is fit.
5	Benchmark Dose Computations:
6	BMR = 1 estimated standard deviations from the control mean
7	BMD = 0.233402
8	BMDL at the 95% confidence level = 0.0864608
9	Parameter Estimates
Variable
Model
(Default) Initial Parameter
Values
Inalpha
3.13464
3.13464
rho
0
0
a
15.4
14.63
b
13.6063
2.69257
c
2.14406
1.98086
d
1
1
10 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std
Scaled Resid
0
9
15.4
15.4
5.8
4.794
6.09 x 10"08
0.1272
10
18.2
18.2
5.7
4.794
-1.09 x 10"08
0.8666
9
27.6
27.6
3.2
4.794
-3.65 x 10"08
11 Likelihoods of Interest
Model
Log(likelihood)
# Pa rams
AIC
A1
-57.88496
4
123.7699
A2
-56.10689
6
124.2138
A3
-57.88496
4
123.7699
R
-68.59968
2
141.1994
4
-57.88496
4
123.7699
12 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1 (Does response and/or variances differ among Dose
levels, A2 vs. R)
24.99
4
< 0.0001
Test 2 (Are Variances Homogeneous, A2 vs. Al)
3.556
2
0.169
Test 3 (Are variances adequately modeled, A2 vs. A3)
3.556
1
0.169
Test 4 (Does the model for the Mean fit, A3 vs. fitted)
0
0
n/a
This document is a draft for review purposes only and does not constitute Agency policy.
C-7	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-3. Summary of BMD modeling results for decreased red blood cells in male
Wistar rats exposed to 1,2,4-TMB by inhalation for 3 months; BMR = 1 SD
change from control mean (constant variance, high dose dropped),
(Korsak et al., 2000a)
Model3
Goodness-of-fit
BMD1sd
(mg/L)
BMDL1sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)
0.8653
59.81949
0.847227
0.467889
Of the models that provided
an adequate fit and a valid
BMDL estimate, the Linear
model was selected based
on lowest AIC
Exponential (M3)
n/ab
61.79073
0.870338
0.469066
Exponential (M4)
0.8653
59.81949
0.847227
0.184658
Linear
0.8864
59.811121
0.851043
0.499419
Polynomial 2°c
Polynomial 3°
n/ab
61.790726
0.869761
0.5002
Power
n/ab
61.790726
0.870176
0.5002
aConstant variance case presented (Test 2 p-value = 0.2848). Although Test 1 p-value (0.091) was greater than
0.05, visual inspection of the dose-response curve indicates that responses do differ between dose groups.
Selected model in bold; scaled residuals for selected model for concentrations 0, 0.1339, and 0.8671 mg/L were
-0.0916, 0.108, and -0.0167 respectively.
b x2 test had insufficient degrees of freedom (due to estimated model parameters = dose groups). However,
inspection of scaled residuals and visual fit indicated appropriate model fit.
c For the polynomial 3° model, the b3 coefficient estimate was 0 (boundary). The models in this row reduced to
the polynomial 2° model.
Data Source: (Korsak et al.. 2000a)
Linear Model with 0.95 Confidence Level
1.5
Linear
11
10.5
10
9.5
9
8.5
8
BMD
7.5
BMDI
O
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
dose
	09:31 04/19 2012	
Note: BMR = 1 SD change from control mean; dose shown in mg/L 1,2,4-TMB (high dose dropped) (Korsak et al.. 2000a)
Figure C-2. Plot of mean response by dose for decreased red blood cells in male
Wistar rats, with the fitted curve for Linear model with constant variance.
This document is a draft for review purposes only and does not constitute Agency policy.
C-8	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1	Polynomial Model.
2	(Version: 2.16; Date: 05/26/2010)
3	The form of the response function is: Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2 + ... beta_n*doseAn
4	A constant variance model is fit.
5	Benchmark Dose Computations:
6	BMR = 1 estimated standard deviations from the control mean
7	BMD = 0.851043
8	BMDL at the 95% confidence level = 0.499419
9	Parameter Estimates
Variable
Model
(Default) Initial Parameter
Values
alpha
2.21157
2.45563
rho
0
0
beta 0
10.0231
10.0231
beta_l
-1.74743
-1.74743
10 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std
Scaled Resid
0
10
9.98
10
1.68
1.49
-0.0916
0.1339
10
9.84
9.79
1.82
1.49
0.108
0.8671
10
8.5
8.51
1.11
1.49
-0.0167
11 Likelihoods of Interest
Model
Log(likelihood)
# Pa rams
AIC
A1
-26.895363
4
61.790726
A2
-25.639495
6
63.278991
A3
-26.895363
4
61.790726
fitted
-26.905560
3
59.811121
R
-29.647442
2
63.294884
12 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1 (Does response and/or variances differ among Dose
levels, A2 vs. R)
8.01589
4
0.091
Test 2 (Are Variances Homogeneous, A2 vs. Al)
2.51173
2
0.2848
Test 3 (Are variances adequately modeled, A2 vs. A3)
2.51173
2
0.2848
Test 4 (Does the model for the Mean fit, A3 vs. fitted)
0.0203948
1
0.8864
This document is a draft for review purposes only and does not constitute Agency policy.
C-9	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-4. Summary of BMD modeling results for decreased clotting time in female
Wistar rats exposed to 1,2,4-TMB by inhalation for 3 months; BMR = 1 SD
change from control mean (constant and modeled variance, high dose
dropped) (Korsak et al.. 2000a)
Constant Variance
Model3
Goodness-of-fit
BMD1sd
(mg/L)
BMDL1Sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.0676
151.6841
0.624689
0.35101
No model selected as Test 2
p-value was < 0.10.
Therefore, as suggested in
the Benchmark Dose
Technical Guidance (U.S.
EPA, 2012a), the data were
remodeled using a non-
homogenous variance
model
Exponential (M4)
n/ac
150.3436
0.118085
0.0006662
Lineard
Polynomial 2°
Polynomial 3°
Power
0.05648
151.99019
0.69465
0.441274
Modeled Variance
Model®
Goodness-of-fit
BMD1Sd
(mg/L)
BMDL1Sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.00949
150.0056
0.829105
0.456483
No model selected as the
only appropriate fitting
model (Exponential model
4) returned an implausibly
low BMDL estimate.
Exponential (M4)f
n/ac
145.2775
0.154524
0.000850437
Lineard
Polynomial 2°
Polynomial 3°
Power
0.007771
150.362869
0.866447
0.533906
a Constant variance case presented (Test 2 p-value = 0.008489).
b For Exponential model 3, the estimate of d was 1 (boundary). The models in this row reduced to exponential
model 2.
CX2 test had insufficient degrees of freedom (due to estimated model parameters = dose groups).
d For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
6 Modeled variance case presented (Test 3 p-value = 0.1159).
Y test had insufficient degrees of freedom (due to estimated model parameters = dose groups). However,
inspection of scaled residuals and visual fit indicated appropriate model fit. However, this model returned an
unreasonably low BMDL value. Therefore, this endpoint cannot be modeled in BMDS and the NOAEL/LOAEL
approach is recommended.
Data Source: (Korsak et al., 2000a).
This document is a draft for review purposes only and does not constitute Agency policy.
C-10	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-5. Summary of BMD modeling results for decreased fetal weight in male
Sprague-Dawley rats exposed to 1,2,4-TMB by maternal inhalation on
GD6-GD20; BMR = 1 SD or 5% change from control mean (constant
variance)(Saillenfait et al., 2005)
BMR = 1 SD change from control mean
Model3
Goodness-of-fit
BMD1sd
(mg/m3)
BMDL1Sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)
0.5714
-84.27301
2,803.48
2,139.69
Of the models that provided
an adequate fit and valid
BMDL estimate, the Linear
model was selected based
on the lowest AIC (BMDLs
differed by less than 3-fold).
Exponential (M3)
0.8333
-83.91341
3,440.45
2,348.58
Exponential (M4)
0.5714
-84.27301
2,803.48
2,052.08
Exponential (M5)
0.5459
-81.91341
3,440.45
2,348.58
Hill
0.5588
-81.936294
3,440.86
2,367.37
Linear
0.6217
-84.509084
2,839.22
2,201.74
Polynomial 2°
0.8828
-84.028802
3,398.61
2,382.65
Polynomial 3°
0.9521
-84.179982
3,444.47
2,408.2
Power
0.8432
-83.937043
3,440.84
2,368.19
BMR = 5% change from control mean
Model3
Goodness-of-fit
bmd5%
(mg/m3)
BMDI_5%
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)
0.5714
-84.27301
2,009.49
1,577.44
Of the models that provided
an adequate fit and valid
BMDL estimate, the Linear
model was selected based
on the lowest AIC (BMDLs
differed by less than 3-fold).
Exponential (M3)
0.8333
-83.91341
2,861.09
1,716
Exponential (M4)
0.5714
-84.27301
2,009.49
1,427.9
Exponential (M5)
0.5459
-81.91341
2,861.09
1,716
Hill
0.5588
-81.936294
2,857.59
1,749.71
Linear
0.6217
-84.509084
2,057.05
1,640.07
Polynomial 2°
0.8828
-84.028802
2,798.98
1,760.54
Polynomial 3°
0.9521
-84.179982
2,841.49
1,777.39
Power
0.8432
-83.937043
2,857.43
1,750.98
aConstant variance case presented (Test 2 p-value = 0.1008), selected model in bold; scaled residuals for selected model for
concentrations 0, 492,1,471, 2,913, and 4,408 mg/mB were -0.336, -0.324, 0.486, 0.906,
-0.694, respectively.
Data source: (Saillenfait et al.. 2005)
This document is a draft for review purposes only and does not constitute Agency policy.
C-ll	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Linear Model with 0.95 Confidence Level
Linear
6
5.8
5.6
5.4
5.2
5
BMDI
BMD
0
500
1000
1500
2000
2500
3000
3500
4000
4500
dose
	10:32 04/19 2012	
Note: BMR = 1 SD change from control mean, dose shown in mg/m3 1,2,4-TMB (Saillenfait et al., 2005)
Figure C-3. Plot of mean response by dose for decreased fetal weight in male
Sprague-Dawley rats, with the fitted curve for Linear model with constant variance.
Linear Model with 0.95 Confidence Level
Linear
5.8
5.6
5.4
5.2
BMD
BMDI
0
500
1000
1500
2000
2500
3000
3500
4000
4500
dose
	10:35 04/19 2012	
Note: BMR = 5% change from control mean, dose shown in mg/m31,2,4-TMB (Saillenfait et al., 2005).
Figure C-4. Plot of mean response by dose for decreased fetal weight in male
Sprague-Dawley rats, with the fitted curve for Linear model with constant variance.
This document is a draft for review purposes only and does not constitute Agency policy.
C-12	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1	Polynomial Model.
2	(Version: 2.16; Date: 05/26/2010)
3	The form of the response function is: Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2 + ... beta_n*doseAn
4	A constant variance model is fit.
5	Benchmark Dose Computations:
6	BMR = 1 estimated standard deviations from the control mean
7	BMD = 2839.22
8	BMDL at the 95% confidence level = 2201.74
9	BMR = 5% Relative risk
10	BMD = 2057.05
11	BMDL at the 95% confidence level = 1640.07
12	Parameter Estimates
Variable
Model
(Default) Initial Parameter
Values
alpha
0.16139
0.170101
rho
0
0
beta 0
5.88846
5.88821
beta_l
-0.000143129
-0.000142292
13 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std
Scaled Resid
0
23
5.86
5.89
0.34
0.406
-0.336
492
22
5.79
5.82
0.3
0.406
-0.324
1471
22
5.72
5.63
0.49
0.406
0.486
2913
22
5.55
5.47
0.48
0.406
0.906
4408
24
5.2
5.26
0.42
0.406
-0.694
14 Likelihoods of Interest
Model
Log(likelihood)
# Pa rams
AIC
A1
46.139026
6
-80.278052
A2
50.018128
10
-80.036256
A3
46.139026
6
-80.278052
fitted
45.254542
3
-84.509084
R
28.974008
2
-53.948016
15 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1 (Does response and/or variances differ among Dose
levels, A2 vs. R)
42.0882
8
< 0.0001
Test 2 (Are Variances Homogeneous, A2 vs. Al)
7.7582
4
0.1008
Test 3 (Are variances adequately modeled, A2 vs. A3)
7.7582
4
0.1008
Test 4 (Does the model for the Mean fit, A3 vs. fitted)
1.76897
3
0.6217
This document is a draft for review purposes only and does not constitute Agency policy.
C-13	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-6. Summary of BMD modeling results for decreased fetal weight in female
Sprague-Dawley rats exposed to 1,2,4-TMB by maternal inhalation on
GD6-GD20; BMR = 1 SD or 5% change from control mean (constant
variance; Saillenfait etal., 2005)
BMR = 1 SD change from control mean
Model3
Goodness-of-fit
BMD1sd
(mg/m3)
BMDL1sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)
0.5056
-101.6488
2,650.97
2,044.51
Of the models that provided
an adequate fit and valid
BMDL estimate, the linear
model was selected based
on the lowest AIC (BMDLs
differed by less than 3-fold).
Exponential (M3)
0.654
-101.1358
3,312.88
2,212.4
Exponential (M4)
0.5056
-101.6488
2,650.97
1,947.94
Exponential (M5)
0.3568
-99.13583
3,312.88
2,212.4
Hill
0.3698
-99.180649
3,311.58
2,241.33
Linear
0.5547
-101.899075
2,692.29
2,108.65
Polynomial 2°
0.7252
-101.342513
3,258.79
2,264.38
Polynomial 3°
0.832
-101.617243
3,322.13
2,306.76
Power
0.6693
-101.182018
3,311.53
2,242.38
BMR = 5% change from control mean
Model3
Goodness-of-fit
bmd5%
(mg/m3)
BMDI_5%
(mg/m3)
Basis for Model Selection
p-value
p-value
AIC
Exponential (M2)
0.5056
-101.6488
1,951.39
1,549
Of the models that provided
an adequate fit and valid
BMDL estimate, the linear
model was selected based
on the lowest AIC (BMDLs
differed by less than 3-fold).
Exponential (M3)
0.654
-101.1358
2,778.64
1,662.76
Exponential (M4)
0.5056
-101.6488
1,951.39
1,398.32
Exponential (M5)
0.3568
-99.13583
2,778.64
1,662.76
Hill
0.3698
-99.180649
2,773.5
1,702.36
Linear
0.5547
-101.899075
2,001.36
1,612.89
Polynomial 2°
0.7252
-101.342513
2,703.42
1,718.54
Polynomial 3°
0.832
-101.617243
2,764.88
1,746.99
Power
0.6693
-101.182018
2,773.32
1,703.72
aConstant variance case presented (Test 2 p-value = 0.3936), selected model in bold; scaled residuals for
selected model for concentrations 0, 492, ,1471, 2,913, and 4,408 mg/m3 were 0.39, -0.187, -0.566, 0.519,
-0.158, respectively.
Data source: (Saillenfait et al.. 2005)
This document is a draft for review purposes only and does not constitute Agency policy.
C-14	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Linear Model with 0.95 Confidence Level
5.8
Linear
5.6
5.4
5.2
4.8
BMDI
BMD
0
500
1000
1500
2000
2500
3000
3500
4000
4500
dose
	10:09 04/19 2012	
Note: BMR = 1 SD change from control mean, dose shown in mg/m3 1,2,4-TMB (Saillenfait et al., 2005)
Figure C-5. Plot of mean response by dose for decreased fetal weight in female
Sprague-Dawley rats, with the fitted curve for Linear model with constant variance.
Linear Model with 0.95 Confidence Level
5.8
Linear
5.6
5.4
5.2
4.8
BMDL
BMD
0
500
1000
1500
2000
2500
3000
3500
4000
4500
dose
	10:15 04/19 2012	
Note: BMR = 5% change from control mean, dose shown in mg/m31,2,4-TMB (Saillenfait et al., 2005)
Figure C-6. Plot of mean response by dose for decreased fetal weight in female
Sprague-Dawley rats, with the fitted curve for Linear model with constant variance.
This document is a draft for review purposes only and does not constitute Agency policy.
C-15	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1	Polynomial Model.
2	(Version: 2.16; Date: 05/26/2010)
3	The form of the response function is: Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2 + ... beta_n*doseAn
4	A constant variance model is fit.
5	Benchmark Dose Computations:
6	BMR = 1 estimated standard deviations from the control mean
7	BMD = 2692.29
8	BMDL at the 95% confidence level = 2108.65
9	BMR = 5% Relative risk
10	BMD = 2001.36
11	BMDL at the 95% confidence level = 1612.89
12	Parameter Estimates
Variable
Model
(Default) Initial Parameter
Values
alpha
0.141584
0.14543
rho
0
0
beta 0
5.59423
5.59388
beta_l
-0.000139761
-0.000138886
13 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std
Scaled Resid
0
23
5.57
5.59
0.33
0.376
-0.309
492
22
5.51
5.53
0.31
0.376
-0.193
1471
22
5.4
5.39
0.45
0.376
0.142
2913
22
5.28
5.19
0.4
0.376
1.16
4408
24
4.92
4.98
0.4
0.376
-0.757
14 Likelihoods of Interest
Model
Log(likelihood)
# Pa rams
AIC
A1
54.992554
6
-97.985109
A2
57.038880
10
-94.077760
A3
54.992554
6
-97.985109
fitted
53.949538
3
-101.899075
R
36.104870
2
-68.209740
15 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1 (Does response and/or variances differ among Dose
levels, A2 vs. R)
41.868
8
< 0.001
Test 2 (Are Variances Homogeneous, A2 vs. Al)
4.09265
4
0.3936
Test 3 (Are variances adequately modeled, A2 vs. A3)
4.09265
4
0.3936
Test 4 (Does the model for the Mean fit, A3 vs. fitted)
2.08603
3
0.5547
This document is a draft for review purposes only and does not constitute Agency policy.
C-16	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-7. Summary of BMD modeling results for decreased maternal body weight
gain in female Sprague-Dawley rats exposed to 1,2,4-TMB by inhalation on
GD6-GD20; BMR = 1 SD change from control mean (constant variance)
(Saillenfait etal., 2005)
Model3
Goodness-of-fit
BMD1sd
(mg/m3)
BMDL1Sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
< 0.0001
1,025.385
3.67497
Bad Completion
Of the models that provided
an adequate fit and valid
BMDL estimate, the
Exponential 3 model was
selected based on the
lowest AIC (BMDLs differed
by less than 3-fold).
Exponential (M3)
0.7552
717.3518
2,821.1
2,247.99
Exponential (M4)b
< 0.0001
773.2296
Not Computed
0
Exponential (M5)
0.4537
719.3518
2,821.1
2,247.99
Hill
0.593
719.075964
2,781.23
2,161.92
Linear
0.1319
720.406291
2,009.47
1,649.63
Polynomial 2°c
Polynomial 3°
0.7004
717.502596
2,888.45
2,132.32
Power
0.7393
717.394507
2,821.04
2,129.53
aConstant variance case presented (Test 2 p-value = 0.4284). Selected model in bold; scaled residuals for
selected model for concentrations 0, 492,1,471, 2,913, and 4,408 mg/m3 were -0.1845,0.5186, -0.4013,
0.1315, -0.2808, respectively.
bThe Exponential models 2 and 4 models did not return BMD and/or BMDL values and were excluded from
further consideration.
c For the polynomial 3° model, the b3 coefficient estimate was 0 (boundary). The models in this row reduced to
the polynomial 2° model.
Data source: (Saillenfait et al., 2005).
Exponential Model 3 with 0.95 Confidence Level
20
10
Exponential
BMDL
BMD
500 1000 1500 2000 2500 3000 3500 4000 4500
11:00 04/19 2012
Note: BMR = 1 SD change from control mean; dose shown in mg/m 1,2,4-TMB (Saillenfait et al.. 2005).
Figure C-7. Plot of mean response by dose for decreased maternal body weight gain
in female Sprague-Dawley rats, with fitted curve for Exponential model 3 with
constant variance.
This document is a draft for review purposes only and does not constitute Agency policy.
C-17	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1	Exponential Model.
2	(Version: 1.7; Date: 12/10/2009)
3	The form of the response function is: Model 3: Y[dose] = a * exp{sign * (b * dose)Ad}
4	A constant variance model is fit.
5	Benchmark Dose Computations:
6	BMR = 1 estimated standard deviations from the control mean
7	BMD = 2821.1
8	BMDL at the 95% confidence level = 2247.99
9	Parameter Estimates
Variable
Model
(Default) Initial Parameter
Values
Inalpha
5.22238
5.21746
rho
0
0
a
29.5127
0
b
0.000314053
0.000203897
c
0
0
d
3.96638
18
10 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std
Scaled Resid
0
24
29
29.51
12
13.62
-0.1845
492
22
31
29.49
14
13.62
0.5186
1471
22
27
28.16
12
13.62
-0.4013
2913
22
15
14.62
17
13.62
0.1315
4408
24
0
0.7804
14
13.62
-0.2808
11 Likelihoods of Interest
Model
Log(likelihood)
# Pa rams
AIC
A1
-354.3952
6
720.7904
A2
-352.4764
10
724.9529
A3
-354.3952
6
720.7904
R
-386.383
2
776.7661
4
-354.6759
4
717.3518
12 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1 (Does response and/or variances differ among Dose
levels, A2 vs. R)
67.81
8
< 0.0001
Test 2 (Are Variances Homogeneous, A2 vs. Al)
3.837
4
0.4284
Test 3 (Are variances adequately modeled, A2 vs. A3)
3.837
4
0.4284
Test 4 (Does the model for the Mean fit, A3 vs. fitted)
0.5615
2
0.7552
This document is a draft for review purposes only and does not constitute Agency policy.
C-18	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-8. Summary of BMD modeling results for increased latency to paw-lick in
male Wistar rats exposed to 1,2,3-TMB by inhalation for 3 months;
BMR = 1 SD change from control mean (constant variance and modeled
variance), (Korsakand Rydzynski, 1996)
Constant Variance
Model3
Goodness-of-fit
BMD1sd
(mg/L)
BMDL1Sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.005704
262.2082
700.938
566.333
No model selected as Test 2 p-
value was < 0.10. Therefore, as
suggested in the Benchmark
Dose Technical Guidance (U.S.
Exponential (M4)
0.5461
254.2393
192.288
107.132
Exponential (M5)
n/ac
255.8749
201.187
111.315
Hill
n/ac
255.874906
185.863
110.398
Lineard
Polynomial 2°
Polynomial 3°
Power
0.01728
259.991214
577.555
442.59
remodeled using a non-
homogenous variance model
Modeled Variance
Model®
Goodness-of-fit
d iv/in
BMDL1sd
(mg/L)
Basis for Model Selection
p-value
AIC
DIVIUisd
(mg/L)
Exponential (M2)b
Exponential (M3)
<0.0001
259.5324
496.844
329.318
No model selected as Test 3
Exponential (M4)
0.301
241.4193
86.2091
46.7265
p-value was < 0.1. This was
Exponential (M5)
n/ac
242.5858
113.028
51.9836
dose group. Therefore, the
Hill
n/ac
265.438765
334.7333
Not calculated
data were remodeled using a
Linear'
Polynomial 2°
Power
0.0003247
254.414778
319.651
195.989
non-homogenous variance
model and with the high dose
dropped (see Table C-9)
a Constant variance case presented (Test 2 p-value = 0.0.0001146).
b For Exponential model 3, the estimate of d was 1 (boundary). The models in this row reduced to exponential
model 2.
CX2 test had insufficient degrees of freedom (due to estimated model parameters = dose groups).
d For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
6 Modeled variance case presented (Test 3 p-value = 0.07076). This p-value indicates that a modeled variance
model does not adequately describe the observed variances.
f For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° model, the b2
coefficient estimate was 0 (boundary). The polynomial 3° did not converge. The models in this row reduced to
the Linear model.
Data Source: (Korsakand Rydzynski. 1996)
This document is a draft for review purposes only and does not constitute Agency policy.
C-19	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-9. Summary of BMD modeling results for increased latency to paw-lick in
male Wistar rats exposed to 1,2,3-TMB by inhalation for 3 months;
BMR = 1 SD change from control meanfmodeled variance, high dose
dropped), (Korsakand Rydzynski, 1996)
Model3
Goodness-of-fit
BMD1sd
(mg/m3)
BMDL1sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.07449
203.2651
192.144
131.627
Of the models that
provided an adequate fit
and valid BMDL estimate,
the linear model was
selected based on the
lowest AIC (BMDLs differed
by less than 3-fold).
Exponential (M4)
n/ac
202.0839
104.546
52.5736
Lineard
Polynomial 2°
Polynomial 3°
Power
0.2016
201.714812
152.065
97.1911
aModeled variance case presented (Test 3 p-value = 0.5008). Selected model in bold; scaled residuals for
selected model for concentrations 0,123, and 492 mg/m3 were -0.102, 0.319, and -0.354, respectively.
b For Exponential model 3, the estimate of d was 1 (boundary). The models in this row reduced to exponential
model 2.
test had insufficient degrees of freedom (due to estimated model parameters = dose groups). However,
inspection of scaled residuals and visual fit indicated appropriate model fit.
d For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
Data Source: (Korsak and Rydzynski, 1996)
Linear Model with 0.95 Confidence Level
13:05 03/29 2012
Note: BMR = 1 SD change from control mean; dose shown in mg/m 1,2,3-TMB (high dose dropped) (Korsakand Rydzynski.
1996)
Figure C-8. Plot of mean response by dose for increased latency to paw-lick in male
Wistar rats, with fitted curve for Linear model with modeled variance.
This document is a draft for review purposes only and does not constitute Agency policy.
C-20	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1	Polynomial Model.
2	(Version: 2.16; Date: 05/26/2010)
3	The form of the response function is: Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2 + ... beta_n*doseAn
4	The variance is to be modeled as Var(i] = expflalpha + log(mean(i]] * rho]
5	Benchmark Dose Computations:
6	BMR = 1 estimated standard deviations from the control mean
7	BMD = 152.065
8	BMDL at the 95% confidence level = 97.1911
9	Parameter Estimates
Variable
Model
(Default) Initial Parameter
Values
alpha
-7.3421
2.58956
rho
3.94293
0
beta 0
9.74214
9.90769
beta_l
0.0148851
0.0131332
10 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std
Scaled Resid
0
30
9.7
9.74
2.1
2.26
-0.102
123
20
11.8
11.6
3.8
3.18
0.319
492
10
16.3
17.1
6.3
6.84
-0.354
11 Likelihoods of Interest
Model
Log(likelihood)
# Pa rams
AIC
A1
-106.147893
4
220.295786
A2
-95.815379
6
203.630758
A3
-96.041973
5
202.083946
fitted
-96.857406
4
201.714812
R
-116.956260
2
237.912520
12 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1 (Does response and/or variances differ among Dose
levels, A2 vs. R)
42.2818
4
<0.0001
Test 2 (Are Variances Homogeneous, A2 vs. Al)
20.665
2
<0.0001
Test 3 (Are variances adequately modeled, A2 vs. A3)
0.453187
1
0.5008
Test 4 (Does the model for the Mean fit, A3 vs. fitted)
1.63087
1
0.2016
13
This document is a draft for review purposes only and does not constitute Agency policy.
C-21	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-10. Summary of BMD modeling results for decreased segmented neutrophils
in male Wistar rats exposed to 1,2,3-TMB by inhalation for 3 months;
BMR = 1 SD change from control mean (constant variance), (Korsak et al.,
2000b)
Model3
Goodness-of-fit
BMDisd
(mg/m3)
BMDL1Sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.7155
189.1052
915.77
534.809
Of the models that
provided an adequate fit
and valid BMDL estimate,
the Exponential 2 model
was selected based on the
lowest AIC (BMDLs differed
by less than 3-fold).
Exponential (M4)
0.4482
191.0108
814.879
261.734
Exponential (M5)
n/ac
192.4867
547.805
137.551
Hill
n/ac
192.486705
564.348
Not calculated
Lineard
Polynomial 2°
Polynomial 3°
Power
0.6711
189.233222
979.089
632.777
aConstant variance case presented (Test 2 p-value = 0.2692). Selected model in bold; scaled residuals for
selected model for concentrations 0,123, 492 and 1,230 mg/m3 were -0.242, 0.5701, -0.4994, and 0.176,
respectively.
b For Exponential model 3, the estimate of d was 1 (boundary). The models in this row reduced to exponential
model 2.
test had insufficient degrees of freedom (due to estimated model parameters = dose groups). Inspection of
scaled residuals indicated appropriate model fit. However, inspection of visual fit indicated uncertain dose-
response characteristics, and therefore, these models were excluded from consideration.
d For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
Data Source: (Korsak et al., 2000b)
Exponential Model 2 with 0.95 Confidence Level

Exponential 	




BMDL

BMD

O
200 400 600
800
1000
1200
3 2012
dose



Note: BMR = 1 SD change from control mean; dose shown in mg/mB 1,2,3-TMB (Korsak et al.. 2000b)
Figure C-9. Plot of mean response by dose for decreased segmented neutrophils in
male Wistar rats, with fitted curve for Exponential model 2 with constant variance.
This document is a draft for review purposes only and does not constitute Agency policy.
C-22	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1	Exponential Model
2	(Version: 1.7; Date: 12/10/2009)
3	The form of the response function is: Model 2: Y[dose] = a * exp{sign * b * dose}
4	A constant variance model is fit.
5	Benchmark Dose Computations:
6	BMR = 1 estimated standard deviations from the control mean
7	BMD = 915.77
8	BMDL at the 95% confidence level = 534.809
9	Parameter Estimates
Variable
Model
(Default) Initial Parameter
Values
Inalpha
3.57763
3.56089
rho
0
0
a
25.2579
19.0843
b
0.000295164
0.00028845
c
0
0
d
1
1
10 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std
Scaled Resid
0
10
24.8
25.26
4.5
5.982
-0.242
128
10
25.4
24.32
5.8
5.982
0.5701
523
10
20.7
21.64
5.8
5.982
-0.4994
1269
10
17.7
17.37
8.3
5.982
0.176
11 Likelihoods of Interest
Model
Log(likelihood)
# Pa rams
AIC
A1
-91.2178
5
192.4356
A2
-89.25328
8
194.5066
A3
-91.2178
5
192.4356
R
-96.16301
2
196.326
4
-91.55261
3
189.1052
12 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1 (Does response and/or variances differ among Dose
levels, A2 vs. R)
13.82
6
0.03172
Test 2 (Are Variances Homogeneous, A2 vs. Al)
3.929
3
0.2692
Test 3 (Are variances adequately modeled, A2 vs. A3)
3.929
3
0.2692
Test 4 (Does the model for the Mean fit, A3 vs. fitted)
0.6696
2
0.7155
This document is a draft for review purposes only and does not constitute Agency policy.
C-23	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-ll. Summary of BMD modeling results for decreased segmented neutrophils
in female Wistar rats exposed to 1,2,3-TMB by inhalation for 3 months;
BMR = 1 SD change from control mean (constant variance), (Korsak et al.,
2000b)
Model3
Goodness-of-fit
BMD1Sd
(mg/m3)
BMDL1sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.6401
177.6514
517.048
334.805
Of the models that
provided an adequate fit
and valid BMDL estimate,
the Hill model was selected
based on the lowest BMDL
(BMDLs differed by more
than 3-fold).
Exponential (M4)b
Exponential (M5)
0.5208
179.1714
365.397
134.354
Hill
0.5692
179.083138
337.442
99.2111
Linear0
Polynomial 2°
Polynomial 3°
Power
0.4533
178.341743
645.521
465.309
a Constant variance case presented (Test 2 p-value = 0.09252). Although this p-value is less than 0.10, it
indicates a marginal fit at the 95% confidence level, and therefore a constant variance is determined to
adequately fit the observed variance data. Selected model in bold; scaled residuals for selected model for
concentrations 0,128, 523, and 1,269 mg/m3 were 0.209, -0.412, 0.312, and -0.108, respectively.
b For Exponential models 3 and 5, the estimate of d was 1 (boundary). Therefore Exponential model 3 reduced
to Exponential model 2, and Exponential model 5 reduced to Exponential model 4.
c For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
Data Source: (Korsak et al., 2000b)
Hill Model with 0.95 Confidence Level
09:03 04/19 2012
600
dose
Note: BMR = 1 SD change from control mean; dose shown in mg/m 1,2,3-TMB (Korsak et al.. 2000b).
Figure C-10. Plot of mean response by dose for decreased segmented neutrophils in
female Wistar rats, with fitted curve for Hill model with constant variance.
This document is a draft for review purposes only and does not constitute Agency policy.
C-24	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1	Hill Model.
2	(Version: 2.16; Date: 04/06/2011)
3	The form of the response function is: Y[dose] = intercept + v*doseAn/ (kAn + doseAn)
4	A constant variance model is fit
5	Benchmark Dose Computations:
6	BMR = 1 estimated standard deviations from the control mean
7	BMD = 337.442
8	BMDL at the 95% confidence level = 99.2111
9	Parameter Estimates
Variable
Model
(Default) Initial Parameter
Values
alpha
26.4982
29.205
rho
0
0
intercept
22.76
23.1
V
-17.5024
-11.2
N
1
1.05772
k
809.89
391.333
10 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std
Scaled Resid
0
10
23.1
22.8
6.1
5.15
0.209
128
10
19.7
20.4
3.4
5.15
-0.412
523
10
16.4
15.9
4.2
5.15
0.312
1269
10
11.9
12.1
7.1
5.15
-0.108
11 Likelihoods of Interest
Model
Log(likelihood)
# Pa rams
AIC
A1
-85.379588
5
180.759176
A2
-82.165225
8
180.330450
A3
-85.379588
5
180.759176
fitted
-85.541569
4
179.083138
R
-95.409822
2
194.819645
12 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1 (Does response and/or variances differ among Dose
levels, A2 vs. R)
26.4892
6
0.0001804
Test 2 (Are Variances Homogeneous, A2 vs. Al)
6.42873
3
0.09252
Test 3 (Are variances adequately modeled, A2 vs. A3)
6.42873
3
0.09252
Test 4 (Does the model for the Mean fit, A3 vs. fitted)
0.323962
1
0.5692
This document is a draft for review purposes only and does not constitute Agency policy.
C-25	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-12. Summary of BMD modeling results for increased reticulocytes in male
Wistar rats exposed to 1,2,3-TMB by inhalation for 3 months; BMR = 1 SD
change from control mean (constant variance), (Korsak et al., 2000b)
Model3
Goodness-of-fit
BMD1Sd
(mg/m3)
BMDL1Sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.2733
89.08418
1112.25
806.744
Of the models that
provided an adequate fit
and valid BMDL estimate,
the Linear model was
selected based on the
lowest AIC (BMDLs differed
by less than 3-fold).
Exponential (M4)
0.1397
90.67033
900.404
308.017
Exponential (M5)
n/ac
91.37006
540.186
140.925
Hill
n/ac
91.370061
554.848
Not calculated
Lineard
Polynomial 2°
Polynomial 3°
Power
0.3105
88.828645
1025.1
652.898
a Constant variance case presented (Test 2 p-value = 0.5223). Selected model in bold; scaled residuals for
selected model for concentrations 0,128, 523 and 1,269 mg/m3 were 0.555, -1.14,0.793, and -0.212,
respectively.
b For Exponential model 3, the estimate of d was 1 (boundary). The models in this row reduced to exponential
model 2.
test had insufficient degrees of freedom (due to estimated model parameters = dose groups). Inspection of
scaled residuals indicated appropriate model fit. However, inspection of visual fit indicated uncertain dose-
response characteristics, and therefore, these models were excluded from consideration.
d For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
Data Source: (Korsak et al., 2000b).
Linear Model with 0.95 Confidence Level
O	200	400	600	800	IOOO	1 200
dose
09:52 04/19 2012
Note: BMR = 1 SD change from control mean; dose shown in mg/mB 1,2,3-TMB (Korsak et al.. 2000b).
Figure C-ll. Plot of mean response by dose for increased reticulocytes in male Wistar
rats, with fitted curve for Linear model with constant variance.
This document is a draft for review purposes only and does not constitute Agency policy.
C-26	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1	Polynomial Model.
2	(Version: 2.16; Date: 05/26/2010)
3	The form of the response function is: Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2 + ... beta_n*doseAn
4	A constant variance model is fit
5	Benchmark Dose Computations:
6	BMR = 1 estimated standard deviations from the control mean
7	BMD = 1025.1
8	BMDL at the 95% confidence level = 652.989
9	Parameter Estimates
Variable
Model
(Default) Initial Parameter
Values
alpha
2.91747
3.0575
rho
0
0
beta 0
2.50021
2.50021
beta_l
0.0016623
0.00166623
10 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std
Scaled Resid
0
10
2.8
2.5
1.3
1.71
0.555
128
10
2.1
2.71
1.7
1.71
-1.14
523
10
3.8
3.37
2.1
1.71
0.793
1269
10
4.5
4.61
1.8
1.71
-0.212
11 Likelihoods of Interest
Model
Log(likelihood)
# Pa rams
AIC
A1
-40.244741
5
90.489483
A2
-39.119955
8
94.239910
A3
-40.244741
5
90.489483
fitted
-41.414322
3
88.828645
R
-45.600613
2
95.201226
12 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1 (Does response and/or variances differ among Dose
levels, A2 vs. R)
12.9613
6
0.04365
Test 2 (Are Variances Homogeneous, A2 vs. Al)
2.24957
3
0.5223
Test 3 (Are variances adequately modeled, A2 vs. A3)
2.24957
3
0.5223
Test 4 (Does the model for the Mean fit, A3 vs. fitted)
2.33916
2
0.3105
This document is a draft for review purposes only and does not constitute Agency policy.
C-27	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-13. Summary of BMD modeling results for decreased fetal weight in male
Sprague-Dawley rats exposed to 1,3,5-TMB by maternal inhalation on
GD6-GD20; BMR = 1 SD change from control mean (constant and modeled
variance)(Saillenfait et al., 2005)
Constant Variance
Model3
Goodness-of-fit
BMD1Sd
(mg/m3)
BMDL1Sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.6927
-66.94125
3,396.62
2,560.01
No model selected as Test 2
p-value was < 0.10.
Therefore, as suggested in
the Benchmark Dose
Technical Guidance (U.S.
EPA, 2012a), the data were
Exponential (M4)
0.6981
-65.6776
2,604.81
1,341.07
Exponential (M5)
0.397
-63.67902
2,603.37
1,341.3
Hill
0.4094
-63.715888
2,572.4
1,274.69
Linear0
Polynomial 2°
Polynomial 3°
Power
0.6496
-66.753074
3,513.03
2,694.51
remodeled using a non-
homogenous variance
model
Modeled Variance
Modeld
Goodness-of-fit
BMD1Sd
(mg/m3)
BMDL1sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.5214
-73.29149
2,523.27
1,779.29
No model selected as Test 3
p-value was < 0.1. This was
due to high variance in
control group. Therefore,
this endpoint cannot be
modeled in BMDS and the
NOAEL/LOAEL approach is
recommended.
Exponential (M4)
0.4304
-71.85947
2,041.7
1,125.34
Exponential (M5)
0.3877
-70.79949
2,044.66
1,237.6
Hill
0.4276
-65.644335
2,407.38
1,295.43
Linear0
Polynomial 2°
Polynomial 3°
Power
0.4791
-73.066751
2,636.36
1,890.46
a Constant variance case presented (Test 2 p-value = 0.002368)
b For Exponential model 3, the estimate of d was 1 (boundary). The models in this row reduced to exponential
model 2.
c For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
d Modeled variance case presented (Test 3 p-value = 0.06027, except the Hill model, for which Test 3 p-value =
0.00544).
Data source: (Saillenfait et al., 2005).
This document is a draft for review purposes only and does not constitute Agency policy.
C-28	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-14. Summary of BMD modeling results for decreased fetal weight in female
Sprague-Dawley rats exposed to 1,3,5-TMB by maternal inhalation on
GD6-GD20; BMR = 1 SD change from control mean (constant and modeled
variance)(Saillenfait et al., 2005)
Constant Variance
Model3
Goodness-of-fit
BMD1Sd
(mg/m3)
BMDL1Sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.9112
-61.96218
3,581.71
2,669
No model selected as Test 2
p-value was < 0.10.
Therefore, as suggested in
the Benchmark Dose
Technical Guidance (U.S.
EPA, 2012a), the data were
remodeled using a non-
homogenous variance
model
Exponential (M4)b
Exponential (M5)
0.7655
-59.96227
3,573.06
1,915.99
Hill
0.7656
-59.962704
3,569.61
1,865.62
Linear0
Polynomial 2°
Polynomial 3°
Power
0.9085
-61.950195
3,676.95
2,794.36
Modeled Variance
Modeld
Goodness-of-fit
BMD1Sd
(mg/m3)
BMDL1sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.01931
-67.53742
2692.79
1827.72
No model selected as Test 3
p-value was < 0.1. This was
due to high variance in
control group and low
variance in the high dose
group. Therefore, this
endpoint cannot be
modeled in BMDS and the
NOAEL/LOAEL approach is
recommended.
Exponential (M4)
0.05097
-69.49883
1481.66
798.275
Exponential (M5)
0.5334
-73.06401
1469.46
1069.57
Hill
0.4769
-59.505126
3161.1
1614.44
Linear6
Polynomial 2°
Polynomial 3°
Power
0.0148
0.01552
-67.061071
2841.13
1969.76
a Constant variance case presented (Test 2 p-value < 0.0001)
bFor Exponential models 3 and 5, the estimate of d was 1 (boundary). Therefore Exponential model 3 reduced
to Exponential model 2, and Exponential model 5 reduced to Exponential model 4.
c For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
d Modeled variance case presented (Test 3 p-value = 0.01301)
6 For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
The Test 4 p-value for the power model (0.01552) was different from the Test 4 p-value for the linear or
polynomial models (0.0148)
Data source: (Saillenfait et al., 2005).
This document is a draft for review purposes only and does not constitute Agency policy.
C-29	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-15. Summary of BMD modeling results for decreased maternal body weight
gain in female Sprague-Dawley rats exposed to 1,3,5-TMB by inhalation on
GD6-GD20; BMR = 1 SD change from control mean (constant and modeled
variance), (Saillenfait et al., 2005)
Constant Variance
Model3
Goodness-of-fit
BMD1Sd
(mg/m3)
BMDLlsd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)
< 0.0001
805.8321
3.36 x 10"51
Bad Completion
No model selected as Test 2
p-value was < 0.10.
Therefore, as suggested in
the Benchmark Dose
Technical Guidance (U.S.
EPA, 2012a), the data were
remodeled using a non-
homogenous variance
model
Exponential (M3)
< 0.0001
807.8353
6.29281
Bad Completion
Exponential (M4)
< 0.0001
701.8275
Not Computed
0
Exponential (M5)
0.00262
649.4267
2,057.15
1,396.23
Hill
0.5141
639.963339
2,035.36
1,353.4
Linearb
Polynomial 2°
Polynomial 3°
0.6919
636.99599
1,982.21
1,655.52
Power
0.4835
638.991033
2,014.88
1,655.77
Modeled Variance
Modelc
Goodness-of-fit
BMD1sd
(mg/m3)
BMDL1sd
(mg/m3)
Basis for Model Selection
p-value
AIC
Exponential (M2)d
< 0.0001
921.089
Not Computed
0
Only the power model
provided an adequate fit
and calculated a BMD and
BMDL, and therefore was
selected.
Exponential (M3)d
< 0.0001
923.089
Not Computed
0
Exponential (M4)
< 0.0001
698.0766
3.76 x 10"46
3.76 x 10"46
Exponential (M5)
< 0.0001
650.9354
1,476.12
601.777
Hill
<.0001
728.727708
29.7037
11.8372
Linear
0.0003338
645.262934
2,749.72
2,330.78
Polynomial 2°
<.0001
710.199993
-9,999
2,491.63
Polynomial 3°'d
0.2014
631.886974
1,797.1
Not calculated
Power
0.1981
631.236865
1,826.86
1,302.02
a Constant variance case presented (Test 2 p-value = 0.003114)
b For the polynomial 2° and 3° models, the b2 and b3 coefficient estimates were 0 (boundary). The models in this row
reduced to the Linear model.
c Modeled variance case presented (Test 3 p-value = 0.2221). Selected model in bold; scaled residuals for selected model for
concentrations 0, 497,1,471, 2,974, 5,874 mg/mB were -0.442, 0.983, -0.47, -0.776, 0.0673, respectively.
d The Exponential model 2 and model 3, as well as the polynomial 3° models, did not return BMD and/or BMDL values and
were excluded from further consideration.
Data source: (Saillenfait et al.. 2005).
This document is a draft for review purposes only and does not constitute Agency policy.
C-30	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Power Model with 0.95 Confidence Level
40
Power
30
20
10
0
-10
-20
BMDL
BMD
0
1000
2000
3000
4000
5000
6000
dose
	11:14 04/19 2012	
Note: BMR = 1 SD change from control mean; dose shown in mg/m31,3,5-TMB (Saillenfait et al., 2005)
Figure C-12. Plot of mean response by dose for decreased maternal body weight gain
in female Sprague-Dawley rats, with fitted curve for Power model with modeled
variance.
1	Power Model.
2	(Version: 2.16; Date: 10/28/2009)
3	The form of the response function is: Y[dose] = control + slope * doseApower
4	The variance is to be modeled as Var(i] = expflalpha + log(mean(i]] * rho]
5	Benchmark Dose Computations:
6	BMR = 1 estimated standard deviations from the control mean
7	BMD = 1826.86
8	BMDL at the 95% confidence level = 1302.02
9	Parameter Estimates
Variable
Model
(Default) Initial Parameter
Values
lalpha
8.3667
5.41079
rho
-1.04093
0
control
30.0752
-12
slope
-0.00209481
628.225
power
1.14244
-0.427017
This document is a draft for review purposes only and does not constitute Agency policy.
C-31	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std
Scaled Resid
0
21
29
30.1
14
11.2
-0.442
497
22
30
27.6
9
11.7
0.983
1471
21
20
21.4
12
13.3
-0.47
2974
17
7
10.6
20
19.2
-0.776
5874
18
-12
-12.3
19
17.8
0.0673
2 Likelihoods of Interest
Model
Log(likelihood)
# Pa rams
AIC
A1
-314.768805
6
641.537610
A2
-306.803486
10
633.606972
A3
-308.999390
7
631.998779
fitted
-310.618432
5
631.236865
R
-352.099997
2
708.199993
3 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1 (Does response and/or variances differ among Dose
levels, A2 vs. R)
90.593
8
<.0001
Test 2 (Are Variances Homogeneous, A2 vs. Al)
15.9306
4
0.003114
Test 3 (Are variances adequately modeled, A2 vs. A3)
4.39181
3
0.2221
Test 4 (Does the model for the Mean fit, A3 vs. fitted)
3.23809
2
0.1981
This document is a draft for review purposes only and does not constitute Agency policy.
C-32	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
C.2. BENCHMARK DOSE MODELING SUMMARY - ALTERNATIVE ANALYSIS WITH
HIGH DOSES INCLUDED
1	The modeling summaries included in this section are for comparison purposes only. After
2	calculation of internal blood dose metrics using the animal PBPK model, the high doses were
3	not dropped in these modeling analyses, even though the PBPK demonstrates poor model fit at
4	high doses. These modeling results were not used in any RfC derivations in Volume 1 of the
5	Toxicological Review.
Table C-16. Summary of BMD modeling results for increased latency to paw-lick in
male Wistar rats exposed to 1,2,4-TMB by inhalation for 3 months;
BMR = 1 SD change from control meanfconstant and modeled variance),
(Korsak and Rydzynski, 1996)
Constant Variance
Model3
Goodness-of-fit
BMD1sd
(mg/L)
BMDL1Sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.00061
190.1611
3.62226
2.73586
No model selected as Test
2 p-value was < 0.10.
Therefore, as suggested in
the Benchmark Dose
Technical Guidance (U.S.
EPA, 2012a), the data were
remodeled using a non-
homogenous variance
model
Exponential (M4)
0.8239
177.4066
0.242222
0.104385
Exponential (M5)
n/ac
179.3571
0.268238
0.105201
Hill
n/ac
179.357065
0.237108
0.0889465
Lineard
Polynomial 2°
Polynomial 3°
Power
0.0009125
189.355645
3.15451
2.22737
Modeled Variance
Model®
Goodness-of-fit
BMD1Sd
(mg/L)
BMDL1Sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.000633
191.8156
3.38239
2.34048
No model selected as Test
3 p-value was < 0.10.
Therefore, this endpoint
cannot be modeled in
BMDS and the
NOAEL/LOAEL approach is
recommended.
Exponential (M4)
0.8604
179.1164
0.231414
0.09854
Exponential (M5)
n/ac
181.0855
0.252014
0.0990336
Hill
n/ac
181.982905
0.292816
Not calculated
Lineard
Polynomial 2°
Polynomial 3°
Power
0.001014
190.872265
2.8175
1.72529
a Constant variance case presented (Test 2 p-value = 0.07651).
b For Exponential model 3, the estimate of d was 1 (boundary). The models in this row reduced to exponential model 2.
CX2 test had insufficient degrees of freedom (due to estimated model parameters = dose groups). Inspection of scaled
residuals and visual fit indicated appropriate model fit.
d For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models, the b2 and
b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
0 Modeled variance case presented (Test 3 p-value = 0.0371)
Data source: (Korsak and Rydzynski. 1996).
This document is a draft for review purposes only and does not constitute Agency policy.
C-33	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-17. Summary of BMD modeling results for decreased red blood cells in male
Wistar rats exposed to 1,2,4-TMB by inhalation for 3 months; BMR = 1 SD
change from control mean (constant variance), (Korsak et al., 2000a)
Model3
Goodness-of-fit
BMD1sd
(mg/L)
BMDL1sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.1671
78.98918
3.68518
2.30432
Of the models that
provided an adequate fit
and a valid BMDL
estimate, the Hill model
was selected based on
lowest BMDL (BMDLs
differed by greater than 3-
fold)
Exponential (M4)
0.7345
77.52579
0.795033
0.241565
Exponential (M5)
n/ac
79.41075
0.842867
0.249166
Hill
n/ac
79.410749
0.835638
0.212686
Lineard
Polynomial 2°
Polynomial 3°
Power
0.1498
79.207001
3.91553
2.5963
aConstant variance case presented (Test 2 p-value = 0.4329). Selected model in bold; scaled residuals for selected model for
concentrations 0, 0.1339, 0.8671, 5.248 mg/L were -1.93 xlO"08,1.75x 10"08, 4.83 x 10"osand -6.99 x 10"os, respectively.
b For Exponential model 3, the estimate of d was 1 (boundary). The models in this row reduced to exponential model 2.
ZX test had insufficient degrees of freedom (due to estimated model parameters = dose groups). Inspection of scaled
residuals and visual fit indicated appropriate model fit.
d For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models, the b2 and
b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
Data source: (Korsak et al.. 2000a).
Hill Model with 0.95 Confidence Level
11
10
9
8
7
BMDL
BMD
O
1
2
3
4
5
dose
	Q9:24 Q4/19 2Q12	
Note: BMR = 1 SD change from control mean; dose shown in mg/L 1,2,4-TMB (Korsak et al.. 2000a)
Figure C-13. Plot of mean response by dose for decreased red blood cells in male
Wistar rats, with fitted curve for Hill model with constant variance.
This document is a draft for review purposes only and does not constitute Agency policy.
C-34	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
1	Hill Model.
2	(Version: 2.16; Date: 04/06/2011)
3	The form of the response function is: Y[dose] = intercept + v*doseAn/ (kAn + doseAn)
4	A constant variance model is fit
5	Benchmark Dose Computations:
6	BMR = 1 estimated standard deviations from the control mean
7	BMD = 0.835638
8	BMDL at the 95% confidence level = 0.212686
9	Parameter Estimates


(Default) Initial Parameter
Variable
Model
Values
alpha
2.08604
2.31783
rho
0
0
intercept
9.98
9.98
V
-2.33466
-2.28
N
1.7672
2.11193
k
0.635516
0.681064
10 Table of Data and Estimated Values of Interest
Dose
N
Obs Mean
Est Mean
Obs Std Dev
Est Std
Scaled Resid
0
10
9.98
9.98
1.68
1.44
-1.93e-008
0.1339
10
9.84
9.84
1.82
1.44
1.75e-008
0.8671
10
8.5
8.5
1.11
1.44
4.83e-008
5.248
10
7.7
7.7
1.38
1.44
-6.99e-008
11 Likelihoods of Interest
Model
Log(likelihood)
# Pa rams
AIC
A1
-34.705375
5
79.410749
A2
-33.333528
8
82.667056
A3
-34.705375
5
79.410749
fitted
-34.705375
5
79.410749
R
-41.888855
2
87.777711
12 Tests of Interest
Test
-2*log( Likelihood
Ratio)
Test df
p-value
Test 1 (Does response and/or variances differ among Dose
levels, A2 vs. R)
17.1107
6
0.008885
Test 2 (Are Variances Homogeneous, A2 vs. Al)
2.74369
3
0.4329
Test 3 (Are variances adequately modeled, A2 vs. A3)
2.74369
3
0.4329
Test 4 (Does the model for the Mean fit, A3 vs. fitted)
1.13687e-013
0
n/a
This document is a draft for review purposes only and does not constitute Agency policy.
C-35	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-18. Summary of BMD modeling results for decreased clotting time in female
Wistar rats exposed to 1,2,4-TMB by inhalation for 3 months; BMR = 1 SD
change from control mean (constant and modeled variance), (Korsak et
al., 2000a)
Constant Variance
Model3
Goodness-of-fit
BMD1sd
(mg/L)
BMDL1Sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.00311
207.7609
13.2329
4.78502
No model selected as Test
2 p-value was < 0.10.
Therefore, as suggested in
the Benchmark Dose
Technical Guidance (U.S.
EPA. 2012a), the data were
remodeled using a non-
homogenous variance
model
Exponential (M4)
0.3078
199.2547
0.119261
0.000258705
Exponential (M5)
n/ac
201.2538
0.12336
0.000534297
Hill
n/ac
201.25379
0.129946
1.20 x 1010
Lineard
Polynomial 2°
Polynomial 3°
Power
0.003013
207.824506
12.5899
5.12676
Modeled Variance
Model®
Goodness-of-fit
BMD1sd
(mg/L)
BMDL1sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.0001725
209.2185
16.2811
5.15229
No model selected as the
only appropriate fitting
models (Exponential
model 5) calculated an
implausibly low BMDL
Therefore, this endpoint
cannot be modeled in
BMDS and the
NOAEL/LOAEL approach is
recommended
Exponential (M4)
0.09227
196.7223
0.297031
0.000698259
Exponential (M5)
n/ac
198.7223
0.235929
7.68 x 10"05
Hill
n/ac
204.758516
0.138361
Not calculated
Lineard
Polynomial 2°
Polynomial 3°
Power
0.0001675
209.276823
15.0257
5.46511
a Constant variance case presented (Test 2 p-value = 0.02286).
b For Exponential model 3, the estimate of d was 1 (boundary). The models in this row reduced to exponential
model 2.
test had insufficient degrees of freedom (due to estimated model parameters = dose groups). Inspection of
scaled residuals and visual fit indicated appropriate model fit.
d For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
6 Modeled variance case presented (Test 3 p-value = 0.2001, except Hill model for which Test 3 p-value = <
0.0001).
Data Source: (Korsak et al., 2000a).
This document is a draft for review purposes only and does not constitute Agency policy.
C-36	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table C-19. Summary of BMD modeling results for decreased reticulocytes in female
Wistar rats exposed to 1,2,4-TMB by inhalation for 3 months; BMR = 1 SD
change from control mean (constant and modeled variance), (Korsak et
al., 2000a)
Constant Variance
Model3
Goodness-of-fit
BMD1sd
(mg/L)
BMDL1Sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.05738
91.21206
5.67056
0.775822
No model selected as Test
2 p-value was < 0.10.
Therefore, as suggested in
the Benchmark Dose
Technical Guidance (U.S.
EPA, 2012a), the data were
remodeled using a non-
homogenous variance
model
Exponential (M4)
0.2784
88.67076
0.107641
0.000190582
Exponential (M5)
n/ac
90.67077
0.111117
0.000273446
Hill
0.3149
88.506257
0.11386
6.85 x 10 15
Lineard
Polynomial 2°
Polynomial 3°
Power
0.04654
91.631076
6.34191
3.62271
Modeled Variance
Model®
Goodness-of-fit
BMD1sd
(mg/L)
BMDL1sd
(mg/L)
Basis for Model Selection
p-value
AIC
Exponential (M2)b
Exponential (M3)
0.01667
75.37239
12.0859
4.65557
No model selected as the
only appropriate fitting
models (Exponential
model 5 and Hill models)
did not calculate BMDLs.
Therefore, this endpoint
cannot be modeled in
BMDS and the
NOAEL/LOAEL approach is
recommended
Exponential (M4)f
Exponential (M5)
0.3582
70.02825
Not Computed
0
Hill
n/ac
89.127269
Not Computed
Not Computed
Lineard
Polynomial 2°
Polynomial 3°
Power
0.009093
76.584735
8.44761
5.29336
a Constant variance case presented (Test 2 p-value = < 0.0001).
b For Exponential model 3, the estimate of d was 1 (boundary). The models in this row reduced to exponential
model 2.
CX2 test had insufficient degrees of freedom (due to estimated model parameters = dose groups). Inspection of
scaled residuals and visual fit indicated appropriate model fit.
d For the power model, the power parameter estimate was 1 (boundary). For the polynomial 2° and 3° models,
the b2 and b3 coefficient estimates were 0 (boundary). The models in this row reduced to the Linear model.
6 Modeled variance case presented (Test 3 p-value = 0.253).
f For Exponential model 5, the estimate of d was 1 (boundary). The models in this row reduced to exponential
model 4.
Data source: (Korsak et al., 2000a).
This document is a draft for review purposes only and does not constitute Agency policy.
C-37	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Toxicological Review ofTrimethylbenzene
APPENDIX D. DOCUMENTATION OF
IMPLEMENTATION OF THE 2011 NATIONAL
RESEARCH COUNCIL RECOMMENDATIONS
Background: On December 23, 2011, The Consolidated Appropriations Act, 2012, was
signed into law (U.S. Congress. 20111. The report language included direction to EPA for the
Integrated Risk Information System (IRIS) Program related to recommendations provided by
the National Research Council (NRC) in its review of EPA's draft IRIS assessment of
formaldehyde (NRC. 20111. The report language included the following:
"The Agency shall incorporate, as appropriate, based on chemical-specific datasets and
biological effects, the recommendations of Chapter 7 of the National Research Council's Review
of the Environmental Protection Agency's Draft IRIS Assessment of Formaldehyde into the IRIS
process...For draft assessments released in fiscal year 2012, the Agency shall include
documentation describing how the Chapter 7 recommendations of the National Academy of
Sciences (NAS) have been implemented or addressed, including an explanation for why certain
recommendations were not incorporated."
The NRC's recommendations, provided in Chapter 7 of the review report, offered
suggestions to EPA for improving the development of IRIS assessments. Consistent with the
direction provided by Congress, documentation of how the recommendations from Chapter 7 of
the NRC report have been implemented in this assessment is provided in the tables below.
Where necessary, the documentation includes an explanation for why certain recommendations
were not incorporated.
The IRIS Program's implementation of the NRC recommendations is following a phased
approach that is consistent with the NRC's "Roadmap for Revision" as described in Chapter 7 of
the formaldehyde review report. The NRC stated that "the committee recognizes that the
changes suggested would involve a multi-year process and extensive effort by the staff at the
National Center for Environmental Assessment and input and review by the EPA Science
Advisory Board and others."
This document is a draft for review purposes only and does not constitute Agency policy.
D-l	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Toxicological Review ofTrimethylbenzene
Phase 1 of implementation has focused on a subset of the short-term recommendations,
such as editing and streamlining documents, increasing transparency and clarity, and using
more tables, figures, and appendices to present information and data in assessments. Phase 1
also focuses on assessments near the end of the development process and close to final posting.
The IRIS TMBs assessment is one of the first assessments in Phase 2 of implementation, which
addresses all of the short-term recommendations from Table D-l. The IRIS Program is
implementing all of these recommendations but recognizes that achieving full and robust
implementation of certain recommendations will be an evolving process with input and
feedback from the public, stakeholders, and external peer review committees. Phase 3 of
implementation will incorporate the longer-term recommendations made by the NRC as
outlined below in Table D-2. On May 16, 2012, EPA announced fU.S. EPA. 2012bl that as a part
of a review of the IRIS Program's assessment development process, the NRC will also review
current methods for weight-of-evidence analyses and recommend approaches for weighing
scientific evidence for chemical hazard identification. This effort is included in Phase 3 of EPA's
implementation plan.
This document is a draft for review purposes only and does not constitute Agency policy.
D-2	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table D-l. The EPA's implementation of the National Research Council's
recommendations in the trimethylbenzenes assessment
NRC recommendations that EPA is
implementing in the short term
Implementation in the trimethylbenzenes
assessment
General recommendations for completing the IRIS formaldehyde assessment that EPA will adopt for all IRIS
assessments (see p. 152)
1. To enhance the clarity of the document, the
draft IRIS assessment needs rigorous editing to
reduce the volume of text substantially and
address redundancies and inconsistencies. Long
descriptions of particular studies should be
replaced with informative evidence tables. When
study details are appropriate, they could be
provided in appendices.
Implemented. The overall document structure has been
revised in consideration of this NRC recommendation. The
new structure includes a concise Executive Summary and an
explanation of the literature review search strategy, study
selection criteria, and methods used to develop the
assessment. The main body of the assessment has been
reorganized into two sections, Hazard Identification and Dose-
Response Analysis, to help reduce the volume of text and
redundancies that were a part of the previous document
structure. Section 1 provides evidence tables and a concise
synthesis of hazard information organized by health effect,
More detailed summaries of the most pertinent epidemiology
and experimental animal studies are provided in Appendix B.
Information on chemical and physical properties and
toxicokinetics is also provided in Appendix B. The main text of
the Toxicological Review is approximately 90 pages, which is a
major reduction from previous IRIS assessments. Technical
and scientific edits were performed to eliminate any
redundancies or inconsistencies.
This document is a draft for review purposes only and does not constitute Agency policy.
D-3
DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table D-l (Continued): The EPA's implementation of the National Research Council's
	recommendations in the trimethylbenzenes assessment
NRC recommeiukitions that EPA is
implementing in the short term
Implementation in the trimethylbenzenes
assessment
2. Chapter 1 needs to be expanded to describe
more fully the methods of the assessment,
including a description of search strategies used
to identify studies with the exclusion and inclusion
criteria articulated and a better description of the
outcomes of the searches and clear descriptions
of the weight-of-evidence approaches used for
the various noncancer outcomes. The committee
emphasizes that it is not recommending the
addition of long descriptions of EPA guidelines to
the introduction, but rather clear concise
statements of criteria used to exclude, include,
and advance studies for derivation of the RfCs and
unit risk estimates.
Implemented. Chapter 1 has been replaced with a Preamble
that describes the application of existing EPA guidance and
the methods and criteria used in developing the assessment.
The term "Preamble" was chosen to emphasize that these
methods and criteria are being applied consistently across IRIS
assessments. The new Preamble includes information on
identifying and selecting pertinent studies, evaluating the
quality of individual studies, weighing the overall evidence of
each effect, selecting studies for derivation of toxicity values,
and deriving toxicity values. These topics correspond directly
to the five steps that the NRC identified in Figure 7-2 of their
2011 report.
A new section, Literature Search Strategy and Study Selection,
provides detailed information on the search strategy used to
identify health effect studies, search outcomes, and selection
of studies for hazard identification. This information is
chemical-specific and has been designed to provide enough
information that an independent literature search would be
able to replicate the results. This section also includes
information on how studies were selected to be included in
the document and provides a link to EPA's Health and
Environmental Research Online (HERO) database
(www.epa.gov/hero) that contains the references that were
cited in the document, along with those that were considered
but not cited.
3. Standardized evidence tables for all health
outcomes need to be developed. If there were
appropriates tables, long text descriptions of
studies could be moved to an appendix of
deleted.
Implemented. In the new document template, standardized
evidence tables that present key study findings that support
how toxicological hazards are identified for all major health
effects are provided in Section 1.1. More detailed summaries
of the most pertinent epidemiology and experimental animal
studies are provided in Appendix B.
4. All critical studies need to be thoroughly
evaluated with standardized approaches that are
clearly formulated and based on the type of
research, for example, observational
epidemiologic or animal bioassays. The findings of
the reviews might be presented in tables to
ensure transparency.
Partially implemented. Information in Section 4 of the
Preamble provides an overview of the approach used to
evaluate the quality of individual studies. Critical evaluation of
the epidemiologic and experimental animal studies is included
in the evidence tables in Section 1.1. Additional information
on study characteristics is found in Appendix B. The study
information for TMBs is presented in table format that clearly
presents detailed study summary information and key study
characteristics. As more rigorous systematic review processes
are developed, they will be utilized in future assessments.
This document is a draft for review purposes only and does not constitute Agency policy.
D-4	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table D-l (Continued): The EPA's implementation of the National Research Council's
	recommendations in the trimethylbenzenes assessment
NRC recommendations that EPA is
implementing in the short term
Implementation in the trimethylbenzenes
assessment
5. The rationales for the selection of the studies
that are advanced for consideration in calculating
the RfCs and unit risks need to be expanded. All
candidate RfCs should be evaluated together with
the aid of graphic displays that incorporate
selected information on attributes relevant to the
database.
Implemented. The Dose-Response Analysis section of the new
document structure provides a clear explanation of the
rationale used to select and advance studies that were
considered for calculating toxicity values. Rationales for the
selection of studies advanced for reference value derivation
are informed by the weight-of-evidence for hazard
identification as discussed in Section 1.2. In support of the RfC
derivations for individual TMB isomers, an exposure-response
array was included that compares effect levels for several
toxicological effects (Figures 2-1, 2-3, and 2-5). The exposure-
response array provides a visual representation of points of
departure for various effects resulting from exposure to TMB
isomers. The array informs the identification of doses
associated with specific effects, and the choice of principal
study and critical effects. In the case of TMBs, the database
supported development of multiple candidate RfC's. Such
values have been developed previously and will be developed
in future assessments, where the data allow.
6. Strengthened, more integrative and more
transparent discussions of weight-of-evidence are
needed. The discussions would benefit from more
rigorous and systematic coverage of the various
determinants of weight-of-evidence, such as
consistency.
Partially implemented. A new section, Hazard Identification
(Section 1), provides a more strengthened, integrated and
transparent discussion of the weight of the available
evidence. This section includes standardized evidence tables
to present the key study findings that support how potential
toxicological hazards are identified and exposure-response
arrays for each potential toxicological effect. Weight-of-
evidence discussions are provided for each major effect
(Section 1.1.1—neurotoxic effects, Section 1.1.2—respiratory
effects, Section 1.1.3-reproductive/ developmental effects,
and Section 1.1.4—hematological and clinical chemistry
effects). A more rigorous and formalized approach for
characterizing the weight-of-evidence will be developed as a
part of Phase 3 of the implementation process.
General Guidance for the Overall Process (p. 164)
7. Elaborate an overall, documented, and quality-
controlled process for IRIS assessments.
Implemented. EPA has created Chemical Assessment Support
Teams to formalize an internal process to provide additional
overall quality control for the development of IRIS
assessments. This initiative uses a team approach to making
timely, consistent decisions about the development of IRIS
assessments across the Program. This team approach has
been utilized for the development of the TMBs assessment.
Additional objectives of the teams is to help ensure that the
necessary disciplinary expertise is available for assessment
development and review, to provide a forum for identifying
and addressing key issues prior to external peer review, and
to monitor progress in implementing the NRC
recommendations.
8. Ensure standardization of review and
evaluation approaches among contributors and
teams of contributors; for example, include
standard approaches for reviews of various types
of studies to ensure uniformity.
9. Assess disciplinary structure of teams needed
to conduct the assessments.
This document is a draft for review purposes only and does not constitute Agency policy.
D-5	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table D-l (Continued): The EPA's implementation of the National Research Council's
	recommendations in the trimethylbenzenes assessment
NRC recommeiukitions that EPA is
implementing in the short term
Implementation in the trimethylbenzenes
assessment
Evidence Identification: Literature Collection and Collation Phase (p. 164)
10. Select outcomes on the basis of available
evidence and understanding of mode of action.
Partially implemented. A new section, Literature Search
Strategy and Study Selection, contains detailed information
on the search strategy used for the TMBs assessment,
including key words used to identify relevant health effect
studies. Figure LS-1 depicts the study selection strategy and
the number of references obtained at each stage of literature
screening. This section also includes information on how
studies were selected to be included in the document and
provides a link to an external database (www.eoa.gov/hero)
that contains the references that were cited in the document,
along with those that were considered but not cited. Each
citation in the Toxicological Review is linked to HERO such
that the public can access the references and abstracts to the
scientific studies used in the assessment.
Section 3 of the Preamble summarizes the standard protocols
for evidence identification that are provided in EPA guidance.
For each potential toxicological effect identified for TMBs, the
available evidence is informed by the mode of action
information as discussed in Section 1.1. As more rigorous
systematic review processes are developed, they will be
utilized in future assessments.
11. Establish standard protocols for evidence
identification.
12. Develop a template for description of the
search approach.
13. Use a database, such as the Health and
Environmental Research Online (HERO) database,
to capture study information and relevant
quantitative data.
Evidence Evaluation: Hazard Identification and Dose-Response Modeling (p. 165)
14. Standardize the presentation of reviewed
studies in tabular or graphic form to capture the
key dimensions of study characteristics, weight-
of- evidence, and utility as a basis for deriving
reference values and unit risks.
Implemented. Standardized tables have been developed that
provide summaries of key study design information and
results by health effect. The inclusion of all positive and
negative findings in each health effect-specific evidence table
supports a weight-of-evidence analysis. In addition, exposure-
response arrays are utilized in the assessment to provide a
graphical representation of points of departure for various
effects resulting from exposure to TMB. The exposure-
response arrays inform the identification of doses associated
with specific effects and the weight-of- evidence for those
effects.
15. Develop templates for evidence tables, forest
plots, or other displays.
Implemented. Templates for evidence tables and exposure-
response arrays have been developed and are utilized in
Section 1.1.
16. Establish protocols for review of major types
of studies, such as epidemiologic and bioassay.
Partially implemented. General principles for reviewing
epidemiologic and experimental animal studies are described
in Section 4 of the Preamble. Standardized systematic review
is an ongoing process.
This document is a draft for review purposes only and does not constitute Agency policy.
D-6	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table D-l (Continued): The EPA's implementation of the National Research Council's
	recommendations in the trimethylbenzenes assessment
NRC recommendations that EPA is
implementing in the short term
Implementation in the trimethylbenzenes
assessment
Selection of Studies for Derivation of Reference Values and Unit Risks (p. 165)
17. Establish clear guidelines for study selection.
a.	Balance strengths and weaknesses.
b.	Weigh human vs. experimental evidence
c.	Determine whether combining estimates
among studies is warranted.
Implemented. EPA guidelines for study selection, including
balancing strengths and weaknesses and weighing human vs.
experimental evidence are described in the Preamble
(Sections 3-6). These guidelines have been applied in Section
2 of the TMBs assessment to inform the evaluation of the
weight-of-evidence across health effects and the strengths
and weaknesses of individual studies considered for reference
value derivation.
In the case of TMBs, the database did not support the
combination of estimates across studies. In future
assessments, combining estimates across studies will be
routinely considered.
Calculation of Reference Values and Unit Risks (pp. 165-166)
18. Describe and justify assumptions and models
used. This step includes review of dosimetry
models and the implications of the models for
uncertainty factors; determination of appropriate
points of departure (such as benchmark dose, no-
observed-adverse-effect level, and lowest
observed-adverse-effect level), and assessment of
the analyses that underlie the points of departure.
Implemented as applicable.
The rationale for the selection of the point of departure (a
95% lower confidence limit on the benchmark dose; BMDL)
for the derivation of the inhalation reference value for
1.2.4-TMB	and 1,2,3-TMB is transparently described in Section
2. The determination of sufficient similarity regarding
1.3.5-TMB	and 1,2,4-TMB, and the decision to adopt the RfC
for 1,2,4-TMB as the RfC for 1,3,5-TMB, is transparently
described in Section 2.
The rationale for the route-to-route extrapolation in order to
use inhalation data for derivation of an RfD for 1,2,4-TMB is
transparently described in Section 2. The determination of
sufficient similarity regarding 1,2,3-, 1,2,4-, and 1,3,5-TMB,
and the decision to adopt the RfD for 1,2,4-TMB as the RfDs
for 1,2,3-TMB and 1,3,5-TMB, is transparently described in
Section 2.
A summary of the benchmark dose modeling for the
derivation of the reference values for effects other than
cancer, including an alternative analysis with high doses
included, is described in Appendix C.
19. Provide explanation of the risk-estimation
modeling processes (for example, a statistical or
biologic model fit to the data) that are used to
develop a unit risk estimate.
Not applicable. The TMB assessment concludes that there is
inadequate information to assess the carcinogenic potential.
Therefore, a unit risk estimate for cancer was not derived.
This document is a draft for review purposes only and does not constitute Agency policy.
D-7	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table D-l (Continued): The EPA's implementation of the National Research Council's
	recommendations in the trimethylbenzenes assessment
NRC recommendations that EPA is
implementing in the short term
Implementation in the trimethylbenzenes
assessment
20. Provide adequate documentation for
conclusions and estimation of reference values
and unit risks. As noted by the committee
throughout the present report, sufficient support
for conclusions in the formaldehyde draft IRIS
assessment is often lacking. Given that the
development of specific IRIS assessments and
their conclusions are of interest to many
stakeholders, it is important that they provide
sufficient references and supporting
documentation for their conclusions. Detailed
appendixes, which might be made available only
electronically, should be provided when
appropriate.
Implemented. The new template structure that has been
developed in response to the NRC recommendations provides
a clear explanation of the literature search strategy, study
selection criteria, and methods used to develop the TMBs
assessment. It provides for a clear description of the decisions
made in developing the hazard identification and dose-
response analysis. Information contained in the Preamble and
throughout the document reflects the guidance that has been
utilized in developing the assessment. As recommended,
supplementary information is provided in the accompanying
appendices.
This document is a draft for review purposes only and does not constitute Agency policy.
D-8	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Table D-2. National Research Council recommendations that the EPA is generally
implementing in the long term
NRC recommeiukitions that the EPA is
generally implementing in the long term
Implementation in the tri methyl benzenes
assessment
Weight-of-Evidence Evaluation: Synthesis of
Evidence for Hazard Identification (p. 165)
1.	Review use of existing weight-of-evidence
guidelines.
2.	Standardize approach to using weight-of-
evidence guidelines.
3.	Conduct agency workshops on approaches to
implementing weight-of-evidence guidelines.
4.	Develop uniform language to describe strength
of evidence on noncancer effects.
5.	Expand and harmonize the approach for
characterizing uncertainty and variability.
6.	To the extent possible, unify consideration of
outcomes around common modes of action
rather than considering multiple outcomes
separately.
As indicated above, Phase 3 of EPA's implementation plan will
incorporate the longer-term recommendations made by the
NRC, including the development of a standardized approach
to describe the strength of evidence for noncancer effects. On
Mav 16, 2012, EPA announced (U.S. EPA, 2012b) that as a part
of a review of the IRIS Program's assessment development
process, the NRC will also review current methods for weight-
of-evidence analyses and recommend approaches for
weighing scientific evidence for chemical hazard
identification. In addition, EPA will hold a workshop on August
26, 2013, on issues related to weight-of-evidence to inform
future assessments.
Calculation of Reference Values and Unit Risks
(pp. 165-166)
7. Assess the sensitivity of derived estimates to
model assumptions and end points selected. This
step should include appropriate tabular and
graphic displays to illustrate the range of the
estimates and the effect of uncertainty factors on
the estimates.
As discussed in Section 1.2, although the nervous system is
the primary and most sensitive target of inhaled TMB toxicity,
there is evidence of effects in other organ systems. Candidate
RfCs for 1,2,4-TMB and 1,2,3-TMB are evaluated together in
Figures 2-2 and 2-4 (respectively), including the uncertainty
factors applied to individual endpoints.
This document is a draft for review purposes only and does not constitute Agency policy.
D-9	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Toxicological Review ofTrimethylbenzene
APPENDIX E. SUMMARY OF AVAILABLE C9
AROMATIC HYDROCARBON FRACTION TOXICITY
STUDIES
As part of a testing program mandated under Section 4(a) of the Toxic Substances Control
Act (TSCA), a series of toxicity tests were performed that investigated the mutagenicity,
developmental and reproductive toxicity, subchronic neurotoxicity, and general inhalation
toxicity of the C9 aromatic hydrocarbon fraction (C9 fraction), which is mostly comprised of the
ortho-, meta-, and para- isomers of ethyltoluene (2-, 3-, and 4-ethyltoluene, respectively) and
the 1,2,4-, 1,2,3- and 1,3,5- isomers of trimethylbenzene fU.S. EPA. 19851. The final testing
criteria required that the representative C9 fraction test substance be comprised of no less than
22% ethyltoluene isomers and 15% trimethylbenzene isomers, and required a total
ethyltoluene/trimethylbenzene content greater than 75% fU.S. EPA. 1985) (see Tables E-l and
E-2 for detailed descriptions of the final test substances used). The results of these toxicity tests
were subsequently published in the following references, and are discussed individually below:
mutagenicity (Schreiner et al.. 1989): developmental and reproductive toxicity (Mckee etal..
19901: subchronic neurotoxicity fDouglas etal.. 19931: and general inhalation toxicity fClark et
al.. 19891.
Table E-l. Composition of the C9 fraction test substance used for toxicity testing in
Schreiner et al. (1989), McKee et al. (1990), and Douglas et al. (1993)
Compound
Weight percent
o-xylene
3.20
Cumene (isopropylbenzene)
2.74
n-propylbenzene
3.97
4-ethyltoluene
7.05
3-ethyltoluene
15.1
2-ethyltoluene
5.44
1,2,4-trimethylbenzene
40.5
1,2,3-trimethylbenzene
6.18
1,3,5-trimethylbenzene
8.37
>C10
6.19
TOTAL
98.74
Source: Schreiner et al. (1989). McKee et al. (1990). and Douglas et al. (1993)
This document is a draft for review purposes only and does not constitute Agency policy.
E-l	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Toxicological Review ofTrimethylbenzene
Table E-2. Composition of the C9 fraction test substance used for toxicity testing in
Clark etal. (1989)
Compound
Weight percent
non-aromatics
0.46
o-xylene
2.27
n-propylbenzene
4.05
4-ethyltoluene
16.60
3-ethyltoluene
7.14
2-ethyltoluene
7.22
1,2,4-trimethylbenzene
32.70
1,2,3-trimethylbenzene
2.76
1,3,5-trimethylbenzene
9.35
>C10

l-methyl-3-n-propylbenzene + 1,2-diethylbenzene
6.54
1 - ethyl- 3,5 -dimethylbenzene
1.77
TOTAL
90.86
Source: Clark et al. (1989)
Schreiner et al. (1989) assessed the mutagenic potential of the C9 fraction (see Table E-l;
total trimethylbenzene content = 55.05%) by measuring five genotoxic endpoints: mutation
frequency in bacteria, mutation frequency in CHO cells (chinese hamster ovary cells), sister
chromatid exchange in CHO cells, chromosomal aberrations in CHO cells, and chromosome
aberrations in rat bone marrow cells. In the bacterial mutagenicity assay, five Salmonella
typhimurium test strains (TA98, TA100, TA1535, TA1537, and TA1538) were exposed to either
negative controls (DMSO), positive controls, or to 0.0025-0.50 [il/plate C9 fraction in the
presence or absence of the S9 microsomal mixture. After 72 hours of incubation, cells exposed
to positive controls exhibited greater rates of gene mutations than negative controls. However,
there was no evidence that the C9 fraction induced gene mutations with or without S9
activation in any S. typhimurium strain up to the highest test concentration, at which signs of
cellular toxicity became apparent In the CHO forward mutation assay, CHO cells were exposed
for 4 hours to either negative controls (DMSO), positive controls, or the C9 fraction at 0.01-0.13
[iL/ml (-S9) or 0.02-0.2 [iL/ml (+S9). After a seven day post-exposure incubation period, CHO
cells exposed to positive controls exhibited statistically significantly greater mutation
frequencies compared to negative controls, while no evidence of C9 fraction-induced mutations
were observed at any test concentration. To test for the induction of sister chromatid exchange
in vitro, CHO cells were exposed to controls or the C9 fraction (2.0-66.7 |J.g/ml - S9 for 22.5
hours or 0.667-50.1 |J.g/ml + S9 for 2 hours). Cell-cycle arrest was not observed at exposure
concentrations lower than 66.7 or 50.1 |J.g/ml C9 fraction (-S9 or + S9, respectively), and %
This document is a draft for review purposes only and does not constitute Agency policy.
E-2	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
38
Toxicological Review ofTrimethylbenzene
SCE/cell was not increased in cells exposed to any concentration of C9 fraction. The ability of
the C9 fraction to induce chromosomal aberrations was similarly investigated in CHO cells: no
exposure concentration of the C9 fraction, up to 90.2 [ig/mL, induced chromosomal aberrations
in the absence or presence of S9. In order to investigate the potential in vivo mutagenicity of the
C9 fraction, Sprague-Dawley rats (30 per exposure group, 15 male and 15 female) were
exposed via inhalation to 0,150, 500, or 1500 ppm C9 fraction for 6 hours on 5 consecutive
days. Following the termination of exposure, ten rats from each treatment group were
sacrificed at 6, 24, and 48 hours, and their bone marrow was harvested, stained, and examined
for chromosome/chromatid aberrations. No induction of chromosomal/chromatid aberrations
were observed at any exposure concentration in animals sacrificed at 6 or 24 hours. No
aberrations were observed in animals sacrificed at 48 hours, but the majority of samples
(approximately 66%) were not analyzed due to inadequate staining. In general, the results of
Schreiner et al. (1989) indicate that, as tested, the C9 fraction did not induce in vitro or in vivo
mutagenicity in multiple assays.
The developmental and reproductive toxicity of the C9 fraction (see Table E-l; total
trimethylbenzene content = 55.05%) was investigated by McKee et al. (1990). In the
developmental toxicity portion of the study, pregnant CD-I mice (30 per group) were exposed
to 0,100, 500, or 1500 ppm C9 fraction for 6 hours/day on gestational days (GD) 6-15 (nominal
concentrations: 102 ± 3.5, 463 ± 5.3, and 1249 ± 16.5 ppm; actual concentrations: 102 ± 2.6,
500 ± 3.7, or 1514 ± 22.9 ppm). Throughout the exposure period, the dams were examined for
clinical signs twice daily, and body weights were taken daily. Blood samples were taken from
the dams on GDI5, and surviving dams were sacrificed on GDI8. The number and location of
live and dead fetuses were recorded, as well as the total number of implantations and corpora
lutea. Fetuses were weighed, sexed, half of the fetuses examined for external malformations, the
remaining fetuses were examined for skeletal malformations. Severe maternal toxicity was
observed in the highest exposure group (i.e., 1514 ppm), with 44% of animals dying before
sacrifice. Only two dams died in the 500 ppm group, and no animals died in the 102 ppm group.
Maternal body weight was statistically significantly decreased at all exposure concentrations on
GD15, but only in the 1514 ppm group on GD 18. Maternal body weight gain was decreased in
both the 500 and 1514 ppm exposure groups when measured on GD6-15 and GDO-18. Clinical
observations of dams revealed some evidence of gross neurobehavioral toxicity, including
abnormal gait (18 animals), labored breathing (9 animals), weakness (7 animals), circling (8
animals), and ataxia (8 animals). There were no differences in maternal organ weights in any
exposure group compared to controls. Hematocrit and mean corpuscular volume were
significantly decreased in dams exposed to 1514 ppm. Exposure to 1514 ppm also resulted in
severe developmental toxicity: the number of litters with viable fetuses was decreased (13 vs.
24 in controls, no statistics provided), the number of live fetuses/litter was statistically
significantly decreased (7.9 ± 4.3 vs. 10.7 ± 1.8 in controls), and postimplantation loss/dam was
This document is a draft for review purposes only and does not constitute Agency policy.
E-3	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
38
Toxicological Review ofTrimethylbenzene
significantly increased (4.3 ± 3.7 vs. 0.9 ± 0.9 in controls). Exposure to both 500 and 1514 ppm
also resulted in decreased fetal body weights (1.16 ± 0.11 g [500 ppm] and 0.82 ± 0.17 g [1514
ppm] vs. 1.25 ± 0.14 g in controls). Increased incidence of cleft palate and unossified sternebrae
(# 5 and/or 6) was observed in the 1514 ppm group. No other evidence of teratogenicity was
observed. The NOAEL identified from the developmental toxicity portion of McKee et al. (1990)
was 100 ppm based on decreased fetal weight.
In the reproductive portion of McKee et al. (1990), male and female CD (30 of each sex) rats
were randomly assigned to one of four exposure groups (i.e., 0,103, 495, or 1480 ppm [nominal
concentrations: 107 ± 24, 513 ± 12.8, or 1483 ± 33.0 ppm; actual concentrations: 103 ± 2.1, 495
± 8.0, or 1480 ± 20.5 ppm), and were exposed 6 hours/day, 5 days/week for 10 weeks. Then
males and females were co-housed (1:1) for a two week mating period. When mating was
confirmed, males were sacrificed. Females were additionally exposed to the C9 fraction 6
hours/day, 7 days/week from GD0 to GD20. Dams were then allowed to deliver. F1 pups were
culled to 8/litter on post-natal day (PND) 8. Exposure was restarted on PND 5 and continued
until PND21 (weaning), at which point dams were sacrificed and F1 pups were counted, sexed,
and weighed. F1 pups were randomly selected for further use in the study (i.e., one week after
weaning, they were exposed for 10 weeks and then mated for 2 weeks to produce the F2
generation). The F2 litters were treated similarly to the F1 litters, but to investigate the effects
of exposure immediately after weaning; exposure of the animals used to produce the F3
generation began immediately after weaning (i.e., PND22). All parental animals were examined
twice daily for clinical signs of toxicity, and body weights were measured weekly. At sacrifice,
organs and tissues were microscopically examined in the control and high exposure groups.
Litters were examined immediately after delivery for litter size, stillbirths, live births, and gross
anomalies. Culled pups and any pups that died spontaneously were necropsied. Pups were
weighed on PND0, 4, 7, and 14.
All Fo males survived exposure, whereas seven Fo females in the 1480 ppm group died (3
prior to mating, 3 during gestation, and 1 during lactation). Weight gain in both Fo males and
females was statistically significantly decreased in the 495 and 1480 ppm groups. No
pathological lesions in the reproductive organs were noted in F0 generation animals (or in any
Fi, F2, or F3 animals). There were no observed alterations in female or male fertility, number of
females delivering a litter, or litter size at birth. There was a small, but not statistically
significant, increase in time necessary for successful mating. In the Fi generation, there were no
decreases in birth weight, or body weights at PND4, compared to controls. Starting on PND7,
and continuing until weaning, body weights were significantly decreased in the 1480 ppm
exposure group relative to control. No differences in post-natal survival were observed. The
decreased body weights of Fi males and females in the 1480 ppm group was still manifest when
exposure was reinitiated (i.e., 10 days after weaning), and during the pre-mating period, body
weight gains were lower than controls in males at 495 ppm and in males and females at 1480
This document is a draft for review purposes only and does not constitute Agency policy.
E-4	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
Toxicological Review ofTrimethylbenzene
ppm. Fi males and females in the 1480 ppm group also exhibited some signs of neurotoxicity
demonstrated by ataxia (18 males, 23 females) and/or decreased motor activity (11 males, 8
females). Male fertility (number of fertile males/number of mated males) was significantly
decreased at 1480 ppm. Lastly, six females in the 1480 died (three during exposure, one during
gestation, one during delivery, and one on PND2). There were statistically significant reductions
in the number of live F2 offspring delivered per litter and the percentage of live F2 births. F2
birth weights were also decreased, but not significantly. The authors report that among mated
Fi females, mating of 24 females (six in the control group, eight at 103 ppm, one at 495 ppm,
and nine at 1480 ppm) was not confirmed, and exposure was carried out until delivery, rather
than being terminated on GD20. When the dams were analyzed as separate groups, the F2 litter
size was only statistically significantly decreased in litters delivered from the dams that were
exposed until delivery. In dams whose exposure was terminated on GD20, F2 litter size was
slightly decreased, but not significantly so. The percentage of live births was decreased in both
groups of dams; among the dams that were exposed until delivery, pup survival was still
decreased at PND4. As with the Fi generation, F2 body weights at birth through PND4 were not
affected by treatment, but starting on PND7 and continuing until weaning, F2body weights were
statistically significantly decreased in the 1480 ppm group.
As stated above, the pre-mating exposure of F2 animals selected to produce the F3
generation was begun immediately after weaning (i.e., PND21). A majority of animals in the
1480 ppm group died during the first week of exposure (36/40 males, 34/40 females). Of the
high exposure group animals that survived, body weights were substantially reduced
throughout the pre-mating exposure period (31-40% below controls in males and 21-35%
below in females). Additionally, body weights were slightly decreased in the 103 ppm (10%
males, 6% females) and 495 ppm (16% males and females) exposure groups. There were no
observed effects on the mean number of live F3 births or post-natal survival. Birth weights of
the F3 generation were statistically significantly decreased in the 1480 ppm group. Birth
weights at PND7 were decreased in the 1480 ppm group, and beginning on PND14 through
weaning, body weights were statistically decreased in the 495 and 1480 ppm group. In general,
the results of McKee et al. (1990) indicate that exposure to the C9 fraction can induce
developmental toxicity (decreased live fetuses, increased postimplantation loss, increased
malformations [cleft palate], and decreased fetal weight) and possibly reproductive toxicity
(decreased male fertility). Multigenerational exposures were also observed to induce
decrements in postnatal weights that occurred at lower doses in later generations compared to
earlier generations. Consequently, the NOAEL identified from the reproductive portion of
McKee (1990) was 100 ppm based decreased fetal weights in the F3 generation. Lastly, pre-
mating exposures of adult animals was observed to elicit some measures of neurotoxicity
(ataxia and decreased motor activity).
This document is a draft for review purposes only and does not constitute Agency policy.
E-5	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
38
Toxicological Review ofTrimethylbenzene
Douglas et al. (19931 investigated the neurotoxic potential of the C9 fraction (see Table E-l;
total trimethylbenzene content = 55.05%) following subchronic exposure to the mixture. Male
CD rats (20 per group) were exposed to 0,100, 500, or 1500 ppm C9 fraction for 6 hours/day, 5
days/week for 13 weeks (nominal concentration: 94 ± 1.0, 481 ± 5.1, and 1334 ± 17.0 ppm;
actual concentration: 101 ± 2.5, 432 ± 2.8,1320 ± 13.0 ppm). Body weights were recorded
weekly during the exposure period, and animals were examined for clinical signs at these time
points. Following termination of exposure, animals were sacrificed and tissues were removed
for histopathology. Specific testing for neurotoxicity was performed 5, 9, and 13 weeks after
exposure was begun. Specific neurotoxicity tests included examination of motor activity
(frequency of movement within a cage), and a functional observation battery (fore and hind
limb grip strength, audio startle response, thermal response [hotplate stimulus test], and hind
limb foot splay). Histopathological examinations were performed on the central and peripheral
nerve tissue, including the proximal sciatic nerves, dorsal root ganglia, spinal cord, and specific
regions of the brain. No animals died during the exposure period, and the only reported clinical
signs reported were urogenital staining, urination, defecation, and vocalization (none of which
were considered treatment related). Animals in the high exposure group (i.e., 1320 ppm)
exhibited statistically decreased body weights at every time point during exposure; animals in
the 432 ppm group had decreased body weights early during exposure, with a statistically
significant decrease at week 4. However, by the end of the exposure period, these animals
weighted more than controls. No consistent treatment-related effects on motor activity were
reported. When analyzed in 10 minute blocks, horizontal activity and total activity during
minutes 10-20 of the test were statistically significantly increased in the 1320 ppm exposure
group during week nine of the exposure period. However, motor activity in this exposure group
returned to control levels during minutes 20-30 of the test, and no effects were observed at the
termination of exposure (i.e., week 13). When results were summarized across the entire 30
minute test period, no effects on motor activity were reported at any time during the 13 week
exposure period. The results of all the neurotoxicity tests comprising the functional observation
battery were generally negative, except for a transient and non-treatment related decrease in
auditory startle response in the 432 ppm exposure group at week nine of exposure.
Additionally, there appeared to be a statistically significant increase in thermal response time
when the endpoint was measured immediately prior to the exposure period. However, this was
most likely a statistical artifact due to an unusually low control group response measure at this
time point No exposure-related incidences of neuropathological lesions were reported
following termination of exposure. In general, the results of Douglas et al. (1993) indicate that
the C9 fraction is not neurotoxic to adult rats: as no consistent neurotoxic effects were noted,
the NOAEL for this study was identified as 1320 ppm. However, this finding appears to be in
disagreement with the reported neurotoxic effects noted in the McKee et al. (1990)
developmental and multigenerational reproductive study, in which pregnant and non-pregnant
This document is a draft for review purposes only and does not constitute Agency policy.
E-6	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
38
Toxicological Review ofTrimethylbenzene
adult animals exposed to similar levels of C9 exhibited gross signs of neurotoxicity, including
abnormal gait, weakness, ataxia and decreased motor activity.
Clark et al. (1989). investigated the inhalation toxicity of the C9 fraction (see Table E-2; total
trimethylbenzene content = 44.81%) following exposure of male and female Wistar rats (50
animals per sex per group) to 0, 450, 900, or 1800 mg/m3 for 6 hours/day, 5 days/week for up
to 12 months (actual concentration: 470 ± 29, 970 ± 70,1830 ± 130 mg/m3). Ten males and
females were sacrificed halfway through the exposure period (i.e., at 6 months), 25 males and
females were sacrificed at 12 months (i.e., at the end of exposure), and 15 males and females
were sacrificed after a four month recovery period. Animals were examined twice daily for
overt signs of toxicity, and body weights were recorded weekly through the first month of
exposure and monthly thereafter. Tail vein blood was taken periodically during exposure
(weeks 1, 2, 3, 4, 6, 8,12, 20, 24, 28, and 32) from 10 males and females in the control and high
dose group, and cardiac blood was collected from 10 males and females in all groups at the 6
and 12 month necropsies, and after the recovery period. Both types of blood samples were
analyzed for common hematological parameters (e.g., erythrocyte count, hemoglobin
concentration). Cardiac and tail vein blood was additionally drawn from 10 males and females
at the 6 and 12 month necropsies and at the end of the recovery period and analyzed for clinical
chemistry parameters (e.g., total protein, alkaline phosphatase). Urine samples were collected
from 12 males and females at 0, 3, 6, 9, and 12 months' exposure and analyzed for common
urinalysis parameters (e.g., glucose, protein). All animals underwent complete necropsies after
sacrifice. The only reported treatment-related clinical sign was an increase in aggression (i.e.,
difficulty in handling) in males in the high exposure group. Three control (two male, one
female) and two males in the low exposure group died during exposure. Body weights were
slightly decreased (2-3%) relative to control during the first 4 weeks in male rats exposed to
1830 mg/m3 and females exposed to 970 mg/m3 and during the first 12 weeks of exposure in
females exposed to 1830 mg/m3. No consistent trends were reported for any of the
hematological parameters analyzed from the tail vein samples. In the interim (i.e., 6 month) and
terminal (i.e., 12 month) cardiac blood samples, the only treatment-related effects reported
were decreased eosinophil counts (30 to 55%, all exposure groups) in female rats at 6 months
and decreased osmotic fragility (5%, all exposure groups) and increased lymphocyte counts
(29%, 1830 mg/m3) in male rats at 12 months. Clinical chemistry effects were generally mild,
with high exposure group females exhibiting increased potassium (6 months), increased
sodium (12 months), and decreased albumen (6 months); the only clinical chemistry effect
noted in males was increased creatinine in the high exposure group at 12 months. There were
no urinalysis parameters affected by treatment At the end of exposure, liver and kidney
weights were statistically significantly increased (11% and 10%, respectively) in high exposure
group males. Gross and histopathological examination generally revealed no consistent
treatment-related lesions. A slight increase in pulmonaiy macrophage infiltration and alveolar
This document is a draft for review purposes only and does not constitute Agency policy.
E-7	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
38
Toxicological Review ofTrimethylbenzene
wall thickening was observed in male and female rats at 12 months, with the average severity
grade for these effects increasing with dose. Although there were no clear treatment-related
increases in tumors at 12 months, one high exposure female had a leiomyoma on the left
uterine horn, one high exposure male had a lymphoma of the spleen, and one low exposure
male had a glioblastoma of the cerebellum. In general, the results of Clark et al. (1989) indicate
that the C9 fraction has low systemic toxicity (NOAEL = 1830 mg/m3) following chronic
exposure.
In summary, the results of Schreiner et al. (1989), McKee et al. (1990), Douglas et al. (1993.),
and Clark et al. (1989) are all well-conducted studies that investigate relevant toxicological
endpoints in appropriate in vitro and in vivo systems. These toxicity tests were mandated by
Section 4(a) of TSCA to investigate the mutagenicity, neurotoxicity, teratogenicity, reproductive
toxicity, and general toxicity of the C9 fraction, and indicated that the C9 fraction elicited limited
toxicity in the test systems used. It must be acknowledged that the specific test compound used
in the C9 fraction was a complex aromatic hydrocarbon mixture containing between 45-55%
TMB isomers, with the remaining mixture primarily consisting of ethyltoluene isomers. Tertiary
constituents (xylene, n-propyl- and isopropylbenzene, and unspecified C10 aromatic
hydrocarbons) comprised as much as 16% of the test compound. Although a conclusion of
sufficient toxicokinetic and toxicological similarity is used in the Toxicological Review to
support the adoption of consistent, cross-isomer reference values, such a conclusion has not
been reached, nor attempted, for the other constituents of the C9 mixture. For some
constituents (i.e., the C10 compounds), such a comparison is not possible as they were not
specifically identified in the compositional analysis. Regarding possible toxicokinetic
similarities, the EPA is currently unaware of any detailed data on the ADME of the C9 fraction
(particularly information regarding the distribution of TMB isomers in the C9 fraction to the
brain and other organ systems). As such, given this particular data gap, an assumption that the
C9 fraction would be an adequate surrogate for individual TMB isomers is not justified.
Additionally, the C9 mixture studies failed to observe clearly adverse effects, except for the
developmental and reproductive toxicity observed in McKee et al. (1990). However, multiple
peer-reviewed studies investigating the toxic effects of individual isomers exist, and serve as
the basis for hazard identification, dose-response analysis, and reference value derivation in the
Toxicological Review of Trimethylbenzenes. These studies include those observing
neurotoxicity fWiaderna etal.. 2002: Gralewicz and Wiaderna. 2001: Wiaderna et al.. 1998:
Gralewicz etal.. 1997b: Gralewicz et al.. 1997a: Korsaketal.. 1995). respiratory toxicity (Korsak
etal.. 2000a. b; Korsak etal.. 1997: Korsaketal.. 1995). developmental toxicity (Saillenfaitetal..
2005). and hematological toxicity (Korsak et al.. 2000a. b). Given the availability of these
studies, and the general lack of observed toxicity in the C9 studies, it is appropriate for the
individual isomer studies to serve as the scientific foundation for the Toxicological Review.
Therefore, although there are available, peer reviewed studies investigating the toxicity of the
This document is a draft for review purposes only and does not constitute Agency policy.
E-8	DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Toxicological Review ofTrimethylbenzene
C9 fraction, the uncertainty regarding any interactive effects other C9 constituents may have on
the ADME of TMB isomers and the general lack of reported toxic effects limit their utility for the
assessment of the human health risk of individual TMB isomers. For these reasons, these
studies were not included in the Toxicological Review.
Additionally, two other industry reports regarding the toxicity of mixtures containing the
isomers were located (Industrial Bio-Test Laboratories. 1992: Chevron. 19851. These
documents were excluded from the Toxicological Review following careful consideration as
they were not peer-reviewed and did not investigate the toxicity of individual TMB isomers.
Ultimately, the decision was made to not seek external peer review for these documents as
these studies would not qualitatively enhance hazard identification, quantitatively enhance
dose-response analysis, or substantially decrease uncertainty in the assessment. Two peer-
reviewed studies investigating the effects of complex mixtures containing TMB isomers were
also found fLehotzkv etal.. 1985: Ungvarv andTatrai. 19851. However, these studies also did
not study TMB isomers individually, and unlike the C9 fraction studies above, provided no
information on the compositional makeup of the test substance. For these reasons, the above
studies were not included in the Toxicological Review of Trimethylbenzenes.
This document is a draft for review purposes only and does not constitute Agency policy.
E-9	DRAFT—DO NOT CITE OR QUOTE
-------
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
Toxicological Review ofTrimethylbenzene
APPENDIX F. RESOLUTION OF PUBLIC COMMENTS
The Toxicological Review of Trimethylbenzenes was released for a 60-day public comment period
in June 2012. After the close of the public comment period, a listening session was held on August
1st, 2012. EPA received comments on the draft assessment from one public reviewer: the
Hydrocarbon Solvents Panel of the American Chemistry Council (ACC). The major comments
received have been synthesized and paraphrased below with a reference to the complete comment
also provided. EPA's responses to the comments as well as information regarding how the
assessment has been revised are also included.
Comment. The Draft IRIS Assessment is subject to EPA and OMB Information Quality Guidelines,
and, as the Draft IRIS Assessment is influential information, it must adhere to a rigorous standard of
quality. EPA must employ "a higher degree of transparency regarding (1) the source of the data
used, (2) the various assumptions employed, (3) the analytic methods applied, and (4) the
statistical procedures employed." As currently presented, the Draft Assessment has failed to
comport with the Information Quality Guidelines (Comments 1.1 and 1.2, pp 4-6)
EPA Response: In response to NRC recommendations, EPA has increased the transparency of IRIS
assessments, particularly in regard to (1) the source of the data used (i.e., inclusion of evidence
tables in the main body of the Toxicological review, and study summary tables included in
Appendix B); (2) the various assumptions used in the document (i.e., extensive discussions of the
interpretation of study data used in the assessment, especially neurotoxicological data); and (3) the
analytic methods applied and (4) the statistical procedures employed (i.e., explicit discussion of
modeling methodologies in the Toxicological Review and Appendix C). Further, this assessment has
been through the Interagency Science Consultation review step (Step 3 of the IRIS Process) which
includes OMB.
Comment: In the Draft IRIS Assessment, EPA has included a section titled "Preamble to the IRIS
Toxicological Reviews" that includes a summary discussion of the scope of the IRIS program,
process for developing IRIS assessments, study selection, data evaluation and derivation of toxicity
values. As currently written, the preamble offers an abbreviated view of EPA policies, guidance and
standard practices but fails to include the detail necessary to provide useful information on how the
Agency reviews or weighs the scientific information for inclusion in its toxicological review as
discussed in the NAS recommendations. (Comment 1.3, p 6)
This document is a draft for review purposes only and does not constitute Agency policy.
F-l	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
Toxicological Review ofTrimethylbenzene
EPA Response: The Preamble to the IRIS Toxicological Reviews has been developed in response to
NRC recommendations to concisely summarize EPA policies, guidance and practices employed in
IRIS assessment development. It is not intended to provide a detailed application of procedures to
the TMB isomers. Rather the Preamble is complemented by evaluation of the available scientific
information found in the body of the assessment document. The EPA will seek comments from the
external peer review panel as to the effectiveness of this structure in IRIS assessments. Comment.
Although the Draft Assessment identified a solid core set of databases to search for relevant data,
EPA has failed to conduct a thorough literature search which has resulted in the omission of data
from two TSCA 4(a) test rules fU.S. EPA. 1993.19851. The omission of the TSCA data suggests EPA
may have additionally missed other studies (Comment II. 1, p 7)
EPA Response: The studies published as a result of the 1985 TSCA 4(a) test rule (Douglas et al..
1993: Mckee etal.. 1990: Clark etal.. 1989: Schreiner et al.. 1989) were identified in the initial set of
references considered for inclusion in the Toxicological Review. However, as these studies use the
C9 fraction as the test substances, they were excluded from further consideration (see next
comment/response and Appendix E for further information).
Comment. EPA's decision to consider the TMB isomers toxicokinetically and toxicologically
equivalent was appropriate. However, given this decision, then data on any of the isomers or on
TMB-containing mixtures (predominantly TMBs with other similar hydrocarbons [e.g. C9 aromatic
including ethyltoluene] can be used to characterize the hazards of TMBs individually or collectively.
This includes the data submitted, and accepted by the EPA under TSCA Section 4(a) test rules fU.S.
EPA. 1993.19851. Inclusion of this data would greatly enhance the database available on TMB
isomers individually, and address many of the uncertainties raised in the Draft IRIS Assessment.
ACC encourages EPA to review all available data on TMBs and C9 mixtures and to reevaluate those
studies in regard to the calculations for the RfC and RfD. (Comment II.2, pp 8-10; Comment V, pp
17-18)
EPA Response: The 1985 TSCA 4(a) test rule (U.S. EPA. 1985) required that "manufacturers and
processors of the C9 aromatic hydrocarbon fraction ... test the C9 aromatic hydrocarbon fraction
for neurotoxicity, mutagenicity, developmental toxicity, reproductive effects, and oncogenicity."
EPA issued the final testing requirements that the C9 fraction be tested based on the findings that
(1) there were there no data to suggest that exposure to individual TMB isomers posed a threat to
human health, that (2) there was no evidence of substantial releases of TMB isomers to the
environment, and that (3) there was adequate data to suggest that TMB isomers would not persist
in the environment (U.S. EPA. 1985).
However, much of this information is dated and no longer correct Information does exist currently
that occupational and residential exposures to TMB isomers do occur (HSDB. 2011a. b, c; Martins et
al.. 2010: Choi etal.. 2009: Guo etal.. 2009: Tiun-Horng et al.. 2008) and that substantial quantities
This document is a draft for review purposes only and does not constitute Agency policy.
F-2	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
Toxicological Review ofTrimethylbenzene
of 1,2,4-TMB are released to the environment (TRI. 20081 (see Preface). Lastly, TMB was nominated
to the IRIS program due to its presence at Superfund sites, indicating that individual TMB isomers,
once released to the environment, are capable of persisting in the environment at contaminated
locations. Therefore, while testing the C9 fraction was originally deemed sufficient given the lack of
evidence that exposure to individual isomers of TMB was likely, current information demonstrates
that TMB isomers are released to and persist in the environment and that human populations are
exposed to TMBs in occupational and residential settings.
In the Federal Register Notice announcing the C9 fraction testing requirements, EPA agreed with
public comments that, in the absence of toxicological information on individual ethyltoluene or
TMB isomers, "assessing the toxicity of the C9 mixture as a complete entity should provide a
reasonable upper bound for the toxicity of the individual ethyltoluene and TMB [isomers] in the C9
mixture" fU.S. EPA. 19851. However, this assumption has been shown to be inaccurate given
current information. In the time since the promulgation of the C9 fraction testing requirements and
subsequent conduct and publication of the C9 fraction toxicity studies, multiple peer-reviewed
studies have been published that demonstrate that individual TMB isomers do elicit clearly adverse
toxicological effects. These include neurotoxicity fWiaderna etal.. 2002: Gralewicz and Wiaderna.
2001: Wiaderna etal.. 1998: Gralewicz etal.. 1997b: Gralewicz etal.. 1997a: Korsak etal.. 19951.
respiratory toxicity f Korsak etal.. 2000a. b; Korsak etal.. 1997: Korsak etal.. 19951. developmental
toxicity (Saillenfaitetal.. 2005). and hematological toxicity (Korsak et al.. 2000a. b). Generally, the
C9 fraction studies failed to observe clear measures of toxicity in the systems investigated. The
ultimate reason for the discrepancy between the individual isomer and C9 fraction studies is
unknown.
However, it must be acknowledged that the specific test compound used in the C9 fraction was a
complex aromatic hydrocarbon mixture containing between 45-55% TMB isomers, with the
remaining mixture primarily consisting of ethyltoluene isomers. The test compound also contained
xylene, n-propyl- and isopropylbenzene, and unspecified C10 aromatic hydrocarbon constituents.
These tertiary compounds comprised as much as 16% of the test compound. Additionally, in Clark
et al. (1989). up to 9% of the test compound was unidentified impurities. For the purposes of
setting a reference value for trimethylbenzenes, it is preferable to analyze the trimethylbenzene
isomers themselves, and not complex mixtures that include other compounds. For these reasons,
these studies were not included in the Toxicological Review. A more comprehensive discussion of
this subject has been provided in Appendix E of the Supplement Information document
Comment. The Draft IRIS Assessment states that "no chronic, subchronic, or short-term oral
exposure studies were found in the literature" for 1,3,5-TMB. This is incorrect; there are oral
toxicity studies performed by the request of EPA Office of Water Chemicals Final Test Rule (U.S.
EPA. 1993). EPA's exclusion of these studies (Koch Industries. 1995a. b) is not justified, as inclusion
This document is a draft for review purposes only and does not constitute Agency policy.
F-3	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
Toxicological Review ofTrimethylbenzene
of the studies provides direct results for oral exposure to 1,3,5-TMB in rats and does, in fact,
enhance both the hazard identification and dose response analysis. (Comment II.2, pp 10-11)
EPA Response: After careful reconsideration, EPA agrees that the 14- and 90-day oral gavage
1,3,5-TMB toxicity studies should be incorporated into the document. Accordingly, the hazard
identification and dose-response sections of the Draft Assessment have been updated to include
information on and discussion pertaining to the Koch Industries studies (1995a, b). One other
industry report investigating the oral toxicity of 1,2,4-TMB was further considered for inclusion in
the Toxicological Review (Borriston. 19831. In this study, male F344 rats (n = 10) were exposed to
either 0.5 or 2.0 g/kg 1,2,4-TMB daily for 28 days. All rats in the high dose and one rat in the low
dose group died during exposure (no times given). Other reported effects were enlarged adrenal
glands, mottled and red thymuses, and congested lungs. Given the limited toxicological information
provided by this report (other than total mortality in the high dose group), this report was not
included in the Toxicological Review.
Comment. EPA has selected decreased pain sensitivity (expressed as increased latency to response)
as the critical effect for TMB toxicity, and Korsak and Rydzynski (1996) as the principal study.
Exposure to TMB isomers resulted in an increased latency in response when measured immediately
after treatment but found no effects 2 weeks post-exposure for animals in the repeat dose study.
The most likely explanation is that exposure to TMB isomers results in acute, reversible responses.
Acute effects are related to the most recent exposures, and are not the consequence of repeated
exposures. In this regard, it is unclear how the Korsak and Rydzynski (1996) study can be selected
as the principal study. Furthermore, results for the pain sensitivity endpoint in the neurotoxicity
study with C9 aromatics (Douglas etal.. 1993) found no adverse effects in animals examined at 5, 9
and 13 weeks during and after exposure to higher levels than employed by Korsak and Rydzynski
(1996). The discussion of pain sensitivity should be revised to accurately emphasize that decreases
in pain sensitivity and increases in response latency were observed only when animals were tested
immediately after 90 days of treatment (Korsak and Rydzynski. 1996). but not when the animals
were held without treatment for any extended period of time indicating the transient nature of the
response. (Comment III, pp 11; Comment IV. 1, p 13; Comment VI.2, pl5)
EPA Response: For the reasons discussed previously, the C9 aromatics studies, including Douglas et
al. (1993). are not considered in this assessment. In the sections pertaining to the selection of the
proposed overall RfCs for 1,2,4-TMB and 1,3,5-TMB (Sections 2.1.5 and 2.2.5, respectively), a
detailed discussion of the suitability of the decreased pain sensitivity endpoint is included. This
discussion has been expanded. Specifically, the U.S. EPA's Guidelines for Neurotoxicity Risk
Assessment (U.S. EPA. 1998) do note that effects that are reversible in minutes, hours, or days after
the end of exposure and appear to be associated with the pharmacokinetics of the agent and its
presence in the body may be of less concern than effects that persist for longer periods of time after
the end of exposure (pg. 8). However, this is subsequently clarified to indicate that reversible
This document is a draft for review purposes only and does not constitute Agency policy.
F-4	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
Toxicological Review ofTrimethylbenzene
effects occurring in occupational settings may be of high concern, particularly if they diminish a
person's ability to survive or adapt to the environment fU.S. EPA. 19981 ( pg. 8); such is the case for
exposure to TMBs in occupations with potentially dangerous surroundings and/ or heavy
equipment, such as dockyard painters or asphalt workers.
As pointed out in A Review of the Reference Dose and Reference Concentration Process (U.S. EPA.
20021. "[i]t is also important to keep in mind that effects that may initially appear to be reversible
may re-appear later or be predictive of later adverse outcomes." (pg. 4-16). Additionally, the
Neurotoxicity Guidelines (U.S. EPA. 19981 state that "latent effects (those that become evident only
after an environmental challenge [e.g., in this case, footshock]) have a high level of concern." The
hot plate test is a relatively simple assessment that may not be sensitive enough to detect subtle
changes fU.S. EPA. 19981. suggesting that the large changes observed immediately after TMBs
exposure may reflect gross effects. It is possible that, at longer durations after exposure, an
environmental challenge is necessary for the more subtle perturbations that persist to become
manifest at a detectable level using this test. The latent decrements in pain sensitivity following foot
shock appear to reflect a prolongation of the numbing effects of foot shock following exposure to
TMBs weeks earlier, as the immediate increases in latency due to foot shock were unchanged by
prior TMB exposure. This indicates that some aspect(s) of the altered pain sensitivity phenotype
may fail to resolve following termination of exposure. No environmental challenge was applied in
the subchronic study by Korsak and Rydzynski (1996): such an experiment may have uncovered
similar latent responses. Conversely, the short-term TMB exposure studies testing pain sensitivity
failed to analyze hot plate latency with a foot shock challenge shortly after exposure, as these
evaluations only occurred at > 50 days post-exposure.
Uncertainty regarding the reversibility of pain sensitivity in non-shocked rats at all tested
1,2,4-TMB concentrations also exists. Reversibility of the pain sensitivity phenotype following
subchronic exposure was only tested at the highest concentration of TMBs (i.e., 1,230 mg/m3). In
multiple other tests of neurological function (including pain sensitivity following a foot shock
challenge), it has been shown that exposure to any of the TMBs isomers causes nonlinear effects
when tested some period of time after exposure, with 1,230 mg/m3 TMB routinely eliciting either
no response or a reduced response, as compared to lower TMB concentrations (e.g., 492 mg/m3).
Thus, from data available, a determination regarding the reversibility of TMB-induced decreases in
pain sensitivity at all concentrations at two weeks post-exposure cannot be made with confidence.
Although it is important to consider the potential for reversibility of neurological effects, "for
chronic lifetime exposures, designation of an effect as irreversible or reversible is academic, as
exposure is presumed to be lifetime (i.e., there is no post-exposure period)" (U.S. EPA, 2002; pg. 3-
27). Thus, the nature of an RfC precludes the possibility of recovery of the critical effect and
supports the choice of the principal study, even if all aspects of the pain sensitivity phenotype were
found to be transient (which does not appear to be the case). Taken together, the database supports
This document is a draft for review purposes only and does not constitute Agency policy.
F-5	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
38
Toxicological Review ofTrimethylbenzene
the characterization of decreased pain sensitivity associated with exposure to TMB isomers as an
effect of high concern, and an appropriate endpoint on which to base the RfC derivation. However,
EPA agrees that the observation of reversibility of the decreased pain sensitivity endpoint is an
important factor to consider. As such, EPA has determined that a full 10-fold uncertainty factor for
extrapolation from a subchronic to chronic duration is not warranted, and has instead applied a 3-
fold uncertainty factor (see discussion of uncertainty factors, below).
Comment. Although Korsak and Rydzynski (1996) was identified as the key study, significant
emphasis was placed on subsequent studies in which animals were exposed for only 4 weeks
duration and held for longer periods and foot shock was introduced fWiaderna etal.. 2002:
Gralewicz and Wiaderna. 2001: Wiaderna etal.. 1998: Gralewicz etal.. 1997b) to support a position
that the observed pain sensitivity was not an acute response but that exposure to TMB isomers
results in persistent impairment as long as 50-51 days post exposure, long after TMB had been
eliminated from the body. However, the studies actually demonstrated that pain sensitivity per se
was not persistent. Moreover, these studies show some inconsistencies in their findings:
[Note: numbering of the bullets provided as comments is used to frame the EPA responses below]
(1)	Korsak and Rydzynski (1996)... 1,2,3- and 1,2,4-TMB... tested them for pain sensitivity after
90 days of exposure... increased latency... immediately after termination of exposure...
tested the rats 2 weeks post-exposure and there were no differences.
(2)	Gralewicz etal. (1997b)...1.2.4 TMB for 4 weeks... tested at days 50-51 using the hotplate
assay and found no effects. They then shocked the animals... finding no effects. They then
tested the rats 24 hours after foot shock, finding a significant increased time to response in
the 100 and 250ppm groups.
(3)	Wiaderna et al. (1998)... 1,2,3 TMB for 4 weeks, tested them at 50 and 51 days after
exposure using a hot plate assay only and no effects were seen... after foot shock was
administered, latency [was unchanged]... when tested 24 hours after foot shock a significant
increase in latency... was found at lOOppm [only],
(4)	Gralewicz and Wiaderna f20011...1.2.3-. 1,2,4-, and 1,3,4-isomers of TMB for 4 weeks...and
then tested them on days 50-51 for pain response, finding no effects. Then they shocked the
animals and tested for pain sensitivity immediately after foot- shock and 24, 72, and 120
hours post-shock. Increased latency time was observed at 24 hours for 1,2,4 TMB and 1,3,5
TMB but... significant reductions in latency time to response were found in experiments
with 1,3,5 TMB at 72 hours post-shock and 1,2,4- and 1,3,5- at 120 hours.
(5)	Wiaderna et al. (2002.).-.1,3,5-TMB... for 4 weeks... tested on days 50-51 and found no
effects in the hot plate test and no effects immediately after foot shock or at any
intermediate point before the 240 hours post-shock assessment at which point a significant
reduction in latency time was found at all exposure levels... Results did not replicate
significant differences reported by Gralewicz and Wiaderna (20011... at 72 and 120 hours
post-shock. (Comment III, pp 11-12)
This document is a draft for review purposes only and does not constitute Agency policy.
F-6	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
Toxicological Review ofTrimethylbenzene
EPA Response: Additional details and clarifying discussions have been added to the Toxicological
Review, and are summarized here. Specifically:
(1-3) The comments submitted (above) are accurate. Immediately following 90 days of exposure,
increased latency in the hot plate test (decreased pain sensitivity) was observed fKorsak and
Rydzvriski. 19961: however, this effect did not persist 2 weeks after termination of exposure. A
statistically significant increased latency in the hot plate test was observed only 24 hr post foot-
shock at 100 or 250ppm 1,2,4-TMB and lOOppm (non-significantly increased at 250ppm)
1,2,3-TMB (Wiaderna etal.. 1998: Gralewicz etal.. 1997b).
(4&5) The data described in the submitted comments (above) do not relate to the results of
performance in the hot plate test [i.e., Fig. 4 in Gralewicz et al. (1997b): Fig. 2 in Wiaderna et al.
(Wiaderna et al.. 2002): Fig. 4 in Gralewicz and Wiaderna (Gralewicz and Wiaderna. 2001).
Rather, the evidence presented in the submitted comments reflects observations of reduced step-
down latency in passive avoidance tests [i.e., Fig. 3 in Gralewicz et al. (1997b): Fig. 1 in Wiaderna et
al. (Waderna et al.. 2002): Fig. 3 in Gralewicz and Wiaderna (Gralewicz and Waderna, 2001)1.
Importantly, although these passive avoidance tests do not directly assess pain sensitivity (these
tests are usually interpreted as measures of impulse control and memory retention), a reduction in
the latency to step down could also reflect decreased pain sensitivity to the negative reinforcement
(i.e., foot shock), as the animals may be exhibiting less fear memory of stepping down onto the
platform where they previously received what was intended to be painful foot shocks (the foot
shocks employed in these tests have a much shorter duration than those used to induce reductions
in pain sensitivity in hot plate tests). Notably, there is no use of a hot plate to detect pain sensitivity
in the passive avoidance tests. This misattribution of the passive avoidance tests as measures of
pain sensitivity is apparent when looking at descriptions of the timing of the endpoint assessment:
e.g., the comment in (4) "Then they shocked the animals and tested for pain sensitivity immediately
after foot- shock and 24, 72, and 120 hours post-shock". Pain sensitivity (the hotplate test) was
only conducted a few seconds or 24 hours after foot shock; impulse control and memory retention
(passive avoidance tests) were conducted at 0, 24, 72, and 168 hours (7 days) after foot shock.
To address the comments related to lack of consistency, the results of the hotplate tests in these
studies report an increased latency (decreased pain sensitivity) at 24 hr post foot-shock at 100
ppm, but not 250 ppm (slightly increased latency only), 1,2,3-TMB (Wiaderna etal.. 1998): at
lOOppm, but not 250ppm (slightly increased latency only), 1,3,5-TMB (Wiaderna et al.. 2002): and
at 100 ppm 1,2,4- or 1,3,5-TMB [latency increases ~75% over controls by 1,2,3-TMB were not
statistically significant; fGralewicz and Wiaderna. 20011], Thus, the results are consistent.
The text, evidence tables, and arrays relating to hot plate tests of pains sensitivity and passive
avoidance tests of cognitive function (Section 1.1.1) have been revised and expanded to more
clearly describe the results of these very different tests. The discussion of the hot plate tests, in
This document is a draft for review purposes only and does not constitute Agency policy.
F-7	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
Toxicological Review ofTrimethylbenzene
particular, now includes a greater emphasis on both the general lack of differences in pain
sensitivity observed in non-shocked rats at 50 days post-exposure as well as the lack of
inconsistencies in the decreased pain sensitivity (increased hotplate latency) at 51 days following
TMB exposure when an environmental challenge (foot shock) is applied 24 hr earlier.
Comment. Evidence of persistence in response of the pain sensitivity endpoint was found only after
foot shock administration. No agreed guidelines for study conduct and rationale for administering
foot shock were cited in the Draft Assessment and thus the varied protocols lead to a lack in clarity
regarding whether or not the testing conducted is scientifically valid and reproducible. The Draft
Assessment acknowledged that incorporation of foot shock complicates the interpretation of these
studies: "[m]ost of the neurotoxicity tests in animals incorporated the application of foot shock
which, depending on the procedure, can involve multiple contributing factors and can complicate
interpretations regarding effects on discrete neurological function." Discussions of the effects of the
neurotoxicity studies demonstrating persistence of the pain sensitivity endpoint should be
expanded to qualify that significant persistent effects were only reported after foot shock was
introduce[d], (Comment III, pp 12-13; Comment IV. 1, p 13)
EPA Response: In rats, it is well accepted that foot shock induces short-lived analgesia. This is a
scientifically valid test and a reproducible effect. In the experiments using foot shock in concert
with analyses of pain sensitivity (i.e., hotplate tests), the protocols are near-identical (i.e., % studies
used 2mA, 100ms pulses every 2 seconds for 2 minutes; the other used 4mA). Protocols employing
foot shock in passive avoidance tests (which, as stated previously, is not a test of pain sensitivity) or
active avoidance tests are different, as the stimulus is intentionally shorter. The limitations
regarding the interpretation of pain sensitivity experiments when the hot plate test is coupled with
foot shock has been clarified to focus on the pain sensitivity endpoints alone, rather than
"neurotoxicity tests",
The consistently observed effect of increased latency to paw lick 24 hours after foot shock was
reported at one or more concentrations for all isomers across studies with the exception of one
study of 1,2,3-TMB by Gralewicz and Wiaderna (2001) [effects of 1,2,3-TMB were significant in
Wiaderna et al., (1998)], where the 75% increase relative to controls was not statistically
significant. As described in the text, the most likely explanation for this finding is that prior TMB
exposure potentiates the duration, but possibly not the magnitude, of the short-lived analgesia
caused by foot shock. However, as outlined in the text, it cannot be completely ruled out that TMB
exposure may alter cognition such that contextual clues related to the sequential combination of
foot shock and hotplate testing are differentially processed between groups. Thus, control groups
may better associate the hotplate environment with the previously-applied aversive stimulus and
more quickly withdraw their paws than their TMB-exposed counterparts who may exhibit a
decreased fear response or shorter retention of that fear-associated memory. Alternatively, since
this test paradigm can cause the hotplate test apparatus to become associated with foot shock,
This document is a draft for review purposes only and does not constitute Agency policy.
F-8	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
Toxicological Review ofTrimethylbenzene
inducing stress-related responses in the shocked animal such that subsequent exposure to the hot
plate test apparatus alone can reduce sensitivity to pain (possibly via the release of endogenous
opiods), prior TMBs exposure could amplify this effect. These possible alternative explanations
underlie why the responses were indicated as difficult to interpret as effects on a discrete
neurological function (e.g., on pain sensitivity or memory alone). Importantly, despite the possible
overlap between contributing neurological processes in this test paradigm, these observations are
still regarded as significant and adverse, and clearly indicate a persistence of neurological effects
long after TMB exposures have ceased.
Comment. Increased clarity is needed regarding selection of the critical effect for derivation of the
reference values for the TMB isomers. In discussion at the Listening Session [August 1st, 2012] it
was stated that IRIS used the "step down" technique to develop the assessment This appears to be
incorrect as the document itself indicates pain sensitivity is the key endpoint. If the "step down"
data are key, then EPA should consider revising the Draft IRIS Assessment as this distinction is not
currently clear from the document (Comment VI.1, p 14)
EPA Response: In the discussion at the Listening Session, EPA stated that the public comments
included erroneous descriptions of data relating to tests of passive avoidance (i.e., decreased step
down latency) as measures of pain sensitivity; specifically, as the previous comments reflect,
decreases in step down latency (in passive avoidance tests of cognition) were interpreted by the
commenters as inconsistent with the observations of increased paw lick latency (in hot plate tests
of pain sensitivity). EPA has not stated that the results of the passive avoidance tests (i.e., decreased
step down latency) were used as the key endpoint. Rather, these "step down" data have been
clarified by EPA as distinct from those resulting from pain sensitivity assays and that the results of
these two different tests were complementary rather than inconsistent (see comments above for
details). Revisions to the text have been made and additional clarifying information is now included
in the evidence tables (Section 1.1.1.) to more clearly portray the findings from, as well as the
differences between, these two, distinct test paradigms, and to more transparently convey the lack
of inconsistencies in the conclusions drawn from the results of each.
Comment: Gralewicz and Wiaderna (2001) reported large individual differences in each group in
step down latency for pain sensitivity and foot shock. "In order to reduce the with-in group
variability, data from two rats with the lowest and highest mean step-down latency in the first post
shock trial were excluded from data sets for each group of rats". This suggests it was necessary to
adjust the data to get significance in the Gralewicz and Wiaderna (2001) study raising further
questions about the suitability of these data for risk assessment purposes. (Comment VI. 1, p 14)
EPA Response: The comment relates only to data derived from tests of passive avoidance (cognitive
effects), and not to data from tests of hotplate behaviors (pain sensitivity). No corrections were
indicated by Gralewicz and Wiaderna (2001) in regards to the hot plate tests. The data used by EPA
This document is a draft for review purposes only and does not constitute Agency policy.
F-9	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
Toxicological Review ofTrimethylbenzene
for quantitative dose-response analyses are from tests of hotplate latency, not passive avoidance;
the suitability of the pain sensitivity data is not questioned in the above comment
As to the specific interpretation of the passive avoidance tests performed by Gralewicz and
Wiaderna (2001), the modification cited above does not appear to apply to the significance of the
observations of decreased step-down latency at day 7 after the foot shock: "Statistical comparisons
of the data from all animals revealed differences between groups in trial 6, i.e. on day 7 after the
footshock (F(4,282) = 2.86, P < 0.05); in the MES [1,3,5-TMB] group the step-down latencies were
significantly shorter than in the C [control] group. In order to reduce...". However, because the
modified analysis quoted in the comment above was somewhat unclear in the paper by Gralewicz
and Wiaderna (2001). EPA has decided to revise the evidence tables to reflect that the data
presented graphically appear to represent groups with excluded rats, drawing uncertainty
regarding statistical significance. Thus, the indication of statistical significance at 7 days after foot
shock for 1,3,5-TMB is the only significance indicator that will remain in the evidence tables
(Section 1.1.1.; the modified statistical analyses are now included as notes only), as this analysis, at
least, was clearly based on all animals tested. As the direction and approximate magnitude of these
responses remain consistent across the database, this clarification does not substantially change
EPA's interpretation of the results of the passive avoidance tests.
Comment. In developing the RfC for 1,3,5 TMB, IRIS chose to discount the developmental toxicity
study performed by Saillenfait et al. (2005) as the key study even in the absence of adequate
neurotoxicity data for this isomer (i.e., neurotoxicity data from an appropriate sub-chronic or
chronic study). EPA should carefully consider the study which provides the most robust response
on which to base the RfC derivation for 1,3,5-TMB. (Comment VI.1, p 14)
EPA Response: The RfC derivation section contains an extensive discussion of the developmental
and maternal toxicity endpoints observed in the Saillenfait et al. (2005) study, and the Draft
Assessment has been revised so that candidate RfCs based these effects are derived for 1,3,5-TMB:
1 mg/m3 based on decreased maternal weight gain and 7 mg/m3 based on decreased male and
female fetal body weight. The most sensitive RfC derived from 1,3,5-TMB-specific data is 20-fold
higher than the RfC derived for 1,2,4-TMB based on neurotoxicity data (1 mg/m3 vs. 5x10-2
mg/m3). The RfC section for 1,3,5-TMB also includes an extensive discussion of the toxicokinetic
and toxicological similarities between 1,2,4-TMB and 1,3,5-TMB, especially the similarities in
toxicity between the isomers observed in short-term neurotoxicity studies. It appears that the
major factor driving the derivation of an RfC for 1,3,5-TMB that is so much greater than the RfC for
1,2,4-TMB is the lack of a subchronic 1,3,5-TMB neurotoxicity test, and not some intrinsic difference
in toxicity between the two isomers. Given the observed similarities in toxicity and toxicokinetics
between the two isomers, EPA concluded that it was not scientifically justified to derive an overall
RfC value for 1,3,5-TMB that is so much higher than derived for 1,2,4-TMB. As such, the decision to
adopt the overall RfC value for 1,2,4-TMB (based on decreased pain sensitivity) as the RfC for
This document is a draft for review purposes only and does not constitute Agency policy.
F-10	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
Toxicological Review ofTrimethylbenzene
1,3,5-TMB is retained in the Draft Assessment The candidate RfC values for 1,3,5-TMB based on
maternal and developmental effects are presented alongside the overall RfC value for comparison
purposes, and for potential further uses such as subsequent cumulative risk assessments that
assess the combined effect of multiple agents acting at a common site.
Comment. EPA applies an uncertainty factor of 10 to account for extrapolation from subchronic
exposure to chronic exposure (UFS) based on the "assumption that effects observed in a similar
chronic study would be observed at lower concentrations for a number of possible reasons,
including potential cumulative damage occurring over the duration of the chronic study or an
increase in the magnitude or severity of effect with increasing duration of exposure." However, the
critical effect observed in the principal study (Korsak and Rydzvriski. 1996) does not demonstrate
any cumulative damage from exposure to TMB as effects are not seen two weeks after exposure is
terminated. In consideration of the fact that pain sensitivity is reversible upon termination of
exposure, EPA should consider a UFS of 3 or less. (Comment VI.2, pp 14-15)
EPA Response: After careful consideration, EPA agrees that a full 10-fold UFs is not supported by the
available data. Given the observation of reversibility in neurotoxicity endpoints reported in
subchronic inhalation studies, an uncertainty factor of 3 has been applied in the Draft Assessment
Lowering the UFS to 1 was not supported as, in the case of neurotoxicity endpoints, chronic
exposure may overwhelm the adaptive responses observed after termination of subchronic
exposure, resulting in a lack of reversibility for the pain sensitivity endpoint at 1,230 mg/m3, a
greater magnitude of this response, and/ or manifestation of more severe latent responses
associated with this effect Additionally, hematotoxicity endpoints were also observed to exhibit
reversibility, and the inflammatory nature of observed respiratory effects suggests that adaptive
mechanisms may alleviate these effects following termination of exposure. Therefore, a UFS of 3
was also applied to these endpoints.
Comment: In determining the uncertainty factor for database deficiencies (UFd), EPA cites the
absence of multi-generation and developmental neurotoxicity studies for all three isomers as
contributing to the rationale for application of a 3-fold UFd. Inclusion of the available 3-generation
C9 fraction study (Mckee etal.. 1990) and Aromatol (blended C9 aromatic hydrocarbon mixture)
developmental neurotoxicity study (Lehotzky etal.. 1985) would provide sufficient data to
overcome any deficiencies in the developmental/reproductive area and eliminate the need for any
additional uncertainty factors to account for database deficiencies, reducing the uncertainty factor
to 1. (Comment VI.3, pp 15-16)
EPA Response: Given the decision to exclude the C9 fraction studies from the Draft Assessment (see
above, and Appendix E), the McKee et al. (1990) study has been removed from the discussion
regarding the selection of the UFd. As explained above and in Appendix E, the C9 fraction studies
were excluded from the Draft Assessment because they are complex solvent mixtures that at most
This document is a draft for review purposes only and does not constitute Agency policy.
F-ll	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
37
Toxicological Review ofTrimethylbenzene
only contain 55% TMB isomers. Thus, there is considerable uncertainty regarding how their
compositional make-up influences the observed general lack of C9-induced developmental toxicity
compared to the individual TMB isomer study (which does observe developmental toxicity
following exposure to either 1,2,4-TMB or 1,3,5-TMB). The Lehotzky et al. (1985) study was
excluded based on the same rationale. Therefore, as the C9 fraction studies have been excluded
from the Draft Assessment, the lack of TMB isomer-specific multigenerational reproductive and
developmental toxicity and developmental neurotoxicity studies remains a weakness of the TMB
database.
Comment. The Panel agrees that the database for TMBs provides "inadequate information to assess
carcinogenic potential" of these isomers. The database for TMBs, however, supports the likelihood
that TMBs are not mutagens and are unlikely to be genotoxic carcinogens. In the only study
investigating TMB-induced genotoxicity, only 1,2,3-TMB was reported to elicit positive results in
the Ames assay (Tanik-Spiechowicz etal.. 19981. The observation that 1,2,3-TMB was genotoxic in
the absence of metabolic activation and the manner in which the data were presented call into
question the conclusion of positive mutagenicity for this particular isomer. Further, although Janik-
Spiechowicz et al. (1998) reports increased sister chromatid exchange in the bone marrow of male
mice following exposure to each individual TMB isomer, no alterations in the frequency of
micronucleus formation was noted. As micronucleus formation is a definitive endpoint for
cytogenetic damage, this indicates that clastogenicity is not expressed following exposure to TMB
isomers. Consideration of the available C9 fraction mutagenicity study (Schreiner etal.. 1989)
supports the conclusion that TMB isomers are not likely to be mutagens. (Comment IV.4, pp 16-17)
EPA Response: There is only one available study (Tanik-Spiechowicz etal.. 1998) that investigates
the mutagenic potential of individual TMB isomers. The EPA concludes in the Draft Assessment that
this study provides at best limited information regarding the mutagenic potential of TMB isomers,
and that the database is inadequate to conclude that any isomer is directly genotoxic. In the absence
of any further evidence that individual TMB isomers do not result in gene mutations or
chromosomal aberrations, a definitive conclusion that TMB isomers are not mutagenic is not
currently supported.
Comment: The most useful study for the determination of the RfC is Clark et al. (1989). a one year
inhalation study in rats at doses of 450, 900 and 1800mg/m3. This study provides a longer duration
of exposure and the outcome is consistent with the 90 day inhalation study of 1,2,3 TMB fKorsak et
al.. 2000bl. and the 90 day oral toxicity study of 1,3,5-TMB fKoch Industries. 1995bl. The 90 day
neurotoxicity study with C9 aromatics (Douglas etal.. 1993) which was performed at higher doses
than Clark et al. (1989) and evaluated standard neurotoxicity endpoints; motor activity, functional
observation battery including the hot plate latency response [without foot shock] at 5, 9 and 13
weeks of exposure is also useful as supporting information as no adverse effects were identified.
(Comment V, p 17)
This document is a draft for review purposes only and does not constitute Agency policy.
F-12	DRAFT—DO NOT CITE OR QUOTE
-------
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
30
31
32
33
34
35
36
Toxicological Review ofTrimethylbenzene
EPA Response: As discussed above and in Appendix E, the available C9 fraction studies have not
been included in the Draft assessment for a number of reasons. Primarily, the C9 fraction is a
complex mixture containing at most 55% TMB isomers. Currently, it is unclear why the results of
the C9 fraction studies disagree with the results of individual TMB isomer studies, although the
possibility exists that interactive effects between the constituents of the C9 fraction and biological
systems alter the ADME of TMB. Therefore, the Clark et al. (1989) study is not suitable as the basis
for the derivation of RfCs values for the TMB isomers. Therefore, the methodology used in the
assessment to identify the RfCs for the isomers (i.e., derivation of RfC values for 1,2,4-TMB and
1.2.3-TMB	using isomer-specific data, and setting the RfC for 1,3,5-TMB equal to the RfC for
1.2.4-TMB	based on toxicological and toxicokinetic similarities between the isomers) is retained in
the Draft Assessment The sections outlining the derivation of the RfC for each individual TMB
isomer have been thoroughly edited to more clearly delineate the process by which the values were
derived.
Comment: For the RfD determination the 90 day oral study with 1,3,5 TMB fKoch Industries. 1995bl
is preferable to extensive extrapolation from inhalation data. Results have been accepted by EPA to
characterize the hazards of 1,3,5 TMB. Reliance on this study would obviate the need for
pharmacokinetic analysis and route to route extrapolation. The more extensive data base
accompanying this study reduces the uncertainties identified with the current investigation and
avoids reliance on studies with interpretational difficulties. Furthermore, since IRIS acknowledges
the similarity in toxicological responses among the TMB isomers, an RfD based on animal data for
1,3,5 TMB could reasonably be extrapolated to the other 2 isomers. (Comment II.2, pp 10-11;
Comment V, p 17)
EPA Response: As stated above, discussion of the 90-day oral gavage Koch Industries (1995b) study
has been added to the Draft Assessment, and it was considered as a possible principal study on
which to derive an RfD. However, although the Koch Industries (1995b) study was submitted to
EPA under a TSCA 4(a) test rule, it had not undergone an independent external peer review. As
stated in Section 3.1 of the Preamble, "[i]f a study that may be critical to the conclusions of the
assessment has not been peer-reviewed, EPA will have it peer-reviewed. As such, EPA sought an
independent external review of the Koch Industries (1995b) study by three experts in
neurotoxicology, human health risk assessment, and general laboratory animal toxicology studies
(Versar. 2013). All three external reviewers concluded that the Koch Industries (1995b) study was
well-written, followed GLP or standard protocols (with only minor deviations) for the time period
in which the study was conducted (i.e., mid-1990s), and that the conclusions of the study were
supported by the reported findings. However, two reviewers specifically commented that Koch
Industries (1995b) study was not an appropriate study on which to base the derivation of a
reference dose for a number of reasons (detailed below). The third reviewer, while not explicitly
This document is a draft for review purposes only and does not constitute Agency policy.
F-13	DRAFT—DO NOT CITE OR QUOTE
-------
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
Toxicological Review ofTrimethylbenzene
stating the study was not suitable for RfD derivation, did provide comments that addressed
multiple shortcomings of the study.
Two reviewers questioned the human relevancy of the chosen route of exposure (oral gavage), with
one reviewer noting that, as the toxicity of 1,3,5-TMB was investigated due to it being a water
contaminant, exposure via drinking water would be preferable over exposure via gavage. Further,
this reviewer noted that the dosing regimen of the Koch Industries (1995b) study (5 days/week)
was not optimal as toxicokinetic studies demonstrate rapid clearance of TMB and its metabolites (<
24 hours). Dosing only 5 days a week results in 48 hours of non-exposure and extended clearance;
this reviewer suggested a dosing regimen of 7 days/week would have been more appropriate. One
reviewer expressed strong concern that the NOAEL identified in the study was most likely an
artifact of the study investigating insensitive endpoints (i.e., body weights, gross pathology). This
reviewer expressed confidence that a lower NOAEL would have been identified had the study
investigated endpoints "more pertinent to human health" (e.g., "behavioral, respiratory, or
electrophysiological" endpoints). A second reviewer commented that, as demonstrated by the
available peer-reviewed literature on TMBs, neurotoxicity is a critical endpoint for the evaluation of
TMB-induced toxicity. This reviewer ultimately concluded that the Koch Industries (1995b) study is
not reliable "for assessing noncancer risk, because the endpoint of concern for TMB exposure,
neurotoxicity, was not evaluated". This reviewer acknowledged that although clinical signs were
observed, these markers of effect were "too general to be predictive of neurotoxicity". This
reviewer notes that although the Koch Industries (1995b) study could be used to quantitatively
derive an RfD, the endpoint of concern (neurotoxicity) may not be protected against.
Given the result of the external peer review noted above, and the critical shortcomings of the Koch
Industries (1995b) study (no testing for neurotoxicity and the general lack of any other observed
toxicity), this study has limited utility for the derivation of an RfD for 1,3,5-TMB. Therefore, the
methodology used in the assessment to identify an RfD for 1,3,5-TMB (i.e., setting the RfD equal to
the RfD for 1,2,4-TMB based on toxicological and toxicokinetic similarities between the isomers) is
retained in the Draft Assessment. The sections outlining the derivation of the RfD for each
individual TMB isomer have been edited to more clearly delineate the process by which the values
were derived.
This document is a draft for review purposes only and does not constitute Agency policy.
F-14	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
REFERENCES FOR APPENDICES1
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.
http://www.acgih.org/Store/ProductDetail.cfm?id=1311
Bakke. OM: Scheline. RR. (1970). Hydroxylation of aromatic hydrocarbons in the rat. Toxicol Appl
Pharmacol 16: 691-700.
Battig. K: Grandiean. E: Rossi. L: Rickenbacher. I. (1958). Toxicologische untersuchungen uber
trimethylbenzol. Archiv fuer Gewerbepathologie und Gewerbehygiene 16: 555-566.
Battig. K: Grandiean. 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: Levnaert. 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/i.envres.2011.02.008
Borriston (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
Brown. RP: Delp. MP: Lindstedt. SL: Rhomberg. LR: Beliles. RP. (1997). Physiological parameter values for
physiologically based pharmacokinetic models [Review], Toxicol Ind Health 13: 407-484.
Cerf. I: Potvin. M: Laham. S. (1980). Acidic metabolites of pseudocumene in rabbit urine. Arch Toxicol 45:
93-100.
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.Org/10.1136/oem.56.6.383
Chevron (Chevron Chemical Company). (1985). One generation reproduction study of PED 5450 in rats
with cover letter. (OTS0206739). Washington, DC: U.S. EPA.
Choi. DW: Moon. KW: Bveon. SH: Lee. EI: 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
Clark. DG: Butterworth. ST: Martin. IG: Roderick. HR: Bird. MG. (1989). Inhalation toxicity of high flash
aromatic naphtha. Toxicol Ind Health 5: 415-428.
Collins. AS: Sumner. SCI: Borghoff. SI: Medinsky. MA. (1999). A physiological model for tert-amyl methyl
ether and tert-amyl alcohol: Hypothesis testing of model structures. Toxicol Sci 49: 15-28.
Cooper. SP: Burau. K: Sweeney. A: Robison. T: Smith. MA: Svmanski. 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: Mauderlv. IL: Rothenberg. SI: Seiler. FA: Mcclellan. RO. (1988). Uptake of 19
hydrocarbon vapors inhaled by F344 rats. 10: 262-269. http: //dx.doi.org/10.1016/Q272-
0590(88)90310-7
1 Multiple references published in the same year by the same author(s) have been assigned a letter
(e.g., 1986a, 1986b) in these Supplemental Material Appendices (and independently in Volume 1 of the
Toxicological Review), based on which publication's list of authors, and then title, comes first alphabetically.
This document is a draft for review purposes only and does not constitute Agency policy.
R-l	DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review ofTrimethylbenzene
Deurenberg. P: Weststrate. IA: Seidell. IC. (1991). Body mass index as a measure of body fatness: age- and
sex-specific prediction formulas. Br J Nutr 65:105-114.
Douglas. IF: Mckee. RH: Cagen. SZ: Schmitt. SL: Beattv. PW: Swanson. MS: Schreiner. CA: Ulrich. CE:
Cockrell. BY. (1993). A neurotoxicity assessment of high flash aromatic naphtha. Toxicol Ind Health 9:
1047-1058.
Dowty. Bl: Laseter. IL: Storer. I. (1976). The transplacental migration and accumulation in blood of volatile
organic constituents. Pediatr Res 10: 696-701. http://dx.doi.Org/10.1203/00006450-197607000-
00013
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
Emond. C: Krishnan. K. (2006). A physiological pharmacokinetic model based on tissue lipid content for
simulating inhalation pharmacokinetics of highly lipophilic volatile organic chemicals. Toxicol Mech
Meth 16: 395-403. http://dx.doi.org/10.1080/15376510600860474
Fiserova-Bergerova. V. (1983). Gases and their solubility: A review of fundamentals. In Modeling of
Inhalation Exposure to Vapors: Uptake, Distribution, and Elimination. Boca Raton, FL: CRC Press.
Fukava. Y: Saito. I: Matsumoto. T: Takeuchi. Y: Tokudome. S. (1994). Determination of 3,4-
dimethylhippuric acid as a biological monitoring index for trimethylbenzene exposure in transfer
printing workers. Int Arch Occup Environ Health 65: 295-297.
http://dx.doi.Org/10.1007/BF00405692
Gage. IC. (1970). The subacute inhalation toxicity of 109 industrial chemicals. Br J Ind Med 27:1-18.
http://dx.doi.Org/10.1136/oem.27.l.l
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/S0161-813Xr00100003-6
Gralewicz. S: Wiaderna. D: Tomas. T. (1997a). 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: Tomas. T: Rydzvriski. K. (1997b). 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
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.Org/10.llll/i.1600-0668.2008.00580.x
Harlan Laboratories. (2012). Sprague Dawley: Outbred Rat. Available online at
http: / /www.harlan.com /products and services/research models and services/research models by
product type/outbred rats/sprague dawley sd (accessed June 4, 2012).
Hissink. AM: Krtise. I: Kulig. BM: Verwei. M: Muiiser. H: Salmon. F: Leenheers. LH: Owen. DE: Lammers. IH:
Freidig. AP: McKee. RH. (2007). Model studies for evaluating the neurobehavioral effects of complex
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/i.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
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 ofTrimethylbenzene
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
Ichiba. M: Hama. H: Yukitake. S: Kubota. M: Kawasaki. S: Tomokuni. K. (1992). Urinary excretion of 3,4-
dimethylhippuric acid in workers exposed to 1,2,4-trimethylbenzene. Int Arch Occup Environ Health
64: 325-327. http://dx.doi.org/10.1007/BF00379541
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.
lanasik. B: lakubowski. M: lalowiecki. P. (2008). Excretion of unchanged volatile organic compounds
(toluene, ethylbenzene, xylene and mesitylene) in urine as result of experimental human volunteer
exposure. Int Arch Occup Environ Health 81: 443-449. http: / /dx.doi.org/10.1007 /s00420-007-Q233-9
lanik-Spiechowicz. E: Wvszyriska. K: Dziubaltowska. E. (1998). Genotoxicity evaluation of
trimethylbenzenes. Mutat Res Genet Toxicol Environ Mutagen 412: 299-305.
http://dx.doi.org/10.1016/S1383-5718r97100202-7
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. (1999). Physiologically based modeling of 1,2,4-trimethylbenzene inhalation
toxicokinetics. Toxicol Appl Pharmacol 155: 203-214. http://dx.doi.org/10.1006/taap.1998.8596
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/S0048-9697r97105482-X
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
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
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.Org/10.1016/i.ihazmat.2008.01.022
lones. K: Meldrum. M: Baird. E: Cottrell. S: Kaur. P: Plant. N: Dvne. 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/annhvg/melOl6
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.
Korsak. Z: Rydzvriski. 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: Rydzvriski. K: laite. I. (1997). Respiratory irritative effects of trimethylbenzenes: An
experimental animal study. Int J Occup Med Environ Health 10: 303-311.
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 ofTrimethylbenzene
Korsak. Z: Stetkiewicz. I: Maicherek. W: Stetkiewicz. I: laite. I: Rvdzvnski. 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: Maicherek. W: Stetkiewicz. I: laite. I: Rydzvriski. K. (2000b). Subchronic inhalation
toxicity of 1,2,3-trimethylbenzene (hemimellitene) in rats. Int J Occup Med Environ Health 13: 223-
232.
Korsak. Z: Swiercz. R: Rvdzvhski. 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.
Kostrewski. P: Wiaderna-Brvcht. 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/Q378-
4274(95)03305-X
Kostrzewski. P: Wiaderna-Brvcht. 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-
9697f97105504-6
Laham. S: Potvin. M. (1989). Identification and determination of mesitylene acidic metabolites in rabbit
urine. Toxicol Environ Chem 24: 57-69. http: //dx.doi.org/10.1080/027722489Q9357477
Lammers. IH: Emmen. HH: Muiiser. H: Hoogendiik. 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.Org/10.1016/i.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.
Lehotzky. K: Szeberenvi. IM: Gv. U: Kiss. A. (1985). Behavioural effects of prenatal exposure to carbon
disulphide and to aromatol in rats. Arch Toxicol Suppl 8: 442-446.
Lutz. P: Gralewicz. S: Wiaderna. D: Swiercz. R: Grzeliriska. Z: Maicherek. 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: IIdx.doi.org/10.1111 /i.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: Lammers. IH: Muiiser. 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
Mckee. RH: Wong. ZA: Schmitt. S: Beattv. P: Swanson. M: Schreiner. CA: Schardein. IL. (1990). The
reproductive and developmental toxicity of High Flash Aromatic Naphtha. Toxicol Ind Health 6: 441-
460.
Meulenberg. C: Viiverberg. 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.200Q.8929
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 ofTrimethylbenzene
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/10.1016/0Q41-
008Xr75190048-4
MOE (Ontario Ministry of the Environment). (2006). Rationale for the development of Ontario air
standards for trimethylbenzenes: 1,2,3-Trimethylbenzene. Ontario, Canada.
NIOSH (National Institute for Occupational Safety and Health). (1988). Testimony of the National Institute
for Occupational Safety and Health on the Occupational Safety and Health Administration's proposed
rule on air contaminants, 29 CFR Part 1910, OSHA Docket No. H020. Presented at the OSHA informal
public hearing, August 1,1988. NIOSH policy statements. Cincinnati, OH.
NIOSH (National Institute for Occupational Safety and Health). (1992). NIOSH recommendations for
occupational safety and health: Compendium of policy documents and statements. (92-100).
Cincinnati, OH: U.S. Department of Health and Human Services, Public Health Service, Centers for
Disease Control, National Institute for Occupational Safety and Health. http: //www.cdc.gov/Niosh/92-
100.html
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/aiim.47002006Q4
NRC (National Research Council). (2011). Review of the Environmental Protection Agency's draft IRIS
assessment of formaldehyde. Washington, DC: National Academies Press.
http: / / www, nap, edu / catalog /13142 .html
Pwkko. K. (1980). Effects of methylbenzenes on microsomal enzymes in rat liver, kidney and lung.
Biochim Biophys Acta 633: 1-9. http://dx.doi.org/10.1016/0304-4165(80)90032-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/i.fct.2005.02.008
Schreiner. CA: Edwards. DA: Mckee. RH: Swanson. M: Wong. ZA: Schmitt. S: Beattv. P. (1989). The
mutagenic potential of high flash aromatic naphtha. Cell Biol Toxicol 5:169-188.
Sulkowski. Wl: Kowalska. S: Matvia. W: Guzek. W: Wesolowski. W: Szvmczak. 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: Rydzynski. K: Wasowicz. W: Maicherek. 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.
Swiercz. R: Wasowicz. W: Maicherek. 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.
Swiercz. R: Wiaderna. D: Wasowicz. W: Rydzynski. 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.
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.
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 ofTrimethylbenzene
TRI (Toxic Release Inventory). (2008). Toxic Release Inventory [Database]: U.S. Environmental Protection
Agency.
Tsuiimoto. Y: Noda. T: Shimizu. M: Moriwaki. H: Tanaka. M. (1999). Identification of the dimethylbenzyl
mercapturic acid in urine of rats treated with 1,2,3-trimethylbenzene. Chemosphere 39: 725-730.
Tsuiimoto. 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/S0045-6535r99100467-l
Tsuiimoto. Y: Warashina. M: Nam. VP: 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.
Tsuiino. Y: Hieda. Y: Kimura. K: Eto. H: Yakabe. T: Takavama. K: Dekio. S. (2002). Distribution ofkerosene
components in rats following dermal exposure. Int J Legal Med 116: 207-211.
http://dx.doi.Org/10.1007/s00414-001-0282-7
U.S. Congress. (2011). Consolidated Appropriations Act, 2012. (Pub. L. No. 112-74; 125 STAT. 786). 112th
U.S. Congress. http: //www.gpo.gov/fdsvs/pkg/PLAW-112publ74/pdf/PLAW-112publ74.pdf
U.S. EPA (U.S. Environmental Protection Agency). (1985). Identification of specific chemical substance and
mixture testing requirements; Ethyltoluenes. trimethylbenzenes, and the C9 aromatic hydrocarbon
fraction. Fed Reg 50: 20662-20677.
U.S. EPA (U.S. Environmental Protection Agency). (1987). Health effects assessment for trimethylbenzenes
[EPA Report], (EPA/600/8-88/060). Cincinnati, OH.
http://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=2 000T8ZG.txt
U.S. EPA (U.S. Environmental Protection Agency). (1988). Reference physiological parameters in
pharmacokinetic modeling [EPA Report], (EPA/600/6-88/004). Washington, DC: U.S. Environmental
Proctection Agency. http: //www.ntis.gov/search/product.aspx?ABBR=PB88196019
U.S. EPA (U.S. Environmental Protection Agency). (1993). Office of water chemicals; Final test rule. 40 CFR
Part 799 [0PPTS-42111C; FRL 4047-2] RIN 2070-AB94. Fed Reg 58: 59667-59682.
U.S. EPA (U.S. Environmental Protection Agency). (1998). Guidelines for neurotoxicity risk assessment
[EPA Report], (EPA/630/R-95/001F). Washington, DC.
http://www.epa.gov/raf/publications/pdfs/NEUROTOX.PDF
U.S. EPA (U.S. Environmental Protection Agency). (2002). A review of the reference dose and reference
concentration processes [EPA Report], (EPA/630/P-02/002F). Washington, DC: Risk Assessment
Forum, U.S. Environmental Protection Agency.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=51717
U.S. EPA (U.S. Environmental Protection Agency). (2007). Acute exposure guideline levels (AEGLs) for
1,3,5-trimethylbenzene (CAS reg. no. 108-67-8), 1,2,4-trimethylbenzene (CAS reg. no. 95-63-6), 1,2,3-
trimethylbenzene (CAS reg. no. 526-73-8) [EPA Report], Washington, DC.
http://www.epa.gov/opptintr/aegl/pubs/123 %20124 %20135 trimethylbenzenes %20interim 11
2007.vl.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2011a). PBPK data files for TMB [PBPK], Research
Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (2011b). Results of the BMD analyses for 1,2,4-TMB,
1,2,3-TMB, and 1,3,5-TMB [BMDS], Research Triangle Park, NC.
U.S. EPA (U.S. Environmental Protection Agency). (2012a). Benchmark dose technical guidance.
(EPA/100/R-12/001). Washington, DC: Risk Assessment Forum.
http://www.epa.gov/raf/publications/pdfs/benchmark dose guidance.pdf
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 ofTrimethylbenzene
U.S. EPA (U.S. Environmental Protection Agency). (2012b). EPA announces NAS' review of IRIS
Assessment development process. Available online at
http://yosemite.epa.gOv/opa/admpress.nsf/0/lce2a7875daf093485257a000054df5470penDocumen
t
Ungvarv. G: Tatrai. E. (1985). On the embryotoxic effects of benzene and its alkyl derivatives in mice, rats
and rabbits. Arch Toxicol 8: 425-430.
Versar (Versar Inc.). (2013). Peer review report: External peer review of the 1995 Koch Industries study
report: 90-day oral gavage toxicity study of 1,3,5-trimethylbenzene in rats with a recovery group. (EP-
C-12-045). Springfiled, VA: Versar, Inc.
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. (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.
Wiglusz. R: Delag. G: Mikulski. P. (1975a). Serum enzymes activity of mesitylene vapour treated rats. Bull
Inst Marit Trop Med Gdynia 26: 303-313.
Wiglusz. R: Kienitz. M: Delag. G: Galuszko. E: Mikulski. P. (1975b). Peripheral blood of mesitylene vapour
treated rats. Bull Inst Marit Trop Med Gdynia 26: 315-321.
Williams. LR: Leggett. RW. (1989). Reference values for resting blood flow to organs of man [Review],
Clinical Physics and Physiological Measurement 10:187-217. http: //dx.doi.org/10.1088/0143-
0815/10/3/001
Yoshida. T. (2010). Estimation of absorption of aromatic hydrocarbons diffusing from interior materials in
automobile cabins by inhalation toxicokinetic analysis in rats. J Appl Toxicol 30: 525-535.
http://dx.doi.org/10.1002/iat.1522
Zahlsen. K: Eide. I: Nilsen. AM: Nilsen. OG. (1992). Inhalation kinetics of C6 to C10 aliphatic, aromatic and
naphthenic hydrocarbons in rat after repeated exposures. Pharmacol Toxicol 71: 144-149.
http://dx.doi.Org/10.llll/i.1600-0773.1992.tb00534.x
Zahlsen. K: Nilsen. AM: Eide. I: Nilsen. OG. (1990). Accumulation and distribution of aliphatic (n-nonane),
aromatic (1,2,4-trimethylbenzene) and naphthenic (1,2,4-trimethylcyclohexane) hydrocarbons in the
rat after repeated inhalation. Pharmacol Toxicol 67: 436-440. http: //dx.doi.0rg/lO.llll /i.1600-
0773.1990.tb00859.x
This document is a draft for review purposes only and does not constitute Agency policy.
R-7	DRAFT—DO NOT CITE OR QUOTE
-------